AN INDEPENDENT SUPPLEMENT FROM MEDIAPLANET BUSINESS INTELLIGENCE, DISTRIBUTED IN THE TIMES

17 SEPTEMBER 2007

Bridging the gap between intelligence and information

BUSINESSINTELLIGENCE

AN INDEPENDENT SUPPLEMENT FROM MEDIAPLANET ABOUT BUSINESS INTELLIGENCE, DISTRIBUTED IN THE TIMES2

IntroductionBUSINESS INTELLIGENCE

CONTENTS

Data base trends p. 4

Top tips p. 6

Visit mathematics p. 8

BT case study p. 9

Software strategies p.11

Virgin media case study p.12

Data visualisation p.13

Goliaths still need the Davids p. 14

Meta Data p. 15

Collaborate to compete and comply p. 17

BUSINESS INTELLIGENCEA TITLE FROM MEDIAPLANET

Project Manager: Tonje Loew HansenAssociate Project Manager: Dominic RodgersProduction Editor: Katherine WoodleyJournalist: Rod NewingDesign: Sophie WesterbergPrepress: Jez MacBeanPrint: News International

Mediaplanet is the leading Europeanpublisher in providing high qualityand in-depth analysis on topical industry and market issues, in print,online and broadcast.

For more information about supple-ments in the daily press, please contactHenrik Kanekrans,

020 7563 8873henrik.kanekrans@mediaplanet.com

www.mediaplanet.com

Business intelligence (BI) continues tobe the most important strategic ini-tiative in many organisations. The re-sults from Gartner’s CIO survey 2007showed that, for most of the 1,400 in-terviewed chief information officers(CIOs), BI is not only in the No.1 spot,but had the largest increase in prior-ity of all technologies for the secondyear running. This is a major shiftfrom the situation in 2001, when BIdidn’t even make the top-ten list.Why? There are several reasons:

First, while there is less money forgrowing the business, the CIO is ex-pected to increasingly contribute togrowing the top line and the bottomline. BI is one technology where, ifimplemented correctly, contributes todriving the business forward andmaking it more agile. In a tough busi-ness environment, companies are re-alising that BI can generate realcompetitive advantage, by driving theright decisions at the right time.

Second, BI is often imposed by ex-ternal factors such as compliancewith regulatory requirements and en-suring that auditors won’t find anyskeletons in the data-managementclosets.

And finally, CIOs are under pres-sure to create value from the systemsthey themselves implemented such asenterprise resource planning (ERP) orcustomer relationship management(CRM).

So how much of the budget is allo-cated to BI these days? A Gartner sur-vey of European companies indicatedthat most organisations aim to allo-cate around 5-10 per cent of theirsoftware budgets to BI. This includesdata warehousing, reporting andanalysis tools, data integration, dataquality, or metadata management, butalso hardware upgrades, professionalservices, and training. CIOs are look-ing to invest in strategic technologytools that can help them deliver in-formation and projects that will helpdifferentiate their businesses compet-itively rather than implementingcommodity technology solutions.

To stay on top of the fast-changingBI landscape, organisations need totrack many developments that influ-ence how BI is consumed. For exam-ple, while many organisations alreadysuffer from information overload, in-

dustries such as manufacturing, retail,and distribution have yet to reachthat stage, as radio-frequency identi-fication (RFID) has not become main-stream. Meanwhile, the convergenceof so-called structured and unstruc-tured data (from databases to com-plex data stores for rich mediacontent), the introduction of searchtechnology, the provisioning of BI “asa service”, data-warehouse appli-ances, open-source technology andhardware developments such as 64-bit architectures all will have an im-pact on today’s BI environments.

Companies that fail to rise to thechallenge of making BI a strategicpriority and give it an organisationalbackbone in form of a business intel-ligence competency centre (BICC),will find themselves seriously disad-vantaged in their respective markets.

In addition and perhaps surprising

is how indifferent businesses can betowards the critical issue of data qual-ity. For many organisations, data isconsidered largely a by-product oftheir business applications, and there-fore implicitly considered correct andof good quality – a big mistake. Or-ganisations must manage data as acorporate asset, with stewards ap-pointed with responsibility for thequality of data, and various technolo-gies, such as profiling, matching,cleansing, or enrichment, being de-ployed throughout the business.

Poor data quality can dramaticallyjeopardise any BI initiative and oftenhas a significant negative impact onthe business. After all, there is novalue in a report that looks great, isnicely designed, numbers all adjusted,corporate colours and logos in place,but the data in the report is wrong. By Andreas Bitterer

Business intelligence must be a strategic priority

Peeping under the BI bonnetWe have done the business issues, now we are looking at the technology, butwill unstructured data be next?

Business intelligence (BI) is a very complex businessstrategy that can be very simply described as the useof interactive end-user software to provide peoplewith the information they need to make decisions thatcontribute to the achievement of corporate objectives.It has been around in various forms for nearly 25years, yet for the second year in a row, Gartner’s 2007survey of chief information officers showed that BIapplications are still the highest technology priorityfor business.

This high level of interest in business intelligence isreflected in the fact that this is the third supplementwe have published in twelve months. On September14th 2006 and March 26th 2007 we took a very comprehensive look at the business issues surroundingthis powerful strategic business tool.

Given that this technology has been around for solong, these business issues are essentially unchanged.For this reason, we have decided this time around to

‘lift the bonnet’ to take a look at the technologies thatdrive BI, at the vendors who supply the systems and atthe industry in which they operate. Taken together, thethree reports will give readers a very complementaryand comprehensive view of BI.

The only issue that has changed in the last twelvemonths is an emerging interest in combining struc-tured and unstructured data. Anybody who set out todesign a business intelligence solution from scratchwould have included both - and unstructured infor-mation management is often referred to as ‘businessintelligence.

However, the BI industry has been firmly centred onstructured data, although combining it text and otherunstructured data has been talked about for 20 years.Several BI vendors have introduced relevant functionality,but so far it hasn’t taken off. We will watch developmentsvery carefully and we hope to report positive news of‘true business intelligence’ in our next supplement.

What if inflation rises rapidly? What if the market spikes suddenly?

What if you counter aggressively? If you had Applix, you wouldn’t just

know the answers. You’d already know what to do with them. With our

analytics software, you’ll be able to instantly create and rapidly evaluate almost any

possible business scenario. Knowing the consequences of your actions before you take them

will enable you to improve your decision making – and give you a serious edge over your

competition.Visit applix.com to learn how we can turn your business into an agile enterprise.

Handleanything.

© 2007 Applix, Inc. All rights reserved. Applix is a registered trademark of Applix, Inc.

Highly complexplanning, budgetingand risk assessment

You

AN INDEPENDENT SUPPLEMENT FROM MEDIAPLANET ABOUT BUSINESS INTELLIGENCE, DISTRIBUTED IN THE TIMES4

Data base trendsBUSINESS INTELLIGENCE

Absolutely fantastic!Data warehouses are costly and hard to manage, butthe cost is coming down and the effort can be worth it.A data warehouse is a relational data-base, optimised to run queries, whichcontains an integrated set of accurateinformation transferred from a rangeof underlying transaction processingsystems. It is based on subject, usu-ally customers, but including opera-tions, human resources and financial.This ‘single version of the truth’ feedsany number of end-user applications,

often through an intermediate ‘datamart’ containing summarised data.

“There are a lot of big data ware-houses, but they are typically ‘creak-ing at the seams,’ because they have somuch highly detailed data,” says Ali-son Whitby, BI practice head atMorse, a consulting company. “Somehave line item detail on orders goingback five years. Such detail fills up

the warehouse without providing us-able information. Organisations needto retain summaries and archive thedetails.”

Another common problem withdata warehouses is maintaining anacceptable level of data quality. Prob-lems arise from input errors, missingdata, duplication, etc. “Even the largerglobal companies have not paid

enough attention to data quality,”says Elaine Fletcher, business intelli-gence practice manager at LogicaCMG. “Departments can lose patienceand trust and go outside for their ownlocal application, so you lose thesingle source of truth.”

The key is to put in business rulesto align the different sources into asingle view. “When that is done, thebig data warehouses are absolutelyfantastic,” she says. “They give youan accurate profile of every customerthat lots of different consumers in theorganisation can feed from.”

Microsoft has worked hard to reduce the cost of building a datawarehouse with its SQL Server data-base, but the market has been shakenup by a rash of new vendors with alternative solutions. Companies likeNetezza and DATAllegro have devel-oped data warehouse appliances.These bundle commodity hardwareand storage; open source operatingsystems and databases; and proprietarysoftware to improve performance.

“Data warehouses don’t always perform well enough, particularly forcomplex analytics, and they take anage to install and get running,” saysPhilip Howard, research director atBloor Research. “The ‘plug in and go’appliances provide outstanding performance at a price that beats thedatabase vendors. They have broughtthe price point of warehouses and

data marts down. People can now dothings that they couldn’t previouslycost-justify. It has been disruptive tothe market, but it has been good forcustomers.”

Ms Whitby argues that organisationsdon’t have to be big and to have thelatest tools to use BI intelligently. Morsecarries out benchmarking studies andknows that systems in some largesuccessful organisations are not the‘latest or the greatest.’

“They are using them much moreintelligently,” she says. “They haveput processes in place to ensure thatthey are using the information theyhave to the maximum, not just stuff-ing it full data. Small companies canbe quite nimble if they use what theyhave very effectively.”

There has long been talk of theneed to integrate the structured datain a data warehouse with unstruc-tured data, such as text, wordprocessing files and presentations.“These often carry vital information yetare only available to a small number ofpeople,” says David Hobbs-Mallyon,SQL Server product manager atMicrosoft. “The industry is taskedwith adding greater value to data ware-houses be ensuring that unstructureddata is stored, accessed, searched,collated and consumed. A lot oforganisations find that how they treattheir data can differentiate them fromeach other.”

In-Memory analytics comes of ageSpeed delivers real business value, says Martin Richmond-Coggan

Projects where the product was selected on the basis ofquery performance have consistently delivered morebusiness benefits than those that selected products basedon functionality. This is one of the findings of the OLAPReport (www.olapreport.com/Performance.htm).

The authoritative web site advises organisations to treat fast perform-ance as a critical success factor, notan after-thought. “Make sure that fastquery performance is one of your key

buying criteria,” it says, “and take advantage of 64-bit applications thatwill perform better if more data isheld in RAM, rather than having to beaccessed from disk.”

Data has traditionally been held inmagnetic form, physically distributedacross a hard disk located well awayfrom the electronics in a computer.When queries are performed the computer’s central processor has tosend instructions to the hard disk,which then has to move a mechanicallaser into the right position to read thedata and then send it back to the cen-tral processor. This is a time consuming process that causes notice-able delays to users, who are condi-tioned by the almost instant responsesfrom massive Google searches.

An alternative approach that hasbeen available for the last 23 years is toload the entire database into random access memory (RAM). This isthe high speed electronic memory on acomputer chip attached directly to thecomputer’s central processor. From hereit can be instantly accessed without de-lays caused by distance and mechanicalmovements. According to Monash In-formation Services, it takes up to 1 mil-lion times as long to find a random byteof data on disk, as it might take to findthe same byte in RAM.

Until recently, computers were limited to 3 gigabytes of RAM bytheir 32 bit architecture. Now thatcomputers use a 64 bit architecture,they have unlimited RAM. The avail-ability of large amounts of inexpensivememory has made it possible to putvery large databases in RAM. Applixcustomers are now running their ap-plications with 64 gigabytes of RAMand we expect some to move to 128gigabytes by the end of the year.

Users expect the same almost instant response from their BI appli-

cations that they get from theirspreadsheets, which they can achievewith interactive memory-based ap-plications. The OLAP Report remindsreaders that Google and games consoles set today’s expectations forusers.

“Slow query performance has beenconsistently the most serious product-related problem reported,” it says. “Ifusers grow impatient with slowqueries, they are more likely to printor copy query results into local Excelspreadsheets or may do less analysis,because they can’t afford to spend thetime or they lose their train ofthought.”

Speed is important for developingthe application, when results can beinstantly tested, and for operations,such as entering month-end journals.A disk-based system could take tenminutes or even need to be updatedovernight.

We see in-memory analytics as anincreasing source of competitive advantage, which is reflected in thefact that Applix is growing at 40 per

cent a year, whilst the rest of the mar-ket is growing at 16 per cent. Ourtechnology has now come of age.

As Gartner puts it, ‘demand for fastqueries against big datasets, coupledwith lower-priced 64-bit computing,will increase the use of in-memorytechnology. By 2012, 70% of Global1000 organisations will load detaileddata into memory as the primarymethod to optimise BI applicationperformance.’

Martin Richmond-Coggan is European vice-precident at Applix

Martin Richmond-Coggan

O I T U K LTDMember of the Genus Group

Hammond Close • Nuneaton • Warwickshire • CV11 6RY • U.K.Telephone (024) 7625 4940 • Facsimile (024) 7638 2319

E-mail sales@oituk.com • Web-Site www.oituk.com

With its award-winning C-Cube Solutions software platform, OITUK has been providing document and information management

solutions to the public and private sector for over a decade.

OITUK’s C-Cube, Electronic Document Management Systems (EDMS) allow organisations to manage their information effectively and

efficiently, delivering savings in terms of staff time and storage issues.

With clients including the Metropolitan Police, NHS trusts and local authorities, OITUK has a proven track

record of delivering solutions that store, protect and manage information, allowing organisations to grow and

develop.

C-CUBE SOLUTIONS FROM OITUK... THE ANSWER TO INFORMATION AND DOCUMENT MANAGEMENT ISSUES

All successful businesses, regardless of the sector in which they operate,have reached their position in the cut-throat world of commerce by reactingswiftly and intelligently to meet the needs of customers or changes in marketconditions. In other words there is no room for complacency; complacencyequals catastrophe, and many big beasts of the business world have fallenfrom grace for resting on their laurels.

No business of any size or scope would like to be accused of beingcomplacent in any of its operations, but when it comes to the issue ofinformation and records management, this can all too often be the painfultruth.

It seems to be that the UKgovernment and public sectorhas taken a lead in addressingthe issues of information andrecords management seriously,making efforts to develop acoherent strategy, which iscurrently being led by theNational Archives as a

spokesman points out.“The vast majority of government records are now produced electronicallyand the National Archives works with central and local governmentorganisations to ensure that this information is both stored securely andeasily accessible. It is also working on improving its processes for archivingsuch records.”

The message from the National Archives seems to be that it is essential tograsp the issue of information and document management before it becomesan unmanageable issue. “With the switch to use of electronic records, TheNational Archives is looking to improve its processes and procedures withregard to appraisal, selection, transfer, storage, sustainability and delivery. Ithas instigated a programme of work under the Seamless Flow banner to bringincreased automation to these areas. The National Archives also adviseslocal government, which has similar responsibilities for online and FreedomOf Information,” the spokesman added.

UGLY SISTERS AND CINDERELLASWhen it comes to the myriad of strategic and operational issues facing abusiness on a day-to-day basis, information and record management seemto come in low on the list of priorities according to Vijay Magon, TechnicalDirector of information systems provider, OITUK: “All too often managers anddirector’s heads are turned by the more dazzling ‘Cinderella’ type web-basedapplications at the expense of the crucially important, but less glamorous‘Ugly Sister’ record and information management systems

“In essence, executives may be devoting many hours to the style of thecorporate website, while crucial information is left vulnerable to loss ormisuse by malicious elements like hackers.”

INFORMATION, INFORMATION, INFORMATIONAt a time when the amount of information and records being dealt with bybusinesses grows year by year, in the form of documents, emails and faxes,the need for businesses to embrace a strategic policy towards records andinformation management grows ever more pressing.

Information is the lifeblood of any company, and methods of storage andretrieval should be seen as a commercial imperative, not as an obligation,says Vijay.

“The collation, indexing and retrieval of key commercial records andinformation are at the very core of most organisations operations. Thereforeit seems odd that the method of how this process is carried out receives solittle attention. A cost-effective records and information management systemcan save hundreds if not thousands of staff hours which are spent looking fordocuments. With staff salaries making up a large percentage of operatingcosts, it can only make sense to have your staff spending their timeefficiently,” he said.

In order for attitudes to change, it is necessary for businesses to seeinformation and records management as something which can aid anddevelop their business, and not an obligation to be shouldered. In essence,they need to take a pro-active approach, not a reactive approach to recordsand information management.

Said Vijay: “Information and records management systems are often seen asnothing more than ‘electronic filing cabinets’, where documents are storedand can then be retrieved. In fact they are much more dynamic than this.Using an Electronic Document Management System (EDMS) means thatdocuments are ‘born’ on the system when they are first created and thenmove through the system where they can be amended, updated or stored forfuture retrieval. Therefore a document has an auditable ‘lifecycle’ which canbe traced.”

In terms of efficiency savings and administrative time saved, this can soonpay dividends.

WHEN THE SKY FALLS Nobody expects the sky to fall in, buton Sunday, December 11, 2005,that’s just what happened, with themassive explosion at the BuncefieldOil Depot in Hemel Hempstead

in Hertfordshire. Many businesses on an industrial estate adjacent to thedepot were severely damaged along with many residential homes. But theeffects were felt much further afield than that.

Cambridge-based Addenbrookes Hospital had its remote servers whichsupport its IT functions located in premises which were affected by the blast.These were severely disrupted for several days. Despite the loss of thesesystems, hospital staff were able to fall back on an EDMS installed by OITUK,and the hospital remained fully functioning.“This episode starkly reveals the very real need for businesses andorganisations of all sizes to consider how they protect their information; withthe right support in place, they can ensure their businesses have theprotection they need, says Vijay.

“It is not too dramatic to suggest that any business which lost the criticalmass of its information would be unable to remains commercially viable forvery long in the face of such a loss.”The overriding message is that while disasters on the scale of Buncefielddon’t happen everyday, businesses should none the less guard againstburying their heads in the sand in the belief that ‘It could never happen to us.’

READY FOR THE NEXT ACT?Since Enron and other huge business scandals rocked the commercial world,the US, UK and other governments from across the world have enacted awhole new tranche of legal requirements regarding how businesses and otherorganisations store and manage their information. Negotiating this minefieldcan be tricky, but EDMS can alleviate many of the compliance issues, as Vijayexplains.

“Businesses irrespective of size and sector are faced with a plethora of legalobligations. If they contravene these they can face fines or legal action. Thegood news is that EDMS can ensure that businesses or organisations cankeep their house in order and that they remain compliant. For example, EDMScan electronically identify and track documents, alerting company staff as tohow long documents need to be legally stored and when they should bedestroyed in order to remain compliant.”

NO MAGIC BULLET, JUST COMMON SENSEUnlike other aspects of information technology used by organisations andbusinesses, EDMS stretches across all aspects of a business from accountsto human resources. Therefore the system needs to tick a large number ofboxes to fulfil its overall objectives.

There is no magic bullet solution to meeting these criteria, just a commonsense approach that ensures the solution delivers what is expected of it.OITUK have been helping businesses to achieve tangible benefits throughcareful and selective implementation of this long established technology.

Overall, it seems that a little time, planning and expenditure on a viableinformation management policy can increase a business’ efficiency, secure itagainst possible disaster and ensure it remains free from legal wrangles, allof which enable it to remain successful and very definitely not complacent.

There is no magic bullet solution to meeting thesecriteria, just a common sense approach that ensuresthe solution delivers what is expected of it. OITUKhave been helping businesses to achieve tangiblebenefits through careful and selectiveimplementation of this long established technology.

“”

BE COMPLIANT, NOT COMPLACENT

Mike Hawes, head of office of thechief financial officer at SAPUnited Kingdom and Ireland

Due to their continued expansion, many growing businessesface the frustration of spending more time then they wantcapturing and inputting data and delivering reports, thus lim-iting the time for analysis and forecasting the future. In orderto transform the way organisations handle their reporting

processes, performance management software is becoming increasingly focused on au-tomating the process of data management, thus allowing more time for decision mak-ing and the subsequent evaluation of data.

Business process flows (BPFs) are a relatively new technology feature of performance management software. They give organisations the power to automate re-porting by monitoring and enforcing enterprise-wide information and technologyprocesses. They do this by leveraging shared approval flows, process and workflow track-ing throughout the whole of the performance process.

Businesses have the choice of selecting pre-packaged functions or tailoring the BPFsto their individual needs. This focus on automation and controlling common businesstasks and reporting processes can reduce reporting time and improve business analysisand insight. This will mean more time can be spent on analysing company data andidentifying industry opportunities and changes.

Andrew Stevenson, head of centres of excellence, Atos Origin

When designing any user interface to a new BI system, makesure that you focus on the business functionality that you aregiving users rather than the technology or tool functionality.

Few users fully understand all of the functionality of theubiquitous Microsoft office toolset. All they really want is toproduce a spreadsheet, write a report or put together a presen-

tation pack with some nice graphics. The same should be true of any BI solution, where allusers really want to do is get answers to their questions in as easy a manner as possible.

With this in mind, they shouldn’t have to learn a new technology or tool or even un-derstand the fundamentals of the data model that underpins the solution. Ideally, accessshould be either through their existing tools, which they are familiar with and use allthe time, or through an easy-to-use portal. A portal can provide ‘a single place to go towork’ that provides access to electronic mail and key applications, as well as to businessintelligence.

This design principle allows any future upgrades, migrations or software tool changesto be made behind the scenes without users being aware. This removes a great deal ofusers’ reluctance to change and increases the flexibility of the organisation to make full useof new software developments or enhanced architectures, such as a service oriented archi-tecture. It would even enable outsourcing management and operation of their data miningand analytical solutions to third parties.

Sayinath A.G., BI practice head atSonata Software

Some vendors offer frameworks, also called ‘blueprints’ or‘accelerators,’ that provide best practice in performance meas-urement. They define standard key performance indicators(KPIs) for an industry or function and are built on top of anextensible data model that provides pre-defined ‘canned’ re-ports. They are either based upon existing industry standardsor the vendor’s experience of building similar BI solutions.

In most BI implementations, irrespective of the size, gathering requirements and de-ploying a solution that fits the needs of the business, both in terms of content and userexperience, is a major challenge. Adopting such a framework helps users to visualise theend solution and IT to implement it.

Demonstrating a framework gives users a graphic representation of the actionable intel-ligence provided by the KPIs. They can then direct their requirements in an organisedway. This approach accelerates the BI implementation cycle, helps the IT function topromote best practices inside the business and builds consensus on standard perform-ance measurement metrics. It substantially mitigates the risk of building a solution thatmisses the business objective by a significant mile and helps build the base for a consis-tent and high performance organisation. Adopting a framework driven approach can helpto solve the BI needs of the business.

Foster Hinshaw, chief executive ofDataupia

So many conversations I have with businesses are about thehassles of managing mountains of data. They focus on thenegative aspects, such as the cost, the drain on resources andthe compromises they have to make. Nevertheless, they havelarge quantities of data that they need to utilise in order tocreate and respond to business opportunities.

The secret is to focus on the opportunity the data offers,rather than the problems of managing it. Find out what users most want and start giv-ing it to them. Sales executives can be given the transaction history for each of their ac-counts, which they can share with the customer service representatives. It is a simple dataset that involves no complex queries, but it can make all the difference to the business.

Making even the simplest data broadly available depends on a data warehousing in-frastructure that is flexible and scalable. There are steps you can take now to getready for mainstream data warehousing and limit changes to dependent BI applica-tions. These will in turn cut down on testing and deployment time. The result will bea data warehouse that quickly serves up new information when the business side getsits next ‘big idea!’

AN INDEPENDENT SUPPLEMENT FROM MEDIAPLANET ABOUT BUSINESS INTELLIGENCE, DISTRIBUTED IN THE TIMES6

News From The ExpertsBUSINESS INTELLIGENCE

Delivering exceptional business intelligence solutions to our clients.

icitbi.com

Data Quality Resource CentreA complimentary online serviceDownload white papers, analyst reports, product overviews, application briefs, case studies, interactive product demonstrations, web seminar replays and podcasts.

http://eu.informatica.com/dq2

Meeting the compliance deadline for Basel II through an enterprise-wide data quality initiative M&S Money is a leading UK financial services company. A wholly owned subsidiary of global banking leader HSBC, it retains a close working partnershipwith UK-based Marks & Spencer, has a very similar customer profile and collaborates on customer activities. Like all financial services providers, M&S Moneyhad the challenge of readying its information environment to comply with the Basel II Capital Accord by 1 January 2007. In doing so, the company wishedto accomplish a number of operational benefits, namely the assurance of customer data accuracy, reduction of time taken to complete business analysisand increase assuredness in the quality of data from which Business Intelligence is created.

M&S Money planned and delivered a data quality projectto ensure compliance with Basel II, which necessitates thetracking and reporting of ongoing exposure to credit,market and operational risks, and introduced improveddata management practices across the company.Headquartered in Chester, M&S Money offers a wide rangeof financial products including the &More credit card,travel money and various insurance policies. It needed tomeet Basel II’s requirements in order to ensure that itscredit risk management practices met internationallyrecognised levels of consistency and therefore enabledmeasurement of the adequacy of its capital.

Besides the compliance requirements, there were broaderdata management benefits that the company wished toaccomplish. These centred on the need to avoid potentialcustomer frustration caused by inaccurate data, enablingthe company to undertake business analysis and, therefore,sales activities with greater confidence. In the company’sown words, it wanted to ensure that customers were given“no surprises” when approached about new products andservices. M&S Money’s attitude to selling to Marks andSpencer customers has always been that if it was notcompletely confident about data accuracy, and therefore acustomer’s likely interest, it would not approach them at all,rather than risk any frustration. Enhancing data qualitywould therefore drive improved business analysis andsales practices.

The ChallengeM&S Money has grown significantly since it was foundedin 1985 and has a longstanding reputation for the qualityand reliability of its products. Compliance with Basel IItherefore represented a special challenge for M&S Moneyas the changes needed to its information environmentalso had to provide a platform for optimum levels ofcustomer service and management.

Basel II is a demanding piece of legislation that requiresextensive expertise and sophisticated data managementcapabilities if the quality and integrity of data is to beconsistently assured. M&S Money therefore needed todeliver a data quality initiative that ensured data wasmanaged according to a set specification throughout thewhole information environment, imposing quality controlsat many points throughout the architecture while retainingcentralised control and management.

“We cannot ever risk upsetting a customer because ofproblems with our data,” said Neil Hershaw, InformationManagement Officer for M&S Money. “Basel II’s compliancedeadline presented us with both a challenge and anopportunity – delivering high quality data to ensuresuccessful risk management, but also improving our datato drive improvements in many other areas of our operations.”

The Solution“There was obviously the need to satisfy the data qualityelements of Basel II, but also taking this enhanced overallapproach to data management would give us greaterconfidence in our activities to create sales and therefore helpdrive competitive advantage in the market,” said Mr Hershaw.

The primary operational driver was that M&S Moneyneeded to be constantly ‘aware’ that its data was of theappropriate quality and that it could always demonstrateclear management of the data. Informatica Data Qualityand Data Profiler were deployed as a data qualitymanagement platform alongside the datawarehouse toensure that data feeds comply with the required qualitylevels ensuring that BI data remains wholly accurate.Equally, the solution had to support M&S Money’s newdata management procedures that were introduced aspart of the project, so that any changes which couldnegatively impact quality would be immediately addressedand nullified. In creating this calibre of data qualitymanagement platform, the company also had to considerthe need to assure the sustained accuracy of data derivedfrom third party feeds, such as those from insurancepolicy providers. Again, Informatica Data Quality and DataCompiler provided the required support for thesedisparate data feeds and therefore enabled all dataquality to be managed within a single framework.

The methodology for the project began with defining dataquality rules for the relevant files and tables, coding thoserules, running the data, then analysing the data andcreating an action plan for quality assurance. Having nowcompleted the project, data quality levels are evaluatedformally on a quarterly basis.

The ResultsThe main driver for the business was compliance withBasel II’s stipulations, which was completed successfullya year ahead of the deadline. In doing so, the companyhas created an assured level of data quality that has

enabled it to undertake business analysis and financialreporting with greater confidence. A further gain has beenthe reduction of the analysis cycle by up to 40 per cent,meaning faster comprehension and validation ofinformation through less time needing to be spentassessing the accuracy of the data.

M&S Money’s IT team is now able to provide quantitativemeasurement of the data held by the organisation, whichwas a broader goal of the initiative. For a financial servicescompany, that resource has proven to be invaluable inproviding the ‘engine room’ that drives business development.

Once Informatica Data Quality and Data Profiler went live,it took business analysts just four days to develop 20Basel II business rules on the fly, deliver a data accuracyscorecard, create profiles on account history tables anddevelop other business rules that were then added to thescorecard.

“Basel II has been a blessing in disguise. Improved dataquality and the ability to measure that quality hasimpacted our business for the better in several areas,”said Mr Hershaw. “It is clear that business analysis hasbenefited, but also there are factors such as how weevaluate our sales and marketing. We are now able toascertain with confidence what activities have caused a sale.

“We live by the mantra that customers must be presentedwith no surprises, but we also experienced no surprises ofour own in meeting the requirements of Basel II. It was atough assignment but the detailed planning and executionenabled us to minimise the scale of the challenge. We’renow even more confident that we will always deliver onour commitments to customers and can assure high levelsof data quality to assist in running our business,” he said.

Contact DetailsInformatica Software LimitedFor all enquiries please contact Suzanne Rozier Tel: +44 (0) 1628 511 311 (extension 332)Email: srozier@informatica.com Web: www.informatica.com/uk

The Value of Data

The question, let alone the answer, getsmore complicated if you start to involvedata over time. That is, how much morevaluable is a year’s worth of data com-pared to six months? Or three years’worth? Clearly, in many situations, moredata is more valuable. For example,suppose that you wish to forecast salesof one or more items over the Christmasperiod. The only logical way to do thatwould be to be able to compare previousChristmases, not just with each other, butwith trading patterns in the periods leadingup to these Christmases. According toSAS you need a minimum of 3 yearsworth data to be able to make meaningfulpredictions about Christmas trading.

Another problem with estimating thevalue of data is that the value can bederived from third parties. For example,if you are a Telco and you store call datarecords (as you have to) you can storethese off-line, in near-line storage oron-line. But only on-line storage willenable the police and security servicesto do real-time searches against tele-phone records. Now suppose that such asearch enables the prevention of a bombplot. How much is that worth? Actually,that’s not quite a rhetorical question: itused to be estimated that a foiled IRAbomb plot was worth £1m. But that wasawhile ago and inflation needs to beapplied to that figure. Moreover, the IRA

used to give warnings about theirbombs while Al-Queda does not, so thepotential loss of life from a successfulplot could be much greater and thevalue of thwarting such an attempt allthe greater. We can, of course, all see thebenefits of this but there is no direct busi-ness value to the Telco who is fundingthe system per se.

So putting a value on data is very hard.Nevertheless, it should be clear thatwhenever analysis of data is required,then the more the better. However, thisflies in the face of what most companieshave been trying to do for a number ofyears. As data volumes have grown theemphasis in many organisations has beenon information lifecycle managementand archival as a means of reducing thecosts of keeping data on-line. But it isarguable that this is self-defeating.Certainly you save costs by archivingolder data. On the other hand, theinsight that you (or third parties in thecase of call data records) can gain willbe enhanced from being able to analyse

more information gathered over alonger time. So the question becomes:how much more valuable is this greaterinsight and how does that compare withthe additional costs of storage?

Moreover, the equation is changing, ascompanies such as Dataupia enter themarket and allow you to store more dataon-line at lower cost (and with a loweradministrative overhead) then thischanges the tipping point between thevalue of holding more data on-line fordeeper analytical insight, as opposed tothe cost savings of archival (whichbecome smaller). The truth is that I don’tknow where that tipping point is, and inany case it will vary not just be industrybut by company, but it is certainlysomething that enterprises should beaware of and considering.

How do you put a value on data? Clearly this is difficult toanswer. Apart from anything else it depends on context. Forexample, knowing that you have a stock level of x is uselessunless you at least know what typical turnover for that stock is.

By Philip HowardDirector of Research, Bloor Research

For information about Dataupia please visit:www.Dataupia.com or email them at info@dataupia.com

AN INDEPENDENT SUPPLEMENT FROM MEDIAPLANET ABOUT BUSINESS INTELLIGENCE, DISTRIBUTED IN THE TIMES8

Visit MathematicsBUSINESS INTELLIGENCE

Data mining for the massesHas data mining’s time finally come?

Data mining software uses complexmathematical algorithms to identifyhidden patterns or relationships inlarge volumes of data. It is used forapplications like customer segmenta-tion, shopping basket analysis, cross-selling, prediction, churn reduction,credit scoring and fraud detection.

“We are seeing a massive upturn indata mining,” says Elaine Fletcher,practice manager for business intelli-gence at LogicaCMG. “We have beendoing it for years, but it has always

been a very specialist activity andonly one in eight BI clients wanted it.Over the last year nearly everybodywants data mining.”

She explains that when peopleslice, dice and drill-down they canunderstand where they are. “Theywon’t necessarily understand whythey are there or what they should doto improve it,” she says, “which iswhat data mining will answer.”

The underlying technologies havechanged little in 20 years. The main

trend has been repeated attempts tomove it away from being a tool forstatistical experts to become moreeasily usable by general businessusers. A few years ago the BI vendorswere buying up small specialist companies and offering ‘data miningfor the masses.’

Data mining is available eitherthrough workbenches, suites of toolsdesigned for experts to carry out a widerange of activities, or as applications,which are more limited in scope and

easier for general users. Gareth Her-schel, research director at Gartner, ex-plains that workbenches offerenterprises the greatest potential fordiscovering unique insights unavailable to competitors, whilst applications deliver rapid return oninvestment. He expects most organi-sations will eventually rely on a com-bination.

An example is market basket analysis.“Ten years ago in most companies itwas being used by data mining spe-cialists or the statisticians,” saysPhilip Howard, research director atBloor Research. “It has become soprevalent and so well understood that ithas been taken out of that environmentand embedded into applications.”

Data mining used to be carried outon a separate computer, often byexpert third parties who may addexternal data. Although it can now bedone inside the database, which ismore powerful, the outsourcing modelis still attractive. “It is popularbecause the data mining skill set isquite specialist,” says Graeme Smith,solution architect at Morse, “and not alot of companies have enough projectsto support a full time role.”

Business intelligence must eventuallyembrace unstructured data, especiallytext, which has been discussed for 20years. Fortunately, Mr Howard is nowseeing data and text mining comingtogether. “You need to combinenumerical and textual information in

AN INDEPENDENT SUPPLEMENT FROM MEDIAPLANET ABOUT BUSINESS INTELLIGENCE, DISTRIBUTED IN THE TIMES 9

Case Study BUSINESS INTELLIGENCE

Dataupia is thepower behindpolice system

Focus Data Services has selected theDataupia™ Satori Server to achieve unprecedented access to data storedin its Focus 112 Disclosure System(112). The 112 system is used bytelecommunications companies acrossthe United Kingdom to process policeand emergency service requests forCall Detail Record (CDR) data and information.

Focus chose the Dataupia SatoriServer to serve as the underlying architecture of the 112 system due toits scalability, ease-of-use, cost-effectiveness and unique ability to optimise legacy data systems. Datau-pia has allowed Focus to eliminate disparate data sources and gain accessto greater volumes of data. Addition-ally, the Dataupia Satori Server has improved the performance of inquirytimes, making it easier and more affordable to process queries.

The Dataupia Satori Server datamana-gement system is an all-in-onesolution – server, storage, and optimisation software packaged as asingle appliance – designed specificallyto deliver persistent access to as muchdata as an organisation needs.

order to get a clear idea of what isgoing on,” he says. “It is still in its infancy, but it has lots of potentialand we are starting to see companiestaking it seriously.”

The latest service pack for Microsoft’s SQL Server database includes a data mining applicationthat works through the Excel spread-sheet. David Hobbs-Malyon, the com-pany’s SQL Server product manager,says that it is exciting to see endusers, who wouldn’t have thought theycould use such power, start using thefunctionality to discover whether theyshould take data mining seriously.

“Data mining tools were expensiveand complicated, but the new gener-ation is helping to overcome thesechallenges,” says Nakis Papadopou-los, joint chief executive at IMGroup,a systems integrator. “Organisationsare starting to understand the valueof data mining and next year we expect it to become more common-place, giving businesses greater insight.”

Banishing the liesBT is benefiting from a group-wide business intelligence strategy

“This arises because transaction sys-tems come with their own function-ality; because different autonomousparts of the organisation seek solu-tions to their own particular needs;and through acquisition. Each indi-vidual system represents a sort of atruth, but when you have many truthswhat you actually have is many lies!”

This fragmentation makes it diffi-cult to answer simple questions, likehow many customers are there orhow many broadband lines BT sup-plies? There are different answers de-pending on how the terms ‘customer’or ‘broadband line’ are defined.

BT is one of the world’s leadingproviders of communications solutionsand services operating in 170 coun-tries. Its principal activities includenetworked IT services; local, nationaland international telecommunicationsservices; and higher-value broadbandand internet products and services. BThas four main lines of business: BTGlobal Services, Openreach, BT Retailand BT Wholesale. It has revenues over£20 billion and more than 100,000employees.

Three years ago the companyrecognised that it needed to refocuson both reducing its costs and im-proving customer satisfaction. “Thepiecemeal approach of internal or-ganic growth of departmental BI sys-tems was not the right way to addressan overall business intelligence re-quirement,” says Mr Griffiths, “wehad to develop and migrate towards along-term architected strategy.”

Following a large-scale review ofthe market, the company selected Or-acle Business Intelligence, EnterpriseEdition (previously Siebel Analytics)as its strategic tool. This would workin association with core data ware-houses and data marts that alreadyran on Oracle databases.

“It meets the vast majority of ourrequirements,” says Mr Griffiths, “butyou can’t have just one end-user tool.We use other toolsets in specialistareas, such as SAS for statisticalanalysis. We had always used Hyper-ion products within finance for plan-ning, budgeting, reporting and

performance management, so Oracle’spurchase of Hyperion was verytimely. We now have a single vendorfor the majority of our work.”

The company already had sixlarge-scale data warehouses, eachcontaining up to 100 terabytes ofdata. However, each was built for a veryspecific purpose, such as marketing, callanalysis for revenue protection, net-work performance, financial andhuman resources. As part of a ‘feder-ated’ approach they have been re-structured and expanded to makethem multifunctional. Most now sup-port long-term analysis, marketing,routine reporting and data mining.

“We are already seeingbetter customer datathan we had before,

which increases our levelof customer service”In parallel with the BI strategy is a

master data management strategy, di-vided between customer and productdata, with a very keen focus on dataquality. “We are getting better sourcesof data and allowing users to get ac-cess to it from BI applications,” MrGriffiths explains. “We are alreadyseeing better customer data than wehad before, which increases our levelof customer service.”

BT has been slowly rolling its new BIplatform out across the organisation,focusing on new projects or when asignificant rework is required to anexisting system. “We don’t yet have astrategy to remove everything that isthere, but we do have quarterly tar-gets for the number of closures,” saysMr Griffiths. “It is a continual battleto bring them into the overall strate-gic framework. This year we need toclose about 60 systems and it will ac-celerate next year. The overall targetis to get down to 20-25 BI systems bythe end of 2008.”

BT’s overall strategy is allowing itto extend BI into new areas, such as

using real-time business intelligenceto assist inbound customer supportagents. It analyses a large amount ofdata about that customer from thewarehouse and then makes a set ofrecommendations to the agent.

If the customer has suffered recentfaults they might be made an appro-priate offer or if they have recentlyplaced an order they might be auto-matically redirected to specialistagents.

This is particularly important forthe ‘customer save agents,’ who arecharged with retaining customers.When a customer calls BT to canceltheir line the system gives the agent amuch wider view of the customer, in-cluding their revenue, profitability,recent faults and recent orders. Theycan fully understand the value of thatindividual customer to BT and willmake appropriate offers to encouragethem to remain as customers.

The value of the BI strategy appearsin the figures that the system outputs.It also reduces a lot of wasted internaltime discussing which particular fig-ures are correct! Better informationallows better targeting of customers,which shows up as increased sales. Itcan be very difficult in an organisationthe size of BT to identify the true costof a product offer to customers. In-creased profitability results from a bet-ter understanding of the profitability ofeach product and service.

Mr Griffiths explains that there is nopoint in selling a product unless it is

BT used to have between 350 and 400 business intelligence systems, sourced from almost every softwarevendor. “This is typical of most large organisations,” explains Simon Griffiths, chief designer for business intelligence and data warehousing at BT Group plc.

inherently profitable. There is currentlya great deal of price competitiveness inthe marketplace, especially on broad-band, and it is difficult to know howlow the company can set its priceswhilst still remaining profitable. “BIreally gives us the insight to set ourprice points competitively,” he says.

BT is using BI to improve the cus-tomer experience. Giving the cus-tomer a quality experience with BTinvolves having a single view of BTfrom a customer perspective. Whenthey contact BT, customers are nowable to go to one place and deal withtheir entire relationship with the com-pany.

Historically that has been very dif-ficult because of the different au-tonomous organisations within BTand the regulatory restrictions thatrequire them to deal with each otherat ‘arm’s length.’ Each has differentbills and different web sites. The newinfrastructure now provides a muchmore substantially ‘joined up’ businessintelligence system that allows thecompany to deal much more effec-tively with the customer.

The new BI strategy has had a bigimpact on understanding where thecosts and profits of the business lie.“Information from the system allowsus to prioritise not just our BI spend,but BTs overall corporate strategy,”Mr Griffiths concludes. “It reveals thefactors that drive profit by reducingcosts, increasing revenue and retain-ing customers.”

Is it time for BI basic training?

Yet despite this, as far as we are aware there is noofficial defined body in the UK for Business Intelligence professionals – no overall councilthat regulates the practice of BI in the UK’sworkplaces. As a result there are no accreditedor non-accredited courses available that dealwith BI as a whole. The courses that are availabletend to have a high level of technical contentthat focuses almost entirely on designing andimplementing Business Intelligence solutionsand largely ignore the impact of BI on strategicdecisions and business scenarios.

Fresh from university, or seconded to the job,the only external training received is likely to bein the form of user groups entirely focused onone specific product/solution. Naturally, BI professionals often develop a career long penchant for a particular solution.

Our members tell us this leaves a large gap inthe marketplace for a general training coursethat tackles both the Business and IT side ofBusiness Intelligence – basic training so to speak.This would be a great help to all those who arejust starting to work in the BI field, particularly

in organisations thatare only starting toset up a defined BIfunction. Eventhose who haveworked withinBusiness Intelli-gence and relatedareas for sometime would find aconsolidation oftheir accumulatedknowledge useful.

This is one ofObis Omni’s keyroles, to provideBusiness Intelli-gence profession-

als with the opportunity to pick-up the knowl-edge and skills that others have developed overmany years in a shorter time period via net-working events and seminars. Having a setcourse in place would ensure that everyonewould have at least a certain level of knowledge,as often it varies from one extreme to the otherand would certainly ease the process for a lot ofpeople, especially at the start of their careers inBusiness Intelligence. A lot of the issues we find thatpeople have are on the “business” side, dealing withthe cultural, people and process issue of design-ing, implementing and using a BI system. Yet,there is no course that brings together all ofthese elements into one training package.

In the United States, on the other hand, thereare professional courses that enable people tocomplete a qualification in all aspects of Busi-ness Intelligence, from management and leader-ship through to the technical aspects of data,reports and analysis.

However, there could be many complexities involved in setting up such a training course inthe UK. It is easy to determine whether someonehas the technical qualifications for a technicalrole, they have an approved qualification andpossibly experience in the area. It is much moredifficult to determine the quality of a qualifica-tion in general Business Intelligence, especiallyone that would deal with conceptual ideasaround BI and people management. And with alack of an overall BI institute who would pro-vide the training course?

A sudden influx of courses providing BI traininghowever could cause a severe headache for employers, which course would employers ac-cept as a recognised qualification? With no offi-cial body in the United Kingdom that governs thepractice of Business Intelligence and any train-ing involved in it, it makes it almost impossibleto govern the quality of such training and verifyits independence from any solution provider – a

criteria that would be vital in a basic BI trainingcourse.

A number of our members are completing degrees at various Universities around the UK inBusiness Intelligence and related fields. Manyof them agree that they would choose to com-plete a degree over any short training course atthe present time simply because a degree is morerecognised and the level and length of learningis more intense.

Coupled with this is the vast number of defi-nitions for Business Intelligence, it is hard to cre-ate a course that would be relevant for allbusinesses. In order to set up a general BI course,one that enables graduates to apply for BI jobs inany arena, there needs to be a definition ofBusiness Intelligence that relates to all areas ofits practice – a nigh on impossible job at thepresent time.

So what to do? It seems a bit of a catch 22 situation. There is certainly a need for a level oftraining for BI professionals in both the busi-ness/people management and the technical aspects of Business Intelligence, our experiencewith our members shows us this. But the actualpractice of putting one in place seems extremelydifficult. In order to start providing an overallqualification in BI in the UK there needs to be acertain consolidation of the industry in order to define what a basic training course needs to contain.

It is certainly a much larger and more complexissue than this article can deal with, but it is foodfor thought. This industry is growing at a rapidrate, largely driven by the needs of the end user,and it is likely that if any accredited course is tobe set up, the need and impetus will come fromthe end user. Until that happens Obis Omni willcontinue to provide those involved in BI the opportunity to learn from their peers and inter-act with industry experts in a neutral and cost-free environment. After that, who knows…maybe I’ll have to start looking for another job.

The role and function of Business Intelligence (BI) and Corporate Performance Management (CPM) within success-ful organisations is continuing to grow across all industry sec-tors – having gone from being a nice-to-have function to anessential part of business strategic decisions.BY: CHRIS MONK

Obis Omni is an independent Business and IT Community designed to dramat-ically fast track our members Business and IT projects.

We cover a broad spectrum of topics including Business Intelligence, CPM,Data Quality, SOA, Customer Insight - to name a few.

Obis Omni delivers thought-provoking, transparent information through awide variety of white papers, case studies, publications, news articles, online fo-

rums, events and suppliers. Membership and the majority of services offered byObis Omni are completely FREE.

By becoming a member you will have access to networking opportunitieswith other Business and IT Professional across ALL industry sectors - join us andbuild long-term relationships with peers, experts and suppliers.

Visit www.obisomni.com to sign up for membership

O F F I C E S : L O N D O N & C H E S T E R | www.visualmetrics.co.uk | 0870 7606467

2Exceptional AnswersCritical Questions

visualmetrics | The leading Business Intelligence Solutions Provider

visualmetrics is a leading BI solutions provider who sees its value not only in assisting its customers to exploit data as information on which to base decisions, but as intelligence to promote insight into businesses performance.

• Delivering best of breed Pre-built Analytical Applications

• DRIVE - Project Methodology - minimising risk and maximising ROI

• visualintegrator - The data integrator for Analytical Applications

Packages or toolkits,speed or flexibility?The BI industry has always accommodated two software approaches,packaged applications and toolkits. A package is a pre-built applicationthat can be configured to the organisation’s needs, but only as far asthe vendor allows. Early examples were Commander and Pilot.

answer unique questions that are very hard totransfer from another company or industry.”

A hybrid approach is a toolkit with pre-builtcomponents, such as SAP Integrated Planning,Oracle Business Intelligence, Enterprise Edition orPeopleSoft Enterprise Performance Management.“They provide building blocks that give you arehead start to build your own application,” saysMr Smith. “They are a compromise over a pack-age, but they give you a starting point. They arescalable and allow you to manipulate and changethem in order to meet your requirements and theyare easy to use for a business user, rather thanneeding an all-out technical person.”

Gerry Brown, senior analyst at Bloor Re-search explains that whereas shrink-wrappedapplications are 80% packaged and 20% con-

figurable, these solutions are 60% pre-built andfully customisable.

Mr Gallagher describes this approach as ‘pre-built analytic environments. A package cannotbe extended very much, whereas a frameworkis very open and extensible. “The vendor doesa lot of the hard work in pre-building a lot ofthe ‘plumbing’ and creates a good metadataenvironment,” he says, “so that you can easilypopulate it with data. The great thing is thatbecause they are open you can bolt more ap-plications onto them.”

Speed to market is always important andbest practice is emerging that can be pre-built.However, organisations must always ensurethat they are able to use their information as asource of competitive advantage.

Toolkits, like Essbase and TM1,allowed usersto build whatever application they wanted,but they had to start from scratch. They wereparticularly popular for reporting, planning,budgeting and forecasting, where there waslittle in the way of industry best practice andmost companies had their own unique ap-proaches.

When Arbor Software purchased HyperionSolutions, it seemed that the toolkit approachof Arbor’s Essbase had lost out to Hyperion’spre-packaged approach. Apparently, the mar-ket found the flexibility of toolkits too timeconsuming and preferred the vendor to dothe work in advance, so they could imple-ment it quickly.

“The trouble is that packages can be over-sold, because they look lovely when peopledemonstrate them,” warns Elaine Fletcher,practice manager for business intelligence atLogicaCMG. “The problem then becomes try-ing to bend the application to the company’sbusiness, particularly trying to get data into it.”

Organisations that use a toolkit approach formajor projects frequently find valuable usesfor them on a range of smaller projects. Theycan quickly create a simple application tosolve a business need that is specific to the or-ganisation and for which no package wouldever be suitable.

Ms Fletcher points out that because manyorganisations have been through so manymergers and acquisitions, there are lot of idio-syncrasies within the business. Trying to ac-commodate those within the pre-builtapplication can be quite difficult. GraemeSmith, solution architect at Morse, points outthat companies that grow by acquisition al-ways need a scalable solution that they canchange quickly as the business changes.

“There will always be an element of be-spoke,” says Stephen Gallagher, European headof business intelligence Accenture. “When youget to the final business requirements, the busi-ness question being asked is always slightlydifferent. Lots of Essbase and TM1 solutions

AN INDEPENDENT SUPPLEMENT FROM MEDIAPLANET ABOUT BUSINESS INTELLIGENCE, DISTRIBUTED IN THE TIMES 11

Software Strategies BUSINESS INTELLIGENCE

AN INDEPENDENT SUPPLEMENT FROM MEDIAPLANET ABOUT BUSINESS INTELLIGENCE, DISTRIBUTED IN THE TIMES12

Case StudyBUSINESS INTELLIGENCE

Worldwide reporting challengesWhat if you could gain immediate access to your operational data?What if your business users could securely and easily generate theirown reports?What if you were able to have the insight to make data-driven decisions?

“Noetix was a lifesaver. The productwas dependable, consistent, easy-to-use and saved us millions by enablingus to see the workings of our business.”

Sandra Smith, Head of IS Toshiba Information Systems (UK) Ltd.

Discover why more than 1,300 customers worldwide rely on Noetixeveryday to leverage their currentbusiness intelligence tools and deliverfaster, easier access to operational datafor Oracle’s Enterprise Applications.

Project from BIwith ARCFinally, a business intelligence solutionthat lives up to it’s purpose – inti-mately sensing the dynamics of abusiness. Much more than a toolkit,ARC is designed to make business intelligence work for retailers. ARChas changed the way retail managersin Game Stores, Haggen Supermarketsand Deutsche Woolworth use businessintelligence. No more are they helddown with getting the right DW architecture, developing the right analytics, or delivering reports on time.

ARC is built all the way for retail. Itbrings together a retail data model, arobust data warehouse, hundreds ofmeasures & KPIs, and a highly intu-itive front end in one pre-built solu-tion. By enabling retail analytics in acollaborative environment, ARC insti-tutes an action-oriented BI culture ina retail organisation. Game Stores UKcomments, “In our dynamic business,ARC has helped us keep our finger onthe pulse at all times”.

www.arc-bi.com

It is magic!A data warehouse appliance is giving Virgin Mediaperformance, simplicity and affordability.

Virgin Media is an entertainmentand communications business thatoffers United Kingdom consumerstelevision, telephone, broadbandand mobile telephony. It is made upof businesses formerly run by ntl,Telewest and Virgin Mobile. It hasrevenues of £3.6 billion, 14,000staff and 10 million customers.

Mr Froggatt used to work in thedata warehousing practice of amajor vendor, so he has considerableexperience of tuning relationaldatabases to get the best out of adata warehouse. “You can alwaysdo more tuning and get a littlemore out of it,” he says, “but wewere shackled by inherent complexity. The law of diminishingreturns applies, so you don’t getmuch performance back for the effort you put in.”

He believed that other relationaldatabases would have the sameproblems of scalability, as well asinherent complexity that would inhibit quickly designing andbuilding new solutions. He was impressed by the concept of data-base appliances and their messageof simplicity, although he found ithard to believe their performanceclaims. A data warehouse applianceintegrates database, server andstorage platforms in a single system specifically designed to perform detailed queries and analy-ses on large volumes of data.

The company carried out a fourweek off-site test with four vendors.They were provided with the datastructure of the largest data ware-house, a full day’s call detailrecords (CDRs) and some actualbusiness queries. Each vendor wasevaluated for performance, sim-plicity of solution and cost of own-ership.

The appliances proved that theycould deliver a big step change inperformance over the existing environment. Netezza PerformanceServer demonstrated an averageperformance 252 times the existingenvironment and the others werenot far behind.

The company decided to move toan in-house eight week proof-of-concept project with Netezza. “We

were confident because they haddemonstrated everything we askedand delivered more,” says Mr Froggatt.

The project revalidated the off-siteperformance; extended the scope tocover additional data structures fromanother data warehouse; and testedconcurrency and high volumethroughput. It also proved integrationwith other BI technologies, such asBusiness Objects, Informatica and SAS.

This was actually achieved in onlyfour weeks. At that stage an urgentbusiness need arose, so it was decidedto go straight into production. The sixterabyte data warehouse was migratedin only four weeks.

“It was easy to achieve because alot of groundwork was done and because the appliance delivers on thepromise of simplicity,” says Mr Froggatt.“It deals with the physical layout ofthe database and optimises it. With arelational database it would havetaken at least four weeks just to design the data input and output.With an appliance you merely createyour database structure and load thedata. That is as complex as it gets. Itis magic!”

The users saw an immediate stepchange in performance. Their twicedaily Business Objects queries previ-ously had to be run against speciallybuilt summaries to improve perform-ance. This limited the scope of the

queries and still took 15 minutes to run.The raw power of the appliance

meant that the summaries were nolonger required. This gave users theability to query all the data in thewarehouse and reduced query time toonly three minutes. This is adequate forthem, although the appliance has thefacility to create dynamic summariesthat are updated in real time, whichwould give sub-second query times.

“People find the performance gainshard to believe,” says Mr Froggatt,“but the appliance takes away all theinherent complexities of traditionalrelational database systems. There isno need to worry about physical data-base lay-out, indexing strategies tooptimise queries or clustering data.

Because the users are now able toaccess more detailed data, they areuncovering new data quality issues.The performance means that the datawarehousing team can quickly runhighly detailed queries and reports toinvestigate the problems.

Another benefit from the rawpower of the appliance is the abilityto manipulate large volumes of data.The company need to change somedata structures, which meant manip-ulating 32 billon rows of data. Thistook less than four hours, yet it would-n’t have been possible to attempt it onthe old environment. Virgin Mediacurrently has a ‘spider’s web’ of data

warehouses and data marts that sharecommon integrated data. This hasarisen partly because when the com-pany started with data warehousing,the technology available then didn’tallow a single data warehouse. Havingproven the performance and value ofthe appliance, the company’s objec-tive is to start consolidating and ra-tionalising them. This will result in amuch simpler architecture, with fewerdatabases to manage, fewer points tointerface into and fewer interfaces be-tween databases.

As far as the cost of ownership isconcerned, the project paid for itselfwithin three months of the secondwarehouse application going live. Theappliance saves several databaseadministrators and allows users to runmore detailed analysis over largerdata ranges than before, allowingthem to spot patterns and trends.

“Our existing data warehouse solu-tions were already delivering benefitsand were begging to be scaled up,”says Mr Froggatt, “but we didn’t havethe technology to do it. The performanceboost from moving to the appliancehas given us the ability to increasethose benefits.”

Interestingly, the server in the olddata warehouse environment was in-creased to the maximum available, alengthy project which only doubledthe speed. This compares with the 252times speed improvement in the off-site test, which was conformed in theon-site test.

“The data warehouse appliance isdelivering its promise of performanceand simplicity,” Mr Froggatt con-cludes. “Business users are now freeto run whatever queries they candream up. It is delivering everythingwe wanted it to deliver and more atless than we expected to pay.”

It all started with rumblings of disquiet from the user community. “They wantedgreater performance and more detailed analysis,” says Paul Froggatt, head of infor-mation management at Virgin Media. “We needed to look at alternative technolo-gies to give us a step-change in performance.”

AN INDEPENDENT SUPPLEMENT FROM MEDIAPLANET ABOUT BUSINESS INTELLIGENCE, DISTRIBUTED IN THE TIMES 13

Data Visualisation BUSINESS INTELLIGENCE

One of the world’s largest fast moving con-sumer goods manufacturer decided to re-lookat the investments in its reporting and BI infra-structure to be able to scale up to the demandsof its users.

The company had built over 50 reporting applications all of whichwere built in isolation for departmental or LOB needs mainly focus-ing on providing interactive analysis with a self-use product. Thesesolutions needed to be scaled enterprise wide which was far moredifficult than just being able to use the same product suite on top ofa consolidated data structure.

Issues and differences in SKU definition internally and with mar-ket research agencies, product hierarchy aggregation, evaluation ofsales channel hierarchy performance and incentives, financial dataconsolidation and reporting proved to be major obstacles in movingto a centralized data warehouse. The products, while being adequatefor LOB needs, were not scalable for the enterprise need of beingable to handle diverse data sources with different complexities ofextraction and transformation of data and delivering reports in avariety of ways. Sonata, a global system integrator, was selected toassess their current situation, advise them on tools and technologies,while ensuring the day to day business operations and decision mak-ing abilities were not impacted.

Sonata used its proprietary BI optimization framework to assessthe current infrastructure across a number of attributes (usage, cov-erage, technology adoption etc) and place it amongst one of the 4categories, BASIC, STANDARD, RATIONALIZED and DYNAMIC toindicate their BI maturity. Sonata then presented a roadmap to takethe client to the ‘To Be’ BI application landscape.

Sonata used its SonnetBI framework in building the financial an-alytics modules thus expediting the implementation cycle. The Son-netBI framework contains KPIs and canned reports covering salesand promotion, production, procurement and supply chain and fi-nancials analytics. The framework enables customers to visualize theend application in the technology of their choice thus expediting re-quirements collection and also reducing the risk of the end applica-tion not meeting the requirements. The framework was thencustomized for the company’s requirement by extending the datamodel and writing the ETL connectors to their ERP and Generalledger applications.

Benefits included a reduction in the time taken to consolidate itstrial balances from multiple units across the country from nearly 10days to a day at the end of the month.

Contact details:+44 (0) 208 863 8833

info-uk@sonata-software.comwww.sonata-software.com

SonnetBI is a sound choice for a tuneful ending

A picture paints thousands of numbersGraphical presentation can reveal vital information hidden in the data.

In order to get meaning from the vast quantities of transactional information in databases, users have traditionally aggregated it and presented it in summary form, asgrids full of rows and columns. This leaves them poring over pages of numbers try-ing to extract meaningful insight.

“Accountants are responsible for presenting data, but they are nottrained in it,” says Robert Bittlestone,chairman of Metapraxis, one of thefirst BI vendors. “They often don’t seethe need and many financial directorshave a gift for finding the key numberon pages full of rows and columns. Therest of us digest data from colour andpatterns, so we need to use charts tohighlight exceptions, not present num-bers that reflect ‘business as usual.’”

Some graph types are fairly well established, such as time trends forsales and bar charts for profits. Thesewere incorporated into early BI systems, which still retained rows andcolumns as ‘traffic light’ charts withcolour coded numbers, according towhether they were on or above target(green), below (amber) or seriouslybelow (red).

Highly specialised ‘data visualisa-tion’ software has been used for manyyears in conjunction with data mining.Vendors like Advizor, FYI, QlikTech,Spotfire and Tableau Software are addressing the wider BI market.

“Data visualisation is a fast-growingsegment of the BI market,” saysLouella Fernandes, principal analystat Quocirca. “Vendors are attempting

to make BI more intuitive for abroader audience, away from the traditional user base of expert usersand business analysts. ‘A picturepaints a thousand words,’ or in thiscase ‘many thousands of numbers.’Users can analyse visual patterns inthe data, interpret trends and note exceptions that would not readily beseen in a typical tabular report.”

“If data is not appeal-ing or is difficult to read,that inhibits the use of

the system”David Hobbs-Malyon, SQL Server

product manager at Microsoft, pointsout that some users prefer to view datain one graphical format, whilst othersprefer it in another. This means thatvendors have to provide a number ofdifferent options for presenting data.

“We are being driven by our customers to widen the scope of BI toever greater numbers of users in theorganisation,” he says. “If data is presented to a wider audience in away which is not appealing or is

difficult to read, that inhibits the useof the system. With BI systems, we arealways striving to make the systemeasier for end users to use.”

Gartner predicts that visualisationtechnology will reach the plateau ofproductivity in two to five years, withthe widespread adoption of Flash,Ajax and other Web 2.0 technologies

that enable animated, interactive displays of data. However, the companywarns that internal BI teams lack theknowledge and understanding of howuse the technology.

“One problem seems to be that themore advanced charts are not imme-diately intuitive for the reader,” saysNigel Pendse, lead author of the OLAPReport. “If users need special trainingto understand new chart types, theyare not likely to become mainstream.”

Ovum’s 2006 Business IntelligenceMarket Analysis explains that visual-isation technology exploits the

human capability to interpret shapes,sizes, graphics, shading or colour. Itcan display and deliver data in a moreaesthetically pleasing and easy-to-useway; provide deeper insights into datathat is stored but not effectivelyexploited; and further the use of BIwithin the organisation.

It’s no secret that business intelligence (BI) applicationsand technologies enable organisations to make more in-formed business decisions and provide the strategic ad-vantage they need in today’s competitive marketplace.Consequently, every company must make a determinedeffort to seek out comprehensive resources that helpthem effectively use business intelligence. One trulycomprehensive resource that features vendor-neutral ar-ticles authored by industry experts is the Business In-telligence Network™, a “network” of sites dedicated toserving the BI, data warehousing, performance man-agement and information quality communities with un-paralleled industry coverage and resources.

Founded in 2004 to give readers a one-stop resourcefor all of their BI needs, the Business Intelligence Net-work has sites throughout the world, including the

United States, United Kingdom, Germany, and France. In addition to timely articles, news and research, the

Network offers readers a chance to embrace Web 2.0 tech-nologies and interact with other users. All of the sitesutilise the most advanced technology available for deliv-ery of content over the web, including RSS feeds, expert-hosted blogs, audio-driven interviews and podcasting.Users can read articles, pick up the latest research, checkout daily blogs, subscribe to relevant newsletters anddownload audio podcasts to their MP3 players.

To experience the best in BI, visit www.b-eye-network.co.uk. To see all the sites in the Network, visit www.b-eye-network.com/global.

Business Intelligence Network: Bringing free BI expertise direct to your desktop

The England Rugby Football Union (RFU) has selected OutlookSoft, an SAP company, to man-age the reporting and delivery of operational intelligence from its information systems thatmonitor performance from grass roots to the senior England team. OutlookSoft will automatethe production and delivery of reports from various financial and operational RFU systems,such as ‘RugbyFirst’, which monitors the measurement and performance of players, referees andcoaches, overall participation in the game and the development of the sport across the UK. Out-lookSoft will help Support the RFU’s commitment to improve the quality of information avail-able within the organisation and encourage accountability from management for performanceand results. “We are responsible for measuring the success of Rugby throughout the nation, from

the grass roots to elite rugby,” explains Nick Bunting,head of planning, funding and resources, at theRFU. “It is vital that we provide senior manage-ment with accurate information at a glance. Anefficient summary of the organisation’s per-formance will allow us to identify any areas ofconcern straight away. Automating the reportingprocess will allow us to shift our focus from data col-lection to analysis of the game.”

AN INDEPENDENT SUPPLEMENT FROM MEDIAPLANET ABOUT BUSINESS INTELLIGENCE, DISTRIBUTED IN THE TIMES14

Oracle nowoffers products that started

in 16 different companies. These areAppSource, Alcar, Arbor, Brio, Decisioneering, Hyperion, IMRS, IRISoftware, PeopleSoft, Pillar, Razza,Sapling, Siebel, Sqribe and Upstream.Business Objects has acquired ALG,Cartesis, Crystal Decisions, Firstlogic,Next Action Technology,OLAP@Work and SRC. Microsoft haspurchased Maximal, Panorama,ProClarity and Stratature; SAP has pur-chased Outlooksoft and Pilot Software;and Cognos has acquired Adaytumand Frango.* Industry observers havelong been predicting that the big vendors will dominate, both becauseof their market power and becausecustomers want a complete end-to-end solution. Consolidation allowsoverhead costs to be slashed as backoffice tasks, management, buildingsand other infrastructure is shared.Sales, marketing and developmentcosts used to be shared as well, as thenew products were taken on by thelarger organisation. However, experi-ence has shown that market share is

Goliaths still need the DavidsWhilst the industry consolidates, it is still being driven by innovation.

optimised if thebranding andmarketing are

kept separate forsome time.

Product development and integra-tion strategy is enormously complex.Integration brings together differentproducts with different strategies,whilst product development must address the overlap of similar productsover the long term and involves migrating users towards a single uni-fied product.

“I am not yet convinced that BIsuites are as integrated and work together as seamlessly as the vendorswould like people to believe,” saysAndrew Stevenson, head of centres ofexcellence at Atos Origin. “There isstill a large amount of integrationthat has to take place, so the full benefits are not there yet. A lot of it isstill ‘smoke and mirrors.’”

The recent purchase by Oracle, thebusiness software giant, of Hyperion,one of the leaders in BI, raises thequestion of whether the BI marketneeds to exist as a separate entity? Itcould be argued that the functionalityprovided by BI can be absorbed intothe larger business software suites,called ‘enterprise resource planning.’

This would effectively combine theseparate functions of transaction processing with reporting and analysis.

However, the experts emphasisethat a strong BI industry fosters inno-vation by attracting new companieswith interesting new ideas. “There is awhole new generation of small softwarevendors that are more innovative,nimble and agile,” says Gerry Brown,a senior analyst at Bloor Research.“They adopt new technologies, suchas hosting or ‘software-as-a-service,’that are gaining traction in the marketplace. There is lots of space forthe small players and innovation creates new opportunities in the mar-ket. Also, large vendors are leavinggaps in large accounts that want specific solutions.”

As a company that has built itspresence in the BI market, boththrough internal development and acquisition, Microsoft welcomessmaller companies. “As companieslike Hyperion and Outlooksoft arebeing acquired, other new companiesspring up in emerging new areas,”says David Hobbs-Malyon, the company’s SQL Server product manager, “such as combined BI andbusiness process or real-time businessintelligence.”

Nigel Pendse, lead author of TheOLAP Report, refers to the process ofacquiring small companies as ‘inno-vation substitution.’ “The bigger vendors have lost all ability to innovate,” he says, “so all they do isscour the market for some ‘little guy’looking for an exit route. They are sobureaucratic and tangled up in theirown history that they lose the abilityto keep innovative people, or else innovative projects somehow getstrangled at birth.”

He also blames US accountingrules, which require internal develop-ment costs to be expensed each year,reducing reported profits. In contrast,the cost of the goodwill in acquiringa company with the relevant code canbe treated as a balance sheet asset andwritten off ‘below the line,’ after reported profits.

AttractiveClearly, an integrated fully functionalsoftware suite from a financiallystrong vendor who will develop theproduct, instead of a selection ofproducts from different vendors, is attractive to users. Ian Harrison, man-aging director at Report Source, a per-formance management consultancy,says that vendor consolidation, andthe consequential merging of differentproducts into a single applicationsuite, is great news for customers.

“Traditionally, clients have beenforced to select different products, resulting in different projects and creating competition for user loyaltybetween systems,” he says. “Consoli-dation provides a complete solution,providing a massive economy of scaleand removing many of the roadblocksto adoption.”

Mr Stevenson agrees that consoli-dation can be good for user organisa-

Consolidation has been going on a long time in the BIindustry and the pace shows no signs of slackening. Themain driver has been stronger vendors buying up smallerspecialist companies in order to increase the functional-ity they are able to offer to customers. The two mainareas they are bolstering are performance management

suites and the software tools needed to man-age the underlying data that feeds them.

BUSINESS INTELLIGENCE The Industry

England Rugby Football Union Selects OutlookSoft to Monitor Performance from Grass Roots to International Rugby

tions who want to standardise on asingle toolset to drive down the costper user. However, he warns clientsagainst becoming too dependentupon the vendor’s prescribed datastructures and data models. This couldprevent the organisation using somenew and very smart piece of technol-ogy that could add significant businessbenefit without considerable addi-tional investment.

“I would always argue for keepinga level of ‘vanilla flavour’ in back enddata structures and data models,” hesays. “That allows you to put as many‘flavours’ into the front end as youwish, which is really an argument fornot buying software that has beenconsolidated. Another problem is thatthe larger more prescriptive solutionsare good for analysing your businessin the way the vendor wants you to,with generic ‘pixel perfect’ reporting.However, most users want to do‘train-of-thought’ and ‘what if’ analysis,which they are not good at.”

Mr Pendse argues that smallprivately held vendors are the bestcompanies to deal with. They are thequite happy to stay out of the ‘mael-strom,’ are much better at supportingtheir customers and have more conti-nuity of staff because it may be along time before the founders want tosell out.

“The assumption that a lot of ITpeople have that it is safer to buyfrom a big vendor is largely not true,”he says. “The software industry couldend up like the car companies. Wewould have a collection of large, notvery successful, bureaucratic compa-nies who bring out new models everyyear with small changes, but nothingvery interesting.”

* source www.olapreport.com

AN INDEPENDENT SUPPLEMENT FROM MEDIAPLANET ABOUT BUSINESS INTELLIGENCE, DISTRIBUTED IN THE TIMES 15

Meta data BUSINESS INTELLIGENCE

Exploiting corporate information assetsCommon definitions are the foundation of effective BI

Figures alone have little meaning. The power of BI systems lies in their ability to putthem into a context. This is explored in more details in our articles on the use of datamining and visualisation software to present data on-screen.

The starting point for understanding afigure is its definition. ‘Sales,’ ‘week’or ‘month’ often means differentthings to different departments andtheir systems, so the starting point isto agree a common definition. In different departments in one retailerthe day stated at 9.00am, 6.00pm ormidnight and the week could end onFriday, Saturday or Sunday. The results was seven different definitionsof a ‘week!’

All business terms must be agreedby everybody in the company, alongwith the method for calculating theall-important key performance indi-cators. These are stored in the data

warehouse as metadata, or ‘dataabout data,’ along with details of thesource of the data, when it was lastupdated, how reliable it is, etc. Thismetadata can then be read by BI applications that present the data tousers in understandable businessterms, sometimes called a ‘semanticlayer.

“It actually accounts for only afraction of overall data volumes, andyet it causes a disproportionateamount of problems,” says HelenaSchwenk, an analyst at Ovum. “Thisis because it is typically more dynamic and changeable than othertypes of data, and can therefore

quickly decay, become inaccurate orsuffer from low data quality.”

Andrew Stevenson, head of centresof excellence at Atos Origin, says thatmany people manage their data dictionaries and metadata in spread-sheets, not a proper metadata tool.“People don’t pay enough attention totheir corporate information assets,” hesays. “There are some good tools outthere that provide governanceprocesses for metadata management.”

One problem is that each transactionand BI system has had its own metadatamanagement functions. Industry con-solidation brings different systems together under one owner, but doesn’talways result in unified metadatamanagement.

“There is a trend in the industry towards having a unified platformacross the product line with a singleunderlying repository of metadata,”says Philip Howard, research director atBloor Research. “It is a lot of work, butonce it is done it brings a lot of advan-tages. I encourage users who are look-ing to invest in products and tools toseek those that are metadata-based.”

According to Stephen Gallagher, European head of business intelligenceAccenture, metadata is key to BI. He

sees it as the foundation that providesa unified view of the business, notonly for performance management,but also for facilitating future projects.

“It doesn’t resonate with peoplebecause it is too hard,” he says. “Theyknow they need to do it, but end upusing ‘point’ solutions. High perform-ing companies all have a single viewof the business and a single layer ofmetadata.” The final warning comesfrom Graeme Smith, solution archi-tect at Morse, who says that datawarehouses can mushroom and growwithout a data management strategy.In hisexperience, many companiescan have the same information infour warehouses, defined differentlyand reported by different means. He asserts that only standard definitionsthat everybody has ‘bought into’ and‘signed up’ to will provide a singlenumber that everybody can use.

Metadata is a very worthy subjectthat is much talked about, but notalways given the attention it deserves.If the drive for BI to embrace unstruc-tured data takes off, it will be impor-tant to ensure that effective documentmanagement is integrated with meta-data, as part of an overall businessinformation management strategy.

In essence, the goal with business in-telligence is to extract useful datafrom a number of different sources,such as customer relationship man-agement (CRM) systems, enterpriseresource planning (ERP) systems. Thisdata is stored in data warehouses andmade available to sophisticated ana-lytical tools that allow users to runcomplex queries to uncover businesstrends and patterns.

As companies produce more busi-ness data, these data warehouses havegrown exponentially in volume, lead-ing to similar growth in the storagesystems they reside on. This explosivegrowth, combined with the increas-ingly central role that data ware-houses have in an organisation,means that enterprise-class storage isjust as important a requirement forbusiness intelligence as it has becomefor other applications. If your storageinfrastructure is not optimised to sup-

port the business intelligence systemor the applications it interfaces with,then there is little chance of success.

According to Steve Murphy, UKmanaging director for Hitachi DataSystems, the two main storage chal-lenges that organisations adoptingbusiness intelligence systems face areperformance and scalability. “Busi-ness intelligence tools impose highlycomplex, resource-intensive demandson storage systems. The database willoften have to process a vast numberof data transactions just to respond toone query. At the same time it has todeal with 50 concurrent requests ofthe same nature. This has a significantimpact on speed and performance.”

Scalability is also a key concern asbusiness intelligence systems grow.For example, the CEO of a drinksmanufacturer asks for reports aboutsales over the last year and an ana-lytics tool is put in place. The benefits

from using this new tool are clear somore areas of the business beginusing it. It’s very likely that the sys-tem will grow and the repositorydatabase will expand to the point thatit becomes unmanageable.

While some organisations wouldthink that the solution to performanceand scalability problems would be toadd more storage capacity, this is justan expensive, short-term approachthat can in fact make things muchmore inefficient. Instead, organisa-tions must ensure that their storageinfrastructure has the availability,scalability and manageability neces-sary to support a robust business in-telligence platform, while ensuringthat costs do not spiral out of control.

The first thing that organisations canlook at is adopting storage manage-ment tools that allow them to improvethe manageability and responsivenessof their storage systems. Storage man-agement software gives organisationssignificant control over areas such aspolicy management, storage allocation,configuration and provisioning. Thisprovides significant benefits in storageperformance and service delivery.

Tiered storage is another approachthat organisations should consider.The premise with tiered storage is thatdata becomes less valuable as it ages

and thus should be stored on less ex-pensive storage medium. The cate-gories used for assigning data todifferent tiers may be based on age,levels of protection needed, perform-ance requirements, frequency of useand other considerations.

“The value of adopting tiered stor-age to support business intelligencesystems is that it can dramatically re-duce the cost of storing business data,while ensuring that the right infor-mation is accessible to the business asand when it is needed,” said Murphy.

Because less important data can bemigrated down the storage tiers, thesystem is far more flexible and scala-ble. Furthermore, the top tier of stor-

age is reserved for the most importantand frequently used data, whichmeans that the majority of businessintelligence requests can be processedfar more quickly.

Ultimately, business intelligencesystems are just as reliant on avail-ability, scalability and manageabilityas the databases and storage systemsthat they rely upon. By designing arobust, scalable storage system that isresponsive to the requirements of theapplications it supports and by un-derstanding how the data will be usedby the business organisations can en-sure they have a far more cost effec-tive, reliable and dynamic businessintelligence system.

Storage critical to business intelligenceBusiness intelligence systems have become a criticaltool for organisations looking to improve business operations by getting a better understanding of theiroperations. But what many organisations neglect tothink about when they begin considering a businessintelligence deployment is whether they have the rightstorage infrastructure in place.

Only a few years ago, climate change figured very low in the public consciousness. Politicians andcampaign groups may have discussed the implications of global warming to raise awareness but fewin the business world or wider society considered it an urgent threat.

That situation has changed, with recent images of melting ice caps and shrinking lakes showing thevery real impact our activities have on the environment. In the wake of the recent UN report onglobal warming, more and more people are worried about the future of the planet and climatechange has become a significant concern for governments, industries and communities worldwide.

The business world also faces its own challenges in the climate change debate. The most immedi-ate problem is rising energy costs and the potential threat of energy surcharges. At the same time,the introduction of new regulations and growing public concern means that organisations also facethe risk of prosecution and irreparable damage to their reputation if they do not meet environmentalstandards.

Data centres and the environmentWhile everyone needs to do their part to help reduce carbon emissions and curb excessive energyuse, business and industry are big contributors to the problem. Our ever-growing reliance on technology means that an organisation’s data centre, which often uses huge amounts of energy, canbe one of the main culprits.

Why are data centres to blame? There are three big offenders:Servers: The UK government estimates that servers are responsible for four per cent of the UK’stotal commercial electricity consumptionStorage: IDC estimates that storage accounts for 37 percent of overall data centre power consump-tion. The problem is that storage is often highly inefficient, with the average storage system usu-ally achieving around 45 per cent utilisation, instead of the more acceptable 85 per cent that someolder mainframes achievedCooling: Data centres contain highly sophisticated equipment that must be maintained at precise temperatures, usually requiring energy-hungry cooling systems. In fact, Gartner estimates that forevery watt of machine power used an additional 1 – 1.5 watts is required for cooling

Compounding this is the fact that virtually all organisations continue to generate and handle ever-increasing amounts of business-critical data. The more data that is produced, the more that has tobe stored, adding an even greater power burden. Not only does extreme energy consumption in thedata centre damage the environment, it also compromises enterprises’ bottom lines. The Carbon Trustestimates that wasted energy cost UK businesses £570 million during the summer of 2006 alone.

All of this is translating into pressure on the business community to address rampant power consumption and the impact that IT operations have on the environment. So what do organisationsneed to know and what should they be doing?

Understanding the issueThe critical element for CIOs trying to get to grips with this issue is to understand the true carbonfootprint of their department. Electricity spending is all too often hidden in the operations depart-ment but this makes it very difficult to benchmark power usage. Where CIOs have traditionally fo-

cused on scalability and performance, they now need to take into account space and energy re-quirements that are top of the facility manager’s agenda. CIOs must extend their cost models andbaseline existing power usage so that they can develop a power-efficient storage infrastructure thatis also cost effective.

IT managers now grasp the importance of energy efficiency. Improving the performance of serverprocessors has been a popular tactic but this only solves part of the problem. Storage plays just asbig a role in excess power consumption and as such must play a role in the solution. IT managersmust look at how they can consolidate their storage infrastructure, reducing the amount of data thatis actually being stored and power being used by taking advantage of technologies such as virtualisation, de-duplication and thin provisioning.Gaining a true understanding of the environmental impact of IT also means researching suppliers.It’s not just about how much power, space and cooling hardware requires once installed but alsohow environmentally friendly the production process has been. What materials does the manufac-turer use? Do they have control of the entire supply chain? How far have they shipped the prod-uct? What are the disposal and recycling plans?

Taking positive actionBy now it is becoming increasingly apparent that organisations must begin building and using moreenvironmentally friendly data centres. So what practical steps can they take to address this challenge?

Maximise the utilisation of existing storage resources. The amount of energy consumed by storagedepends on the number of drives used, not the volume of data held. So as capacity density increasesstorage systems become more energy efficient and enable an organisation to reduce its power andcooling requirementsLook for storage solutions that have lower power consumption and heat output, that can be configured to further reduce cooling requirements and also meet environmental directives regard-ing construction and recyclingReduce the amount of data throughput. High-performance storage is not only expensive but alsoconsumes a significant amount of energy. By using tools such as archiving, data de-duplicationand compression organisations can reduce the amount of data that actually has to be processed andstoredLook at technologies such as virtualisation and thin provisioning to consolidate the number of systems in use. In addition to power, space and cooling improvements, these technologies also alloworganisations to re-configure their data centre to take advantage of more balanced cooling configurations

How can Hitachi Data Systems help?Hitachi Data Systems is strongly committed to providing eco-friendly solutions and is helping organisations move towards green storage through its intelligent virtual controllers and softwareand services that extend material power, cooling and space advantages to customers.

How do we do this?Hitachi has far more control over environmentals than competitors by virtue of its unique abilityto separate the controller from the back-end disks, which means it can manage thousands of terabytes behind a single engineHitachi’s thin provisioning software provides several significant benefits by reducing administra-tion costs and improving utilisation rates which result in lower ‘carbon footprints’, improved ap-plication availability and power and money savingsHitachi Consulting has launched a Carbon Management Service that provides customers with fullbenchmarking, analysis, business case development and project management to achieve more eco-friendly data centre operations

Not only does Hitachi Data Systems help its customers implement greener technology, it was oneof the pioneers in incorporating environmentally friendly processes into its entire product life cycle.It voluntarily set an early date to be compliant with the relevant government environmental direc-tives and was one of the early manufacturers to prevent rather than dispose of hazardous wastes.

Hitachi Data Systems is a partner of choice for organisations looking to implement more environ-mentally friendly data centres, reduce their power, cooling and facilities costs and comply with en-vironmental directives. Hitachi achieves this by pursuing a three-fold strategy of designing,manufacturing and supporting environmentally friendly storage infrastructures right throughouttheir entire life cycle.

Five steps to reducing your storage carbon footprint� Reduce data centre hotspots. Storage and server racks should be configured with hot and coldrows. Use virtualisation capabilities to reconfigure hardware without disruption to take advantageof more balanced cooling configurations.� Implement virtualisation strategies. Reposition hot spots without disruption. Utilise a single controller and interface and increase utilisation of existing resources� Consolidate storage. Disparate systems are inefficient and consume unnecessary resources. Carefully evaluate future storage needs and consolidate, lower the number of devices you use andconsider centralised network-attached storage systems to further reduce costs of power, coolingand data centre space.� Deploy services to design the right infrastructure. Consider professional services to help with designing storage architectures, deploying virtualisation and dispose of end-of-life equipment.� Investigate the vendor’s environmental track record. Dig deep and investigate whether environ-mental issues and sustainability are really part of the DNA of a storage vendor. How much controldo they have over the total supply chain? Do they have a defined roadmap to minimise their environmental impact?

How green is your storage?

The growing data burden and inefficient storage management and systems are having a clear environ-mental impact. Steve Murphy, UK Managing Director,Hitachi Data Systems, examines how businesses canreduce their carbon emissions and power consumptionby optimising their data centre and improving procurementpractices.

AN INDEPENDENT SUPPLEMENT FROM MEDIAPLANET ABOUT BUSINESS INTELLIGENCE, DISTRIBUTED IN THE TIMES 17

BUSINESS INTELLIGENCEInformation Management

In many organisations, managing riskhas taken precedence over value creation as the primary objective ofinformation management practices.This often means that far more resources are dedicated to reducingexposure than to creating new solu-tions and processes that confer competitive advantage. Collaborationoffers a way to regain control and putvalue creation back on an equal foot-ing. Organisations must do both tosurvive.

A recent survey (www.snia-dmf.org/100year) conducted by theSNIA (Storage Networking IndustryAssociation) Data ManagementForum (DMF) found that:• Less then 50% of business peoplebelieve they get value from theirlong-term information archive;• Over 80% of business and IT professionals think the cost of long-term archiving is too high;

Collaborate to compete and complyKay Mushens, Storage Networking Industry Association Europe Marketing ChairOrganisations have come to realise that information is both an asset and a potentialliability. The need to remain competitive, coupled with the ever stronger demands ofregulatory compliance, legal discovery, and security risks have created a critical needfor organisational transformation and collaboration among records and informationmanagers, the IT department, security professionals and business users of informa-tion. Failure to respond can expose the company to lost productivity, lost business opportunities, and loss of reputation.

• Business management does notvalue archive work.

This is not surprising. Business usersare not usually concerned with theircompanies’ archiving infrastructure;they just need information to be read-ily available when it’s needed. How-ever, improvements in digital archive technologies, and development of anew interface that allows applicationsto search, access, and migrate fix-content data across heterogeneousstorage environments is creating a renewed interest in archive, because ofthe increased value such capabilitiesplace on data as information.

It is not easy to identify who is responsible for bridging the gap between raw data and valuable infor-mation. Who should determine whenbusiness records should be purged?what should be preserved? and forhow long? The SNIA survey found in-consistent answers to these questions.

The good news is that 14% of survey respondents said that businessrecords information managers (RIM),legal, security and IT collaborate to establish information retention policies.The bad news is that only 14% of busi-nesses are collaborating to establishthe value of their information, thestarting point of many storage andarchiving strategies. Another 14%said they leave the decision to IT. Thisis where the issue stems from: 95% ormore of IT administrators are notaware of the value of the informationthey are managing, which meansthose organisations potentially havecompletely unsuitable data storage andinformation management policies.

Successful information-based man-agement practices require a new typeof collaboration, as information ownersand administrators work together tounderstand and classify the businessvalue and requirements for informa-tion. It is crucial that the strengths ofall participants, including businessstakeholders, IT, RIM, legal and secu-rity are understood and brought tobear in establishing best practices forthe organisation. These range from es-tablishing guidelines for how the mem-bers of the collaborative process work together, to agreeing on uniformmethodologies. The following list is astarting point on which each organi-sation can expand.

• Corporate information policy formulation – Information policiescannot be established in a vacuum orby a single group within the enter-

prise; it is a collaborative endeavourinvolving all stakeholders. Complianceand governance initiatives are excel-lent and requisite opportunities forcollaborative partnerships to develop,educate, and better meet policy requirements.

• The tools to implement and measureconformance to information policies– The collaborative team drives opera-tional requirements, forms the policiesthat serve as governances over theserequirements, and translates policiesinto processes with supporting man-agement tools. All parties have a resulting stake in putting those toolsinto daily practice and monitoring theresults.

• Identification of all information andrecords repositories– Organisationsare struggling to manage the prolif-eration of information across the en-tire organisation and need acontribution from all stakeholders toidentify all information and recordsrepositories, their contents, and theirvalue and relationship to businessprocesses.

• Creation and capture of business-related metadata – The capture ofbusiness-related metadata is a criticalpart of the information lifecycleprocess, with all parties invested inthe who, what, when, why, and howof the metadata capture process.

• Shared responsibility with regard torisk – It is critical that all stakehold-ers acknowledge and understand therisks attached to records. Recordsoriginating from both within and outside the enterprise, and formal andinformal records of all electronic andphysical media types, must be assessed for their value as records aswell their risk potential. Specific,measurable actions should be out-lined when records are stored, retained, moved, or destroyed.

In summary, to achieve informationmanagement compliance within yourorganisation then collaborationbetween the data centre, lines of busi-ness and key enterprise stakeholderssuch as legal, security and recordsand information managers is critical.

One implementation method usedby many companies is ILM (Informa-tion Lifecycle Management), as acomplete information-based manage-ment practice.

The Storage Networking IndustryAssociation is currently developingtwo important standards in this area:the Storage Management Initiative-Specification (SMI-S) for ILM services,and the eXtensible Access Method(XAM). SMI-S for ILM services,essential for automation of ILM-basedpractices, will allow heterogeneousservices to be instrumented anddriven by central ILM managementtools and ILM-enabled applications.The first such specifications willappear in SMI-S 1.2.0, with moremanagement interfaces to follow insubsequent releases. XAM providesapplications with a standard interfaceto storage and with the capability towrite metadata relevant to ILM practices.

Although both standards are still indevelopment, there are tools in themarketplace today that can be used toimplement ILM-based practices now.For example, building and automat-ing ILM-based practices around ver-tical applications such as e-mail ordatabase archiving, and implement-ing and automating tiering, protec-tion, compliance, and archivingsolutions using ILM managementtools or utilities integrated into virtu-alisation platforms.

To learn more about how SNIA isworking to define open technologystandards and best practices in in-formation and storage managementfor the Information-Centric Enter-

prise, visit the SNIA Data Manage-ment Forum at www.snia-dmf.org

ver the last twenty years decision support systems, executive information systems and BIsystems have successively been increasinglysuccessful in delivering keen and precise insight

about what is happening in the organisation. The nextchallenge for organisations is to harness BI to improveperformance by helping to make decisions on changes indirection, changing initiatives and developing effectiveaction plans that will enable them to respond to varia-tions in the business environment. This can only beachieved by stepping back from the functional viewpointthat looks at the organisation departmental by department.It requires taking a systematic view of the entire management process across the enterprise.

The next generation of BI and enterprise performancemanagement (EPM) systems have learned from the success of enterprise resource planning (ERP) systems.ERP suites revolutionised transaction processing by taking a systematic approach to managing businessprocesses right across the organisation, all the way fromtaking an order to banking the money.

EPM takes a similar systematic view of the processes of managing the business and usesit to drive the organisation to higher levels of performance. This is only possible becausethe underlying ERP systems can now provide a rich set of clean and aligned data. It hasalso depended on a great deal of work being done to review the management principlesrequired for EPM and to establish best practice that can be embedded in the softwaresuites.

The success of ERP systems arose from a user debate over the value of an integrated solution from a single vendor across the enterprise, compared with a best-of-breed solution for each department from different vendors. The market gave a clear messagethat it wanted integration. This is because no individual part of a transaction process hasmuch value, compared with the enormous value of having efficient integrated enterprise-wide processes.

This is where the critical differences between ERP and EPM start to appear. In many organisations, individual parts of a management process can be a clear source of competitive advantage, whether better reporting, more detailed analysis, more accurateforecasting or planning processes that are focused on the drivers of profit.

This means that to be successful an EPM system has to be not only fully integrated acrossall processes, but to be best-of-breed in each. Chief financial officers will just not accepta module of an EPM suite that will not give them a source of competitive advantage, nomatter how integrated it is. Each module of an EPM system, whether planning, consoli-dation or analytics, must be able to compete against similar systems from specialist vendors.

People often ask why we need separate systems, when EPM could just be a set of functionality within an ERP suite? They need to be separate because they differ so muchin their basic philosophy. ERP systems are used to run the business, whilst EPM systemsare used to manage it.

Running the back office through an ERP system requires a very structured and controlledenvironment. In order to bring efficiency, it enforces rigid adherence to standardprocesses across the organisation. The IT function builds them and controls when changesare made.

In contrast, the EPM system is under control of the business users, who must have theflexibility to modify it very quickly in response to sudden changes in the business envi-ronment. The IT function installs and runs the software and makes sure that it is correctlyconnected to the data in the ERP systems, but it is the business users who change thefunctionality and the logic.

Even though one system drives the other and they need a degree of separation to allowthem to operate in these different modes, it is still vital that they are fully integrated toeach other, even when they are supplied by different vendors.

In the past, integration was about systems exchanging messages, passing data to eachother and having a similar ‘look-and feel.’ However, users now want to make a stepchange to a much higher level of integration.

So far, computer systems have been ‘application-centric,’ Even though the applicationswere tightly integrated, the user knew which application they were using and had toleave it, start another one and log onto it in order to carry out another part of their task.

Users now want to use software systems that are ‘task-centric.’ They want the system toknow who they are, what activities they have to carry out on a daily basis and what information they need access to. They then want the system to be so integrated that itmakes those activities very quick and easy for them to carry out a task, even guiding themthrough it, invisibly cutting across different applications.

This means that customers need a single integrated technology stack that can manageboth their ERP and EPM systems. Some may prefer it to come from a single vendor andothers will expect the same level of integration between different ERP and EPM vendors.Whatever approach the organisation takes, it will be concerned to avoid parallel worldsof ERP on one side and EPM on the other, with very few links between them. Avoidingthis will allow users to efficiently do their job, seamlessly moving between the worlds ofERP, EPM, enterprise content management (ECM) and e-mail as they complete each task.

We have been talking about combining structured and unstructured data since 1991.Business users spend their day using both, so it has always made sense. However, thiscombination has been hindered by business people being slow to adopt new technology,with some executives even getting their e-mails printed out and hand writing replies!

However, what goes on in the home eventually drives what happens in business. We nowhave a much more computer literate society that is used to using unstructured and structured the world-wide web, so executives now want the same experience at work. Oracle believes that the business world is now ready to move forward with this approach,so EPM systems will finally encompass all business activities.

Embracing ECM allows organisations to end the practise of using e-mail as the de factosystem of record. A robust ECM system of record. A robust ECM system, tightly coupledwith the ERP and EPM systems, can capture the decision processes and information information from the EPM system that led to them. This allows the organisation to re-view and mine the information in order to learn from past decisions, and can assist with regulatory compliance.

This vision of EPM depends upon pervasive use of BI throughout the organisation toalign individual behaviour and actions behind the organisation’s objectives: the neces-sary depth and breadth to comprehensively cover the needs of an entire enterprise, as op-posed to departmental needs; and the ability to operate in a heterogeneous world, so thatthey can leverage their investments in other technologies.

John Kopcke is senior vicepresident for business intelligence andperformance managementat Oracle.

Parallel WorldsThere are strong similarities between enterprise performance management andenterprise resource planning, says John Kopcke, but the differences are crucial.

O

http://www.oracle.com/solutions/business_intelligence/index.html.

AN INDEPENDENT SUPPLEMENT FROM MEDIAPLANET ABOUT BUSINESS INTELLIGENCE, DISTRIBUTED IN THE TIMES 19

News From The Industry

INFORMATION IN MOTION

Register online now for Europe's largest independent data storage, compliance and management conference that takes place in Frankfurt,Germany between 29 - 31 October 2007. Featuring over 50 of the leadingsuppliers in the industry and over 120 educational and thought-provokingsessions, including 30 presentations from a broad spectrum of IT end-usersfrom across Europe and the Middle East. This is backed up by over 30 of SNIA’s own vendor-independent Tutorial sessions and a range of otherindustry ‘hot-topics’ and issues that will be discussed over the three-day event.

Register now for your FREE pass at:

www.snweurope.com

Stephen Gallagher, European head ofbusiness intelligenceat Accenture

Organisations often invest a lot of moneyin data warehousing without gettingmuch meaningful information. That isbecause they didn’t do the preparatorywork on the data, but went straight intothe technical building process.

Creation of metadata is key to datawarehousing, but it must be properly managed. It mustn’t be left to somebody inIT to do, nor must the organisation rely on buying a piece of software to do itfor them.

Organisations should create a data governance management body to administertheir data warehouses. It must have full board sponsorship and be led by somebodywith the authority and influence to make difficult decisions between departments.

Its first task must be to bring together all parties and departments within the or-ganisation to agree definitions of data and key performance indicators. It must ensurethat every piece of source data has an ‘owner’ who is responsible for its managementand accuracy. 100% quality doesn’t exist, so the body must determine what level ofaccuracy is appropriate for each piece of data. Metadata provides a foundation for suc-cessful BI, so it is vital is to invest a lot of time in it at the beginning.

Top Tips

BUSINESS INTELLIGENCE

Visualmetrics, the leading businessintelligence solutions provider, isseeing the success of its own productdevelopment programme through aseries of new business wins. The com-pany, which delivers corporate per-formance management (CPM),reporting and analytical applications,has recently been awarded contractsby a number of blue chip companies,including Balfour Beatty Utilities andUnipart Automotive.

A CPM application has been devel-oped for BBUL that records the finan-cial and operational performance ofcontracts. It allows directors and con-tract managers to see costs by indi-vidual contract, with clear

Businesses and organisations need to harness the powerof the knowledge they have to ensure their commercialfuture, according to information systems providerOITUK. “Knowledge is power, but is nothing withoutcontrol,” says OITUK’s Technical Director, Vijay Magon.

“From a one-person small business to a multinationalcorporation, the same principle applies. Information

needs to be stored in a secure, yet easily accessiblemanner for it to fulfill its potential,” Vijay added.

The OITUK KMS - Knowledge Management System -helps businesses to leverage information held in EDMand EDRM solutions, by extracting content and makingthat content available via the KMS search engine, foruse within businesses.

The growing trend in the BI mar-ket has been consolidation – butwhat about the end user, is thisgood or bad for the customer?

The last 18 months have seen a clear shift inthe BI landscape and the realisation that pureBI players need to offer wider overall solu-tions. Undoubtedly the buzz word has been‘consolidation’- in terms of supplying the fullrange of BI capabilities from Enterprise re-porting to complex interactive planning -bringing BI and BPM into a unified BusinessAnalytics framework.

Unfortunately for the customer, most acqui-sitions have filled gaps in the product lines, cre-ating a patchwork of yet more disparatesolutions, now from one supplier, but with nogreater unification for them to achieve their ob-jectives. Clearly, there is value in working withone supplier, but still “best-of-breed” solutionsare frequently selected.

The hope of industry consolidations is thatcomplete Business analytic products willemerge. These will provide a single data storefor all BI purposes, so that the enterprise re-

porting is driven by the same data as the plan-ning applications. Comments from leading an-alysts would imply that these stores are likely tobe memory-resident in order to provide the per-formance and the versatility – the agility - re-quired by today’s business users.

Applix recognises the real value of an inte-grated solution through bringing together BIand BPM. Applix has led the way in memory-resident modelling, initially applied for finan-cial performance management but more andmore for operational business intelligence. Per-haps the vision of unified Business Analyticscan become reality.

identification of variances. It does thisat a level of detail and timescale thatenables actions to be taken to ensuredelivery against plan.

“We provide our clients with CPMapplications that enhance their Oper-ational efficiency and competitiveedge," commented Chris Coan, man-aging director of visualmetrics. "Weidentify and measure key performanceindicators for enhanced CPM, throughapplications that are aligned to an in-dividual organisation's strategy,processes and objectives. We see ourvalue in not only assisting customersto exploit data on which to make de-cisions, but as intelligence to promoteinsight into business performance."

Visualmetrics seesnew business wins

Let’s get together

Knowledge has no power without control

BI as the catalyst for performance management

Business Intelligence is about making better use of the masses of information that threatens to drown yourorganisation. Customer and product information, financial data, sales reports: how can you ensure your managersare getting the right information to help them make informed decisions? And do they have the right tools to helpthem plan and predict what may happen in the future? Correctly deployed, business intelligence technology cansolve these problems and help you and your managers improve the performance of your business.

Understand how to make your business act intelligently. Attend the Gartner Business Intelligence Summit

and experience the BI technologies and strategies at the cutting edge of best practices.

• 18 Gartner analysts

• More than 80 conference sessions

• Keynotes, panels, interviews, case studies, workshops, Analyst One-on-Ones

• The exhibition with more than 25 leading BI technology and service providers

• More than 900 BI executives from across Europe and all industries

europe.gartner.com/bi

Bringing clarity to your BusinessIntelligence issues

Register before 15 October and receive a 25% discount off the standard delegate price.

To register please emailemea.events@gartner.com and quote BI25