Automating Customer Analytics.
DynaMine® Data Mining Automation powered by KNIME.
2 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Outline
Overview DYMATRIX CONSULTING GROUP
Automating customer analytics
DynaMine & KNIME Framework
Case Studies
3 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Outline
Overview DYMATRIX CONSULTING GROUP
Automating customer analytics
DynaMine & KNIME Framework
Case Studies
4 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
DYMATRIX - The Analytical CRM Company
• Advanced Analytics and Customer Intelligence
Solution Provider located in Stuttgart, Germany
• Since 2000 successfully managed more than 600 projects
• Consulting focus on Analytical CRM, Campaign
Management & Optimization and Data Mining Automation
• Owner-managed and independent
5 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Consulting Expertise
Business Intelligence Advanced Analytics
CRM Consulting Trainings
• Data Warehousing
• Reporting und Dashboards
• Business Performance Management
• Data Quality Management
• Data Cleansing
• Integration of Webtracking Data
• Integration of Social Media Data
• Analytical CRM
• Risk Management
• Web Analytics
• Social Media Analytics
• Text Mining
• Forecasting
• Campaign Optimization
• …
• Feasibility Studies
• Software Evaluation
• Business Process Reengineering
• Project Management
• Customer Experience Management
• Customer Journey
• …
• Data Mining Trainings
• Business Intelligence Trainings
• Partner Trainings (Microsoft, SAS,
KNIME)
6 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Solution Expertise – Customer Insight Suite
7 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Established closed loop solution for successful multi-touchpoint
Campaign Management
• DynaCampaign covers the complete campaign process – from
planning & resource management up to campaign execution.
• Parallell processing of multi-step and event-triggered campaigns
across all touchpoints.
• Open and flexible architecture guarantees rapid solution
deployment.
• Modular design ensures scalability and extensibility.
• Noticeable raise of marketing performance due to higher
conversion rates.
• High degree of automation guarantees cost savings.
• More than 40 Customer installations all over Europe across all
industries.
Multi-Touchpoint Campaign Management
8 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Award-winning Data Mining Automation framework for adaptive
self-learning model training and scoring
• Advanced analytic framework for automated self-learning model
optimization, if changes occour in customer or market behaviour.
• Seamless integration into inbound and outbound campaign
processes.
• Model Manager for realtime Next-Best-Offer/-Activity to deliver
integrated and consistent realtime recommendations across all
customer touchpoints.
• Customer individual best offers and contact dates lead to
verifiable growth in sales.
• Remarkable reduction in support and maintenance of data
mining processes (verifiable up to 200 man-day per year).
Data Mining Automation for adaptive model training & scoring
9 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Innovative solution for intelligent Realtime Decision Management
(Next Best Offer, Credit Limit Optimization etc.)
Benefits:
• Decisions with great accuracy guarantee sales increases.
• DynaCision combines predictive models, business score cards and
business rules via a user-friendly frontend.
• Competitive advantage by combining historical and realtime
behavioural data (e.g. market basket) in realtime.
• Plug&Play-Integration into existing architectures (e.g. Web Shops)
via Web Service – Standard-Interfaces.
Realtime Decision Management
High-quality realtime Next Best Activities based on intelligent and
up-to-date decision models running on online- and offline data.
USP:
10 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Outline
Overview DYMATRIX CONSULTING GROUP
Automating customer analytics
DynaMine & KNIME Framework
Case Studies
11 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Gartner Hype Cycle 2011 – Predictive Analytics on its way to productivity
12 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Customer needs
Possibilities
Decision
Approach
Purchase
Delivery
Usage
Customer Experience* and Customer Intelligence
• Traditional CRM-
approaches focus on
analysis of product
usage.
• Valuable information along
the entire customer
experience chain is not
considered and is not
integrated with others
touchpoints information.
• Intelligent integration of all
customer touchpoints
along the customer
experience chain allows
the identification of
customer needs.
* Source: Customer Experience, Phil Winters, 2011
13 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Challenge: Intelligent Integration of all Customer Touchpoints
Transactional &
Usage Data
Product Information
External Data Sociodemographic Data
Online Tracking Data
Customer
Master Data
Campaign &
Contact Data
Integrated
Analytical
Data Mart
Mobile &
Context Data
Social Media Data
14 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Automation of Customer Analytics to identify customer needs
Customer Behaviour
Predictive Model
Decisions
Learn from the past and detect valuable correlations to predict customer behaviour in the future.
15 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Verhalten
Predictive Model
Aktion
The demand for multiple predictive models will increase continuously to predict up-to-date and precise customer needs in the future.
Verhalten
Predictive Model
Aktion Verhalten
Predictive Model
Aktion Verhalten
Predictive Model
Aktion Verhalten
Predictive Model
Aktion Verhalten
Predictive Model
Aktion Customer Behaviour
Predictive Model
Decisions
Social Media Analytics
Churn
Prevention
Product Affinities
Fraud Detection
Credit Limit
Optimization
Credit
Scoring
Text Mining
Demand
Forecasting
Automation of Customer Analytics to identify customer needs
16 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Automation of Customer Analytics to identify customer needs
Best Product Offer?
Best Contact Time?
Best Touchpoint?
Best Context?
The intelligent interaction of different predictive models serves as decision basis for the next best activity.
17 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Outline
Overview DYMATRIX CONSULTING GROUP
Automating customer analytics
DynaMine & KNIME Framework
Case Studies
18 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
DynaMine® – Framework for data mining automation
DynaMine® is a revolutionary Advanced Analytics Framework for automated, adaptive model training and model management. DynaMine® enables streamlined, cost-effective and fully-integrated data mining processes and supports real-time model training and scoring for any KNIME predictive modeling algorithm.
• DynaMine® automates time- and resource-consuming data mining tasks (data preparation, creation, re-training, assessment and deployment of data mining models)
• Data Mining Models are always up-to-date and are easily administrated and adjusted by the DynaMine® closed-loop control.
• DynaMine® offers best-practice set-ups for data mining processes – gathered, combined and optimized from numerous data mining projects.
• DynaMine® overcomes the limitations imposed by existing modeling approaches - train as many data mining models with as many input variables as you want without any limits.
19 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
DynaMine® Advanced Analytics Loop – powered by KNIME
DynaMine® Data Mining Automation Customer Touchpoints
Models & Scores
Operational & packaged applications
Data Integration
Data Preparation
Customer Behaviour
CRM Social Media Website Other
Parts Date Planning Values Actual Values Values Car Configurator Market Data KPIs
Cruise Control Week 01 2009 600 899 4.567
Cruise Control Week 02 2009 650 988 4.598
Cruise Control Week 03 2009 670 788 4.678
Cruise Control … 680 987 5.012
Navigation System Week 01 2009 1.200 1.098 5.678
Navigation System Week 02 2009 1.300 907 6.012
Navigation System Week 03 2009 1.200 976 5.987
Navigation System … 1.300 876 6.093
Car Audio Week 01 2009 600 898 6.789
Car Audio Week 02 2009 700 954 5.678
Car Audio Week 03 2009 700 879 7.897
Car Audio … 700 901 6.592
Model Deployment
Evaluation
Pre-Processing
Training & Recalibration
Model Assessment
20 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
DynaMine® in a distributed analytical environment: Plug & Score!
Mobile
Applications
ERP SCM CRM Legacy other
Batch / Real-time Business Intelligence Hub
Model Deployment
Evaluation
Pre-Processing
Training & Recalibration
Model Assessment
Model Upload
Cloud based
Model Repository
Scores
& Decisions
Requests
Scores
& Decisions
Requests
Web
Applications
CRM, Sales
& Service
Applications
ERP & SCM
Applications
Social
Media
Plattform independent Scoring “on the cloud“
On-site Scoring via DynaMine®
21 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
DynaMine® – Fields of application in Customer Analytics
Next Best Offer / Next Best Activity
Affinity modeling (Up Selling / Cross Selling)
Churn-/Retention Modeling
Credit Scoring / Credit Limit Allocation
Fraud Detection
Social Media Analytics & Text Mining
…any other application of Predictive Modeling…
22 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Outline
Overview DYMATRIX CONSULTING GROUP
Automating customer analytics
DynaMine & KNIME Framework
Case Studies
23 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Case Study Utilities: Advanced Analytics Enterprise Platform
Business Scenario
• One standard platform for all
customer analytics szenarios
Churn Prevention,
Uplift Modeling
Fraud & Application Scoring,
Forecasting & Simulation…
• E2E-Integration into SAP
environment
Solution
• DynaMine Data Mining Automation powered by KNIME Enterprise Server
• Industrialization of analytical model management (adaptive training &
retraining, evaluation & scoring of models)
• Model- & Score-Integration in Campaign Management and SAP Applications
Constraints
• Budget
• Manpower
• Lack of data mining skills
• Time to market
24 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Case Study Utilities: Advanced Analytics Enterprise Platform
Benefits
• Cost-effective implementation of a next generation advanced analytics
closed loop platform, which ensures a continuous ROI.
• Easy to use DynaMine-Frontend encapsulates data mining complexity.
• KNIME backend offers an unrivaled range of data mining algorithms (KNIME-
native, R, WEKA, user written nodes etc) all being parametrized via
DynaMine Web Services.
• Scalability and easy extensibility for new models and business scenarios.
25 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Case Studies: Social Media Analytics & Customer Touchpoint Integration
Business Scenario
• Automation of intelligent Social
Media Content Parsing &
Sentiment Detection
• Integration with other touchpoint
information
• Enrichment of existing „channel
driven predicitve models“
• Early Warning System
Solution
• DynaMine Data Mining Automation powered by KNIME Text Mining Utility for
Sentiment Detection & Classification
• DynaMine-API for realtime access to Social Media Platforms (Facebook,
Twitter & others)
Constraints
• Social Media Data access &
availability
• Sentiment Parser for German
language
• Methodology
26 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Case Study Car Rental – Tag Cloud as Early Warning System
In June 2011 an online campaign of German car rental company Sixt was
revealed. The ad is both insensitive and provocative as it invites Greeks to
take advantage of the low Sixt rates and spend their “drachmas”.
27 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Case Study Car Rental – Tag Cloud as Early Warning System
Text Mining of Twitter Communications: Tag Cloud as Early Warning System
28 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Case Study eCommerce – Social Media Monitoring
Management
Dashboard
Sentiment
Visualization
Integration of KNIME Text Mining in DYMATRIX SocialMediaAnalytics
29 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Case Study eCommerce – Social Media Monitoring
Details
Sentiment
Visualization
Integration of KNIME Text Mining in DYMATRIX SocialMediaAnalytics
30 © DYMATRIX CONSULTING GROUP UGM 2012 – Automating Customer Analytics
Identification of customer needs along the customer lifecycle
• Classify Messages into need-oriented messages
• Map & identify product/service-oriented needs
Case Study Telecommunication – Identification of customer needs
CONTACT
DYMATRIX CONSULTING GROUP GmbH
Ansprechpartner: Stefan Weingaertner
Zeppelin Carré
Lautenschlagerstrasse 2
D-70173 Stuttgart, Germany
Telefon
Telefax
Web
+49.711.22.007.88 - 0
+49.711.22.007.88 - 88
www.dymatrixconsulting.com
Top Related