2009.10.05 --e marketer---marketing-to-online-video-audience

Executive Summary: A significant barrier to online video advertising’s fullest potential is the audience’s attitudetoward such ads—most people do not trust them. Of 16 advertising tactics and media, online video ads ranked nearthe bottom for trust in an April 2009 Nielsen Online survey, among both the total respondent group and even teens,who are presumably more accustomed to online marketing.

105666

To surmount this hurdle,marketers will need to develop a mix ofad formats that suits the increasingly varied content consumed bythe audience.For example,most video streams watched are 5minutes or less.At the same time,more and more short content isprofessionally created,making it safe for brand marketers.Therefore,a less intrusive mix of preroll and other ad deliveryformats,such as overlays,could help reduce audience resistance.

In addition, content providers are making more full TV showsand movies available online, slowly shifting audience viewingtime away from television to the computer screen.That trendcreates support for more-extensive video ads, such as longerpreroll or midroll placements with creative that rivals TVcommercials for influencing the audience.

Surveys over the past few years consistently show a significantportion of the online audience—around one-quarter—is readyto pay some subscription fee to avoid advertising. But for manypeople this is not an issue of ad-only versus ad-free. Instead,the audience is focused on less-intrusive advertising.

The time is ripe for a video destination that offers a deep catalog ofprofessionally created content supported by a blend of advertisingand subscription fees.The reduced advertising would likely be lessannoying and therefore more acceptable to audience members—and could attract a significant viewer base,giving it substantial scalefor online video ad placements.This mixed business model is,ofcourse,common in cable TV,newspapers and most magazines.

Key Questions■ Why do many people distrust online video advertising, and

how can advertisers overcome that?

■ What ad methods are needed with short video content?

■ Is the online video audience as large as it appears?

Marketing to the OnlineVideo Audience

October 2009

®

David Hallerman,Senior [email protected]

Digital Intelligence Copyright ©2009 eMarketer, Inc. All rights reserved.

Advertising Tactics/Media Trusted* by Internet Usersin North America, by Age, April 2009 (% ofrespondents)

<20 Total

Recommendations from people I know 96% 92%

Consumer opinions posted online 66% 72%

Editorial content such as a newspaper article 81% 70%

Brand Websites 82% 69%

E-mails I signed up for 55% 67%

Ads in newspapers 68% 66%

Brand sponsorships 77% 62%

Ads in magazines 70% 62%

Ads on TV 75% 61%

Ads on radio 73% 61%

Ads before movies 75% 53%

Billboards and outdoor advertising 65% 53%

Ads served on search engine results 46% 37%

Online video ads 35% 33%

Online banner ads 26% 24%

Text ads on a mobile phone 26% 18%

Note: *participants responded that they trusted each tactic "completely"or "somewhat"Source: Nielsen Online, "Nielsen Global Online Consumer Survey," July2009

105666 www.eMarketer.com

The eMarketer View

Many online audience members dislike or distrustonline video advertising. And yet many of the same peoplelike and trust TV commercials.The disparate responses to the twosimilar ad formats are probably due to audience expectations—since much of online video content is ad-free, those clips orprograms with advertising seem off-kilter or uninviting.

That audience perspective will shift as marketers increasinglyimplement two key concepts:

■ Making the length of video ads suitable to the length of content,so that they are not too pushy

■ Devoting resources to develop high-quality video creative that iswell-targeted to the intended online audience

Audiences prefer less-intrusive advertising. Marketerresponse to that perspective will help shape trends in the onlinevideo advertising market. For instance, less may be more.An MTVNetworks/InsightExpress study found their audience liked howshort preroll and overlay ad combinations appeared and thendisappeared, making them lower-impact and less intrusive—butwith higher awareness metrics, too.

Most people would prefer no advertising, in anymedium. For online video advertising, this may mean a shift fromthe implicit agreement between advertisers and audience—theunsaid “we show you ads so you get to watch this video”—to anexplicit one based on greater transparency.That could involvemaking the trade-off between ad-sponsored video (perhaps forvalued content not available on TV) and audience time an obvioustransaction. In this way, ad-supported video could be seen as a formof e-commerce, where people purposely purchase online access tovideo with their attention to the advertising instead of money.

While social media and video are a naturalcombination,marketers may find it difficult to includevideo advertising in social media settings.First off,much ofthe video viewed on social media sites is user-generated,and notattractive for many advertisers.And for those companies that look tojump-start a viral video campaign through social networking,realitysets in when they realize that relatively few social media users forwardcommercial video links.However,since successful viral marketingneeds only a small number of carriers to spread the message, thatmight be less of a problem than it initially appears.Furthermore,manyviral videos can be low-cost,which allows marketers to test differentapproaches without much drag on the bottom line.

Most research indicates that video ads tend to bemore effective at moving the brand needle than,say, static banner ads. And because this is the Internet,some marketers or advertising technology firms expect alladvertising to be interactive, including video.

However, expecting a lot of audience interaction with video ads ismisguided. Research shows fewer clicks with video ads than otheronline formats, which makes sense.When people’s mindset ismainly on watching video content, they are less likely to be drawnaway by some commercial message.

Marketers and media companies both questionwhether the online video audience will acceptsubscription-based video content or look toadvertising alone. As the previous report in this series—“Digital Video Advertising:Where’s the Money?”—concluded,a hybrid of ad support and subscriber fees for video content couldsupport substantial growth for Internet video advertising.

For example, three-quarters of respondents to an IBM survey saidads are OK while the other one-quarter indicated they wouldrather pay a fee to avoid advertising. However, for many peoplethe free/fee question is a gray area, where they will accept someads and reduced fees in exchange for a diversity of content.

This three-quarters/one-quarter split is much like earlier researchon the free/fee question.The fact that a significant minority iswilling to pay for their content hints at two business modelsbesides advertising-only:

■ Fee-only, as with Netflix’s streaming video

■ A blend of fee and advertising, where ads reduce the fees andthe mixed revenues can support a wide range of video content

Video destinations that offer extensive content in exchange forfewer ads and lower fees would likely attract a large audience.That, in turn, would create a robust platform for video ads in anuncluttered environment.

Of the three key ways to evaluate video sites—number of visitors, number of streams and streamsviewed per visitor—the last one indicates the mostabout potential engagement. And when you layer thetime people spend with video upon the type of content (short orlong, professional or user-generated), a far clearer pictureemerges of where to place ad dollars.

Furthermore, the same data indicates that the online videoaudience is becoming more engaged.While the number of visitorshas grown by only single-digit percentages between February andJuly 2009, the average number of videos viewed by each audiencemember has jumped by anywhere from 18% to 50% (dependingon the source) during the same period.

That growth indicates that brand marketers that are not at leasttesting some video advertising at this point will soon be behind thecurve, trailing their competition.

Marketing to the Online Video Audience®

2

http://totalaccess.emarketer.com/Reports/Viewer.aspx?N=0&Nr=P_ID:2000560


3

The eMarketer View

The numbers game falls apart for online video,though, when compared with video viewed ontelevision. When you contrast the total time US consumers spendwatching TV with the total time they spend viewing online video, theWeb delivery channel looks like a mere rounding error at 1% of allvideo time in Q2 2009,according to data from Nielsen Online.

However, the share of total online video time nearly doubledcompared with last year. So if it were to continue to double, thenthe audience would be spending 16% of their total video time infront of computers by 2013. Of course, such clear-cut methods forestimating future trends seldom work out. Further, the breakdownamong the three video screens—adding mobile to the mix—willlikely accelerate over time, with increasing hours of Internet-streamed video viewed on TVs.At that point, the question ofwhether video ads will reach an online audience or a TV audiencewill likely become irrelevant.

Audience Attitudes AboutAdvertising

The dance between advertisers and audiences too

often becomes a push/pull struggle, where

advertisers look to push their messages and

audiences look to pull away.That dynamic is

particularly stark for online video advertising.The

cost of such advertising is high and the opportunity

for branding is strong, so marketers strive to get

the most from it. However, the potential

intrusiveness of video advertising makes many

consumers shy away or consider it unacceptable.

People tend to have less tolerance for online video ads thanseemingly equivalent television commercials.While only 7% ofonline video viewers in an April 2009 survey from Frank N. MagidAssociates and Metacafe said video ads on the Internet weremore acceptable than TV commercials, nearly three times asmany found online less acceptable than TV.

But perhaps the most revealing response: 28% of viewers wereundecided regarding video ads.That group signifies theundeveloped state of online video advertising, with its potential toreach target audiences through brand messages.

105281

The large portion of “not sure” respondents also suggests thatsome people are unclear about what is or is not a video ad. Forexample, some consider a banner that runs next to a video playerto be a video ad, while eMarketer does not.

When video ads appear on network TV Websites, general attitudesimprove.According to a study from Knowledge Networks, 34% ofpeople who stream TV programs online find the ads relevant(which helps make them acceptable), with nearly one-half of thosewho download TV shows perceiving the ads as relevant.

Attitudes of US Online Video Viewers Toward OnlineVideo Ads, April 2009 (% of respondents)

Ads in online videos are more acceptable

7%

Ads in online videos are about the same as ads in TV shows

45%

Ads in online videos are less acceptable

20%

Not sure

28%

Note: n=1,279 ages 12-64Source: Frank N. Magid Associates, "Magid Media Futures 2009:Opportunities in Online Video" sponsored by Metacafe, July 7, 2009



4

Audience Attitudes About Advertising

This result suggests that video ads attached to traditional content,such as TV programs, will evoke more positive audience attitudesthan other types of online video.

102722

A drill-down look at audience attitudes from Nielsen Online showsthat audiences ages 30 and younger are more likely than olderviewers to find online video advertising funny, emotionally touchingand informative—qualities essential for brand marketing.

Younger people tend to be more comfortable than their oldercounterparts with online media, such as video.That comfort canlead to a higher level of engagement, such as the 55% of those 30or younger who find video ads funny. Further, some of video adsare geared more toward a younger demographic, are more edgy incases, and are therefore perceived as funnier.

106683

Attitudinal surveys only go so far, though, because sometimespeople remember incorrectly, wish to please the pollster or look togive the “right” answer. Data-based research might be more ontarget.And time spent might be a more accurate measure ofengagement than ideas such as touching, funny or informative.

In that light, research from Eyeblaster shows that Internet users inthe US and Canada in 2008 spent 260% more “dwell time”engaging with an online video ad (85.16 seconds) than a plainbanner (32.73 seconds).Video ads work better than banneradvertising in that respect.

Note: Dwell time is a newly developed metric that attempts tomeasure time as an engagement factor. Eyeblaster defines it as“the average number of seconds consumers intentionally spendengaging with an online ad.The metric sums the following useractions: amount of time the mouse was over an ad, user-initiatedvideo duration, user-initiated expansion duration, and any otheruser-initiated custom interaction duration.”

106682

For additional information on the above chart, seeEndnote 106682 in the Endnotes section.

Online video ads work in large part because they are, in fact, quitecomparable to TV commercials.

Similar to TV, simply noticing the advertised product/brand wasthe top response among active online video viewers in an August2008 Veoh Networks survey conducted by Forrester Consulting.

Attitudes of US Online Video Viewers TowardNetwork TV Online Video Advertising, 2008 (% ofrespondents)

Ads on favorite TV programs areusually relevant to me

Inclined to buy from companies thatadvertise on my favorite TV programs

Streamers(n=710)

34%

26%

Down-loaders(n=209)

49%

42%

Total(n=1,915)

34%

27%

Note: ages 13-54Source: Knowledge Networks Inc. "How People Use TV's WebConnections," March 18, 2009 as cited by Marketing Charts, March 26,2009


Level of Engagement with Online Video Advertisingby Internet Users in North America, by Age, April 2009(% of respondents)

Funny Emotionally touching Informative

<20 66% 41% 52%

<30 55% 30% 44%

30-34 40% 18% 30%

35+ 42% 26% 31%

Total 47% 27% 34%

Note: respondents who "strongly agree" and "agree" that online video adsengaged them on each of these aspectsSource: Nielsen Online, "Nielsen Global Online Consumer Survey," July2009


Average Time Spent on Online Advertising AmongInternet Users in the US and Canada, 2008

Dwell rate* Average dwell time(seconds)

With video 7.53% 85.16

No video 5.56% 32.73

General 6.54% 57.58

Note: excludes messenger ads; *cumulative instances of where a userinteracted with the ad divided by served impressionsSource: Eyeblaster, "Trends of Time and Attention in Online Advertising,"provided to eMarketer, July 22, 2009



5


Clicking on a video ad, in contrast, was relatively rare, done by 12%or less of each group—even the super-active “connectors.”Whilethat group (which is typically younger than the other two segments)represented just 7% of online viewers, it consumed 20% of all onlinevideo and was responsible for 42% of all online video-sharing.

098734

For additional information on the above chart, seeEndnote 098734 | 098735 in the Endnotes section.

A different survey, conducted in August 2008 by iPerceptions, alsofound that video advertising rarely created interactivity, with only11% of respondents saying they were likely to click.

098895

Similar to the Nielsen Online data, younger video viewers in theiPerceptions survey were more likely to click on video ads thanolder ones.That might point to longer-term trends, where videoads will offer a greater degree of interactivity to an increasinglyavailable audience.

098896

Ad Responses of US Active* Online Video ViewersDuring Their Most Recent Online Video Site Visit, bySegment, August 2008 (% of respondents)

Noticed the product or brand that was advertised during myvideo viewing experience

48%

34%

34%

Clicked on an ad and registered for a sweepstakes or giveaway

12%

3%

4%

Clicked on an ad that was part of or next to a video I waswatching and learned more about the product or service

9%

3%

5%

Clicked on an ad and signed up for more information

6%

3%

2%

Purchased a product

4%

2%

2%

Connectors Controllers Watchers

Note: n=415 (watchers); n=397 (controllers); n=193 (connectors); ages 13+;*those who watch more than 1 hour of online video per weekSource: Veoh Networks, "Watching the Web: How Online Video EngagesAudiences" conducted by Forrester Consulting, October 8, 2008


Likelihood of US Internet Users to Click on OnlineAds, by Format, August 2008 (% of respondents)

Text link 25%

Right banner 20%

Top banner 12%

Video ads 11%

Interactive ads 7%

Interstitial 4%

Source: iPerceptions Inc. survey provided to eMarketer, October 6, 2008


Demographic Profile of US Internet Users Who AreLikely to Click on Online Ads, by Format, August 2008(% of respondents)

Text links

Right banner

Top banner

Video ads

Age

<25 14% 14% 23% 31%

25-34 19% 19% 22% 21%

35-44 19% 22% 22% 14%

45-54 22% 22% 18% 16%

55-64 18% 16% 10% 13%

65+ 7% 6% 4% 5%

Income

<$50,000 36% 33% 39% 49%

$51,000-$75,000 18% 20% 18% 18%

$76,000-$100,000 15% 16% 11% 11%

$101,000-$150,000 16% 15% 17% 10%

$151,000-$250,000 10% 11% 9% 7%

$250,000+ 5% 5% 6% 6%

Frequency*

First visit 17% 12% 14% 19%

Daily 29% 31% 36% 33%

Weekly 33% 34% 31% 31%

Monthly 15% 16% 14% 12%

Sporadic 7% 7% 5% 6%

Note: numbers may not add up to 100% due to rounding; *frequency ofvisiting the site on which the ad appearsSource: iPerceptions Inc. survey provided to eMarketer, October 6, 2008



6


Richly populated video destinations supported by a blend ofadvertising and subscriber fees could draw large audiences and therefore contribute to substantial growth for Internet video advertising. (For more information on video advertisingrevenue models, see eMarketer’s August 2009 report,“Digital Video Advertising: Where’s the Money?”)

When the IBM Institute for Business Value asked US Internet usersif they preferred fee or free for online video content, 74% said adswere OK and 26% would pay to avoid advertising.

This three-to-one split is consistent with prior research on thefee/free question.Those willing to pay subscription fees hint attwo business models besides advertising-only: fee-only, as withNetflix’s streaming video; or a blend of fee and advertising, wherethe ads support lower costs and a broad range of video content.

100325

That blend—common to other media such as cable TV, mostmagazines and newspapers—would require large-scale offeringsof online video content to attract a large audience.

The audience’s acceptance of video ads depends on the nature ofthe content.This is not surprising.

For example,while 86% of the respondents in an April 2009 IpsosMediaCT survey said ad-supported full-length TV shows were very orsomewhat reasonable,50% said the same thing about ad-supportedamateur or homemade video clips.

106714

That relatively high figure for advertising on user-generated videoclips is somewhat surprising, when common wisdom saysvirtually all Internet users think user-generated media should bead-free. Ipsos’ research might encourage some advertisers—thatwere previously skittish about putting their brand messages intouser-generated video clips—to do so.

Video Ad Format PreferencesThe debate over how much Internet users will accept preroll videoads continues. Furthermore, some video advertising advocatesdecry the preroll format because it fails to take advantage of theInternet’s interactivity.

Internet Users in Select Countries Who Are Interestedin Online-Video-Ad-Supported Content vs. Paid OnlineContent, Q3 2008 (% of respondents)

Japan

80% 20%

Australia

75% 25%

US

74% 26%

UK

73% 27%

Germany

69% 31%

India

64% 36%

Would watch advertising before or after quality, free contentWould pay to avoid advertising

Note: n=2,800 ages 13+Source: IBM Institute for Business Value, "Digital Consumer Survey ResultsNovember 2008" provided to eMarketer, November 17, 2008


US Online Video Viewers’ Acceptance of Free,Advertising-Supported Online Video, by Type ofContent, April 2009 (% of respondents)

Very reasonable

Somewhat reasonable

Not very reasonable

Not at all reasonable

Full-lengthTV shows

37% 49% 7% 7%

Full-lengthmovies

35% 43% 12% 10%

Movie/TVtrailers orpreviews

26% 39% 17% 18%

Amateur orhomemadevideo clips

22% 28% 20% 31%

Note: n=939 ages 12+Source: Ipsos MediaCT, "MOTION study," September 9, 2009




7


Nevertheless,audiences find preroll ads familiar, the closest onlineanalogue to TV commercials.While most people prefer no advertising,in any medium,29% of the IBM Institute’s respondents preferredpreroll and postroll ads over any other online video ad format.

100326

When Razorfish contrasted preroll with other types of onlineadvertising,however,more active online consumers preferredbanners than streaming video ads.However,substantially morerespondents preferred preroll to tickers,also called video overlay ads.

099240


The audience preferences in the Veoh/Forrester study pointed to abasic fact of advertising:The less intrusive, the more people accept it.Among those who watched full-length TV programs online, slightlymore preferred static banner ads (43%) than preroll video ads (37%).Far fewer cited a preference for midroll or postroll placements.

098741

In addition, 37% of people who viewed network TV shows online in2008 also watched preroll ads, according to Knowledge Networks.Of course, a higher share of online TV shows and clips containpreroll than do run-of-the-mill online videos.

102725

For more about ad formats, the video “Online VideoPrimer: Video Ad Types” is available on eMarketer’sWebsite. Download video

Online Video Advertising Formats that Internet UsersWorldwide* Prefer to View, Q3 2008 (% ofrespondents)

Video before/after content

29%

Ads in frames surrounding content

12%

Video during content

8%

Sponsored with showcased products

8%

Note: n=2,800 ages 13+; *Australia, Germany, India, Japan, the UK and theUSSource: IBM Institute for Business Value, "Digital Consumer Survey ResultsNovember 2008" provided to eMarketer, November 17, 2008


Online Advertising Formats Preferred by USConnected Consumers, June 2008 (% of respondents)

Banner ads on same page 44.44%

Preroll 33.63%

Interstitials 13.10%

Tickers 8.83%

Source: Razorfish, "Digital Consumer Behavior Study" as cited in "FEED:The Razorfish Consumer Experience Report," October 2008


Ad Types Preferred by US Active* Online VideoViewers Who Watch Full-Length TV Programs Online,August 2008 (% of respondents)

Banner ads that are next to the video Iam watching

Video ads that come on before thevideo I watch

Video ads that show in between videosor segments of videos I watch

Video ads that come on after the video I watch

Ads that come up over the bottom of thevideo I am watching but don't interruptthe video

Ads that come up when I pause the video I am watching

Do notwatch

full- length

39%

36%

11%

18%

21%

11%

Watch full-

length

43%

37%

19%

14%

21%

17%

Watchedfull-length

in mostrecentsession

38%

39%

23%

18%

20%

19%

Note: n=1,013 ages 13+; *those who watch more than 1 hour of onlinevideo per weekSource: Veoh Networks, "Watching the Web: How Online Video EngagesAudiences" conducted by Forrester Consulting, October 8, 2008


US Network TV Video Streamers Who Watch PrerollAds, 2006 & 2008 (% of respondents)

Watch video with preroll ads

30%

37%

Watch video without preroll ad

15%

10%

2006 2008

Note: ages 13-54Source: Knowledge Networks Inc. "How People Use TV's Web Connections"as cited in press release, March 18, 2009


http://totalaccess.emarketer.com/Video.aspx?R=5000504


8


Response to Ad FormatsWhile the preroll is ubiquitous, that may not make it the mosteffective form of online video advertising.

A July 2009 study from MTV Networks and InsightExpressexamined how people react to preroll alone and to preroll blendedwith two other ad formats:

■ What MTV calls a “lower 1/3,” which is a form of overlay

■ The “sideloader,” which is a banner ad that slides out of the right-hand side of the video screen approximately 10 secondsafter video playback begins

For unaided brand awareness, the addition of the lower-1/3overlay to a 5-second preroll, for example, performed betteragainst the control group (who had not seen any of the videoformats) than the 30-second preroll alone.The implication is thatless-intrusive advertising can also be more effective.

Note, however, that all these ads were tested against short-formcontent, where a 30-second preroll might be disproportionate.Most short-form content, even professionally made, tends to have15-second prerolls at most.

106669

For additional information on the above chart, see Endnote106674 | 106676 | 106669 in the Endnotes section.

“30-second prerolls are a nightmare on Webvideo.Nobody wants to see that.”—Dina Kaplan,co-founder and chief operating officer,blip.tv,in a video interview on The Business Insider,August 31,2009

Certainly, more respondents to the MTV/InsightExpress surveyfound 30-second preroll ads more intrusive than the short prerolland overlay combination.

However, longer video ads offer more potential to engage theaudience.That can be seen among the 42.2% who said the 30-secondpreroll was an ad they would stop and look at,7 percentage pointshigher than for the other two types of advertising.

The flip side appears in the 68.4% who agreed that 5-secondpreroll plus lower 1/3 was short and let them get to their videoquickly. Note too that clicking or interacting with any of the threead formats got the lowest positive responses.

106674


When MTV/InsightExpress examined the elements of video ads tosee what people liked best, they found relatively broad support forseveral features. For example—and not surprising—approximately 39% of respondents favored animation or motion inads, regardless of how the ads were delivered.

Brand Metrics for US Online Video AdvertisingCampaigns on MTV Networks, by Type of Ad,January-April 2009 (% of respondents)

Control 5-second preroll

+ lower 1/3

5-second preroll

+ sideloader

30-second preroll

Unaided awareness 12.6% 22.9% 17.6% 19.8%

Aided awareness 72.9% 81.0% 79.4% 81.7%

Online ad awareness 29.2% 56.2% 56.1% 47.5%

Purchase intent 41.4% 44.2% 36.3% 37.9%

Note: n=2,306 (control); n=403 (5-second preroll + lower 1/3); n=427(5-second preroll + sideloader); n=353 (30-second preroll)Source: MTV Networks, "Project Inform" conducted by InsightExpress, July2009; eMarketer calculations, September 2009


US Internet Users’ Attitudes Regarding Online VideoAdvertising Campaigns on MTV Networks, by Type ofAd, January-April 2009 (% of respondents*)

More intrusive than other ad formats

Liked that it didn’t interrupt the video

An ad I would stop and look at

The way the ad was delivered was cool

An ad I would click on or interact with

Length of the ad was appropriate

Ad was short and let me get to myvideo quickly

An ad I’d like to see again

5-secondpreroll

+lower 1/3

24.7%

64.5%

35.0%

44.1%

23.4%

61.0%

68.4%

24.6%

5-secondpreroll

+sideloader

29.2%

57.9%

35.2%

43.2%

25.1%

60.5%

64.2%

26.5%

30-secondpreroll

30.0%

61.8%

42.2%

45.7%

25.1%

63.1%

58.6%

29.8%

Note: n=403 (5-second preroll + lower 1/3); n=427 (5-second preroll +sideloader); n=353 (30-second preroll); *respondents who answered“agree” or “strongly agree” on 5-point scaleSource: MTV Networks, "Project Inform" conducted by InsightExpress, July2009



9


However, remember that most people prefer less-intrusive ads. Soit is also no surprise that the highest response rate was forappearing/disappearing 5-second preroll ads combined with lower1/3 or sideloader, and lower for the longer 30-second preroll.

106676


Dislike for Video AdsThe MTV/InsightExpress focus on drill-down details of what theaudience likes is rarely seen research. But the audience’s generaldislike of advertising, including video, comes through loud and clear.

When Nielsen Online asked Internet users worldwide which of 16advertising tactics or media they trusted somewhat or completely,online video ads ranked 14th, cited by only 37% of respondents. Incontrast, 62% said they trusted TV commercials.

105383

Why the gap between the two similar advertising formats?Expectations are likely a big part of it, since people areaccustomed to TV commercials but expect free online content.

That audience attitude might evolve over time, as marketers learnhow to better blend high-quality video ad creative into professionalvideo content, which is becoming increasingly available.

However, even though only about one-third of Internet userstrusted video ads to some degree, two-thirds (67%) of therespondents agreed or strongly agreed with the statement thatadvertising supports low-cost and free content on the Internet,TVand other media.

US Internet Users Who Liked Elements of OnlineVideo Advertising Campaigns on MTV Networks, byType of Ad, January-April 2009 (% of respondents*)

Animation or motion of the ad

Movement of ad across the screen

How the ad appears and then disappears

Ability to expand the ad

Ability to click a link in the ad

Ability to see a longer video ad with aclick

5-secondpreroll

+lower 1/3

38.8%

32.4%

43.9%

31.7%

36.4%

34.2%

5-secondpreroll

+sideloader

39.8%

37.9%

47.9%

33.9%

38.8%

36.3%

30-secondpreroll

40.4%

30.9%

38.5%

29.4%

30.9%

29.5%



Advertising Tactics/Media Trusted* by Internet UsersWorldwide, April 2009 (% of respondents)

Recommendations from people known 90%

Consumer opinions posted online 70%

Brand Websites 70%

Editorial content (e.g., newspaper article) 69%

Brand sponsorships 64%

TV 62%

Newspapers 61%

Magazines 59%

Billboards/outdoor advertising 55%

Radio 55%

E-mails signed up for 54%

Ads before movies 52%

Search engine results ads 41%

Online video ads 37%

Online banner ads 33%

Text ads on mobile phones 24%

Note: *participants responded that they trusted each tactic "completely"or "somewhat"Source: Nielsen Online, "Nielsen Global Online Consumer Survey" as citedin company blog, July 7, 2009



10


That understanding of the implicit contract between advertisersand audience offers hope to marketers that, with time creatingfamiliarity, people will come to trust online video ads more.Further, that understanding opens the door to more transparency,such as when marketers sponsor video content of substantialvalue. In that scenario, the audience is made aware of both themarketer’s message and the content’s worth.

105659

The Veoh/Forrester study also illustrated the imbalance betweenaudiences finding video ads annoying (which reduces trust) andthe trade-off of ads for free content.Whether the respondentswere categorized as active connectors or more passive watchers,about 20% more called video ads annoying than those whothought of them as fair to help pay for video services.

098735

For additional information on the above chart, seeEndnote 098734 | 098735 in the Endnotes section.

Attitudes of Internet Users Worldwide Toward theValue of Advertising, April 2009 (% of respondents)

Advertising and sponsorshipsare important to fund sportingevents, art exhibitions andcultural events

By helping companies succeed,advertising creates jobs inthose companies and in theadvertising industry

Advertising stimulatescompetition which leads tobetter products and lowerprices

By providing me withinformation, advertising allowsme to make better consumerchoices

Advertising funds low-cost andfree content on the Internet, TV,newspapers and other media

Advertising contributes togrowth of the economy

Advertising often gets myattention and is entertaining

Advertising provides usefulinformation on importantissues in society, such assafety and health

Stronglyagree

22%

15%

16%

14%

14%

13%

13%

13%

Agree

59%

65%

52%

53%

53%

58%

53%

50%

Dis-agree

10%

12%

22%

22%

18%

17%

21%

23%

Stronglydis-

agree

3%

2%

5%

6%

4%

3%

7%

7%

Noopinion

6%

7%

6%

5%

11%

8%

5%

7%

Source: Nielsen Online, "Nielsen Global Online Consumer Survey," July2009


Attitudes Toward In-Video Ads Among US Active*Online Video Viewers, by Segment, August 2008 (% ofrespondents)

In-video ads are annoying

61%

59%

58%

In-video ads are fair as they help pay for online video services

41%

37%

34%

I avoid Websites that use in-video advertising

19%

21%

23%

In-video ads are useful when they feature something I aminterested in

28%

19%

19%

In-video ads can be entertaining

23%

17%

10%





11


Social Media and Video AdvertisingThe audience dislike of online video advertising encourages somemarketers to believe that social media will help them break downbarriers.That perspective is reinforced by responses to a GfK Ropersurvey,where only 29% of respondents cited marketing on socialnetworking sites as annoying,compared with 50% who were annoyedby ads during online or offline video-on-demand programming.

100164

The thinking is that video’s prevalence on social network sites—and the social elements of sharing, forwarding and commentingon video sites—helps put the “media” in social media.And,therefore, video and social media would naturally go together.

It is clear why social media users attract marketers interested inonline video. For one, 53% of social network users watched moreonline video than TV in Q4 2008, according to LiveRail.

101066

However,much of that watching was undoubtedly of user-generatedcontent (which most advertisers shun).That trend is seen in aMarch–April 2009 survey by the Pew Internet & American Life Project,which found that 62% of respondents viewed user-generated videoson a sharing site.

105811

Types of Ads/Promotions that US Internet Users FindAcceptable or Annoying, September-October 2008 (%of respondents)

Free product sample distributed by people orincluded in magazines/newspapers

Newspaper ads

Magazine ads

Radio ads

Ads on TV

Product placement in movies/TV programs/videogames

Unconventional marketing*

Social networking sites or blogs

Paid ads in search engine results

Product placement in virtual worlds

Ads during video-on-demand programming

Web ads that play before landing/obscurehomepage

Pop-up or banner ads on Websites

Ads on mobile phones

Unsolicited e-mail

Quite acceptable

67%

56%

50%

43%

42%

31%

24%

23%

21%

15%

10%

8%

6%

5%

4%

Find annoying

12%

12%

19%

29%

37%

36%

37%

29%

48%

24%

50%

79%

83%

70%

84%

Note: n=2,000 ages 18+; *ads in unconventional places or the use ofindividuals to spread the word about products/servicesSource: GfK Roper, "Media, Advertising & Consumers," provided toeMarketer, November 20, 2008


US Social Network Users Who Watch More OnlineVideo Than TV, Q4 2008 (% of respondents)

Watched more online video than TV

53%

Watched about the same

19%

Watched more TV than online video

28%

Note: n=400+ ages 18-24Source: LiveRail, "State of the Industry: LiveRail’s Q4 2008 review of onlinevideo advertising," January 19, 2009


Online Social Media Activities of US Internet Users,March-April 2009 (% of respondents)

Watch user-generated video on a sharing site 62%

Use social networking sites 46%

Use Twitter/share updates 11%

Note: n=1,687 ages 18+Source: Pew Internet & American Life Project, "The Audience for OnlineVideo-Sharing Sites Shoots Up," June 17, 2009



12


In response to the connection between social media and video,63% of brand marketers and 60% of agency executives in June andJuly 2009 told Equation Research that online videos were part oftheir social media efforts, right up there with Facebook and Twitter.

106250

A lesser,but still significant,percentage of marketers (41%) in aMarch 2009 Michael A.Stelzner survey said they made video—eitheron YouTube or other sites—part of their social media outreach.

Note that a link to the survey was originally posted on Twitter and wasreposted by other marketers on Facebook and blogs.Thatmethodology naturally draws in those already involved in social media.

102683

For some companies, the primary way to blend video advertisingand social media is through viral marketing.

Social Media Used for Marketing According to USBrand Marketers and Ad Agencies, June-July 2009 (%of respondents)

Brand Agency

Facebook 80% 84%

Twitter 69% 76%

Online videos 63% 60%

Blogs 57% 72%

Podcasts 34% 28%

MySpace 34% 25%

Webcasts 33% 31%

User forums 33% 22%

Widgets 30% 23%

Wikis 24% 19%

Other 21% 25%

Note: n=123 brand marketers; n=264 ad agenciesSource: Equation Research, "2009 Marketing Industry Trends Report,"August 18, 2009


Social Media Used by US Marketers, January 2009 (%of respondents)

Twitter

86%

Blogs

79%

LinkedIn

78%

Facebook

77%

YouTube or other video

41%

Social bookmarking sites (i.e., Delicious)

38%

Forums

38%

StumbleUpon

28%

Digg, reddit, Mixx or similar site

26%

FriendFeed

18%

Source: Michael A. Stelzner, "Social Media Marketing Industry Report,"March 17, 2009



13


At this stage of the game, viral video is a low-cost process.According to a MarketingSherpa survey in October and November2008 of marketers who created one or more viral videos during theprior year, the budgets for 62% of respondents ranged from only$101 to $10,000.That low financial impact encourages companiesto try viral video marketing, since the cost of failure is so small.

103091

However, forwarding video links—the main element marketerslook for with viral video—may not be all that common. In theVeoh/Forrester survey, only 21% of active video viewers said theypassed along a video or a link during their last visit to an onlinevideo site. In comparison, 37% performed no type of social activity.

098731

The January 2009 survey from Lightspeed Research found 39% ofrespondents shared a video clip—not just during their most recentvisit, as in the Veoh/Forrester poll, but ever.When you considerthat successful viral marketing needs only a small number ofcarriers to spread the message, that figure gives ammunition tothose looking to combine video with social media.

104104

However, much of that video-sharing has little to do withmarketing, since forwarding links to commercial videos is lesscommon. In 2006, for example, 26% of those people who viewednetwork TV sent a friend a video ad link. But by 2008, that figuredropped to only 17%, according to Knowledge Networks. In thiscase, it seems the more common video ads become online, theless noteworthy as well.

102723

The adage holds true:You cannot make an ad go viral. Socialnetworks can help because of some of the built-in notificationoptions that are in place—if users view an ad or comment on it, anotification goes to their friends on their newsfeed. But if theaudience doesn’t find the ad interesting enough to view in the firstplace, it doesn’t matter.

Amount US Marketers* Spend on Viral Video,October-November 2008 (% of respondents)

$0 11%

$1-$100 15%

$101-$500 16%

$501-$1,000 16%

$1,000-$10,000 30%

$10,001-$100,000 12%

$100,000+ 1%

Note: n=1,083; numbers may not add up to 100% due to rounding; *whocreated one or more viral videos in the past yearSource: MarketingSherpa, "Video Marketing Survey," April 10, 2009


Social Activities of US Active* Online Video ViewersDuring Their Most Recent Online Video Site Visit,August 2008 (% of respondents)

Read comments from others

40%

Looked at the ratings others had given the video

28%

Sent a video or link to a video to someone else

21%

Rated a video I was watching

17%

Posted a comment on a video that I watched

11%

Posted a video or link to a video on my blog or social networkingpage (e.g., Facebook, MySpace)

6%

Bookmarked to a social media or bookmarking site (e.g., Digg,StumbleUpon, Delicious, reddit, etc.)

5%

None of the above

37%



Online Video and Social Networking Activities of USInternet Users, January 2009 (% of respondents)

Watched a video clip 72%

Read a blog 46%

Managed a social network profile 41%

Shared a video clip 39%

Uploaded a video clip 32%

Wrote a blog 18%

Note: ages 16-64Source: "Global Web Index" conducted by Lightspeed Research, May 28,2009


US Network TV Video Streamers Who HaveForwarded a Link to a Commercial or AdvertiserVideo, 2006 & 2008 (% of respondents)

2006 26%

2008 17%

Note: ages 13-54Source: Knowledge Networks Inc. "How People Use TV's Web Connections"as cited in press release, March 18, 2009



14

What People Watch (and Why)

To overcome audience resistance to online video

advertising,marketers must find an appropriate

match between the ad and the associated content.

The plethora of video types that people watch on

the Web complicates matters for advertisers.

The majority of video content people watch online is short video,at least when measured by streams instead of viewing time.Among the active viewers polled by Veoh/Forrester, 51% or morewatched various short-form content (such as user-generated);clips from movies, music or TV shows; and local or national news.

098724

“Our research shows the Internet is afundamentally different experience forconsumers than television. Online, they arewatching shorter clips in general.They’rebrowsing.They’re discovering.They’researching. I think it just confirms somethingthat I think we’ve known for a while but themarket overlooks, which is the newconsumer behavior and new opportunity isin video.” —Erick Hachenburg, CEO, Metacafe,in an interview with eMarketer, July 2009

Similar data from the Magid/Metacafe report also show thatpeople mostly watched short content: user-generated video, newsstories and music videos topped this list.

Only 25% of respondents in this survey regularly watched full-length TV shows,or one-half of the figure in the Veoh/Forrester poll.However,Magid/Metacafe surveyed the broad audience of people who haveever watched video content online,while Veoh/Forrester surveyedonly those who watched more than 1 hour of video per week.

105276

Types of Online Video Content Viewed by US Active*Online Video Viewers, August 2008 (% of respondents)

Video posted by people like me (e.g., on YouTube, Google Video)

71%

Movie clips or trailers

63%

Music clips or trailers

59%

Clips from current TV programs

57%

National/local news

55%

Comedy or other humorous online videos

51%

Full-length current TV episodes

50%

Sports news/highlights/events

39%

Full-length movies

33%

Clips of TV shows that aren't on TV anymore

32%

Full-length TV episodes that aren't on TV anymore

28%

Commercials or promotional video created by companies topromote their brands

25%

Animation, anime and cartoons

25%

Adult entertainment

20%

Note: n=832 ages 13+; viewed in the past month; *those who watch morethan 1 hour of online video per weekSource: Veoh Networks, "Watching the Web: How Online Video EngagesAudiences" conducted by Forrester Consulting, October 8, 2008


Types of Online Video Watched Regularly by USOnline Video Viewers, April 2009 (% of respondents)

Videos shot and uploaded by consumers to sites like YouTube 42%

News stories 32%

Music videos 31%

Movie previews 29%

Comedy/jokes/bloopers 26%

Full-length TV shows 25%

Weather 20%

Clips of TV shows 18%

Sports content 16%

Full-length movies 15%

Adult entertainment* 14%

Video game content 12%

Short films 10%

TV previews 9%

TV commercials 7%

Other 3%

None 6%

Note: n=1,279 ages 12-64 who have ever watched online video content;*asked of respondents ages 18+Source: Frank N. Magid Associates, "Magid Media Futures 2009:Opportunities in Online Video" sponsored by Metacafe, July 7, 2009


http://totalaccess.emarketer.com/Interview.aspx?R=6000253


15


Among US Internet users in general, the UK Office ofCommunications (Ofcom) found that 47% watched short videoclips, but only 23% watched longer video content such as featurefilms or complete TV programs.

100522

Even though short videos limit marketers’ options, professionallymade short content is often welcomed by both advertisers and audiences.

The Magid/Metacafe report focused on short-form professionalcontent that averaged 90 seconds, including music videos, movietrailers,TV clips and sports highlights.Among Internet users whowatched such pieces, 37% said the content was equally or moreentertaining than full-length TV programs shown on a TV set.

That result reinforces the predominance of video snacking on theWeb, where the lean-forward nature of the medium encouragesshort consumption.That also implies brand marketers might dowell to develop more video ads that suit the growing universe ofshort professional video content.

105275

Ad Length versus Content LengthHow long a video ad runs is often directly related to the length ofthe content. Logical. Many marketers understand this, as in a fall2008 survey from MarketingSherpa, where 48% of respondentssaid that for each 60 minutes of online video content, the ideallength of advertising would be 1 to 5 minutes.At the high end, 5minutes of advertising would mean one 30-second streaming ad(preroll or midroll) for every 6 minutes of content.

103551

Audience surveys indicate that even though many enjoy shortprofessional clips, they still express limits on the advertisingattached to the content. Perhaps those limits are not as strict asanecdotal evidence suggests.

A 2007 Dynamic Logic study found that a 10-second-plus preroll adwas all that Internet users were willing to watch before a 5-minuteclip—although ad length tolerance nearly doubled to over 20 secondswhen the study excluded those unwilling to watch any advertising.

Online Video Consumption of Internet Users in SelectCountries, by Type, October 2008 (% of respondents)

Watch or download short

video clips*

Watch or download longer video

content**

Canada (n=1,000) 50% 22%

Germany (n=1,002) 47% 14%

UK (n=1,001) 47% 23%

US (n=1,010) 47% 23%

Italy (n=1,003) 45% 27%

France (n=1,000) 43% 21%

Japan (n=1,003) 39% 18%

Note: ages 18-64; *including those made by other people; **such asfeature films or complete TV programsSource: Office of Communications (Ofcom) - UK, "The InternationalCommunications Market 2008” conducted by Synovate, November 20,2008


US Online Video Viewers* Who Find ShortProfessional Online Clips** Equally or MoreEntertaining than Full-Length TV Shows on a TV Set,April 2009 (scale of 1-5)

1 (strongly agree) 13%

2 24%

3 41%

4 14%

5 (strongly disagree) 8%

Note: n=883 ages 12-64; *who watch short professional online clips;**includes music videos, movie trailers, TV clips, video game trailers andsports highlightsSource: Frank N. Magid Associates, "Magid Media Futures 2009:Opportunities in Online Video" sponsored by Metacafe, July 7, 2009


Ideal Length of Ads per Online Video Content-HourAccording to US Marketers, October-November 2008(% of respondents)

0 minutes/find other methods 10%

1-5 minutes 48%

5-10 minutes 29%

11-15 minutes 11%

16-20 minutes 2%

Note: n=1,083Source: MarketingSherpa, "Video Marketing Survey," November 2008 ascited in "Marketing With Video Report: Online, TV & Mobile," December2008



16


While there have been many changes in the online video universesince this study was done two years ago, the greater abundanceof video today makes many Internet users more accustomed toadvertising, even as it also makes some more resistant to onlinevideo ads.

106681

However, when audiences watched short videos of 5 minutes orless, their ad time tolerance got restricted, along with the ad’splacement.The same Dynamic Logic study found that for midrollads, a 5-second spot was OK for 75% of respondents, but only 11%of respondents or fewer were willing to watch when the ad lengthgrew to 10 seconds or more.

092814

Professional versus User-GeneratedThe US audience’s taste for online video appears balanced,according to Deloitte’s annual “State of the Media Democracy”report, with 70% watching user-generated and 71% watching theprofessional kind.

105617

Not only do brand marketers trust professionally created videosfar more than user-generated, but nearly all longer (10 minutes ormore) online videos are professional content.The universe ofprofessional video content can be categorized into several types:

■ Full TV shows

■ Full movies

■ Clips from TV shows or movies

■ Made-for-the-Web shows (often less than 10 minutes long)

■ News, weather, sports (often from TV networks or local stations)

■ Educational and informational

“Most advertisers want to have a high degreeof comfort around the nature of the contentthat they are sponsoring.And so the sheerprocess of managing the millions of assets outthere—it would really only come down to themost popular.OK,squirrel on a skateboard,that gets 10 million views.Why not? But by thetime it got to 10 million views, it would be toolate.Maybe you’d catch the last millionviews.”—Brian Wieser,SVP global director of forecasting,MAGNA, in an interview with eMarketer, June 2009

The online video audience spends more time watchingprofessional content, too.That extended time gives marketersmore opportunities to place ads in front of them, and moreopportunities to test different ways of delivering video-basedmarketing messages.

Average Length of Ad that US Internet Users AreWilling to Watch Before an Online Video, September2007 (seconds)Video length Length of ad

among all users (seconds)

Length of ad excluding users

unwilling towatch any ad

(seconds)

10-minute clip 14.70 26.90

5-minute clip 10.78 20.36



30-second clip 4.61 8.58

15-second clip 3.48 6.13

Note: n=933 ages 18+Source: Dynamic Logic, "AdReaction 2007: Consumers' Perceptions ofAdvertising," provided to eMarketer, October 2007


Length of Ad* US Internet Users Are Willing to WatchDuring Their Online Video**, September 2007(seconds)

Before thevideo

During thevideo

After thevideo

<5 seconds 45% 75% 47%

10 seconds 23% 11% 19%

15 seconds 16% 5% 14%

30 seconds 12% 6% 14%

45 seconds 2% 1% 2%

1 minute 2% 2% 3%

2+ minutes 0% 1% 1%

Note: n=933 ages 18+; *before, during or after the online video; **fiveminute news clip or videoSource: Dynamic Logic, "AdReaction 2007: Consumers' Perceptions ofAdvertising," provided to eMarketer, October 2007


US Internet Users Who Watch User- and ProfessionallyGenerated Video, September-October 2008 (% ofrespondents)Watch user-generated video 70%

of which: % who regularly watch a complete video 80%

Watch professionally generated video 71%

of which: % who regularly watch a complete video 79%

Note: n=2,056Source: Deloitte, "State of the Media Democracy Third Edition," provided toeMarketer, January 12, 2009




17


According to Deloitte’s survey from September and October 2008,when US Internet users watched user-generated videos, theyaveraged 8 minutes, compared with 12 minutes for professionalvideo content.

100979

The likelihood that people will watch full-length TV programsappears to shift far more based on age and income than bygender. In the Veoh/Forrester study, younger audience memberswith lesser incomes represented the biggest groups consumingTV episodes.

For now, at least, that indicates that online video advertising for ayounger target demographic might be more efficient, as theadvertising messages will reach a larger audience. Further, itpoints to how brand marketers that are making video ads just forthe Web should look to add creative elements that will appeal tothe same demo.

098737

“I’ve seen some really brilliant work on howthey’re making UGC safe.” —Tracey Scheppach,senior vice president and video innovation director,Starcom USA, in an interview with eMarketer, June 2009

In the Magid/Metacafe study, the same under-34 demographicwas most likely to watch short professional clips, too.And in everyage category other than teens, more males than females watchedsuch video content.

105277

Whether professional or user-generated, video content thatcaptures audience attention will better sustain any attachedadvertising.According to a May 2009 Lightspeed Research report,“Online Video: 2009 is Primetime,” three types of professionalvideo that got relatively high attention were news highlights,movie trailers and music videos. Since the latter two are, in fact,forms of advertising all by themselves, it would seem that newsclips might be one of the best types of short-form content in whichto place video ads.

Note, too, that clips passed on by friends got the most attention inthe survey. For professional content, that implies a viral qualitythat lends itself to sharing.While many marketers want to maketheir edgy or outrageous ads go viral, that cannot be controlled—only attempted.

104110

Average Length of User- and Professionally GeneratedVideos Watched by Internet Users in Select Countries,September-October 2008 (minutes)

Brazil Germany Japan UK US

User-generated 7 6 11 7 8

Professionally generated 12 9 13 10 12

Source: Deloitte, "State of the Media Democracy Third Edition," provided toeMarketer, January 12, 2009


Demographic Profile of US Active* Online VideoViewers Who Watch Full-Length TV Programs Online,August 2008 (% of respondents)

Gender

Female

Male

Age

13-34

35-54

55+

Income

<$75,000

$75,000+

Do not watchfull-length

50%

50%

44%

34%

22%

69%

31%

Watch full-length

52%

48%

64%

27%

9%

66%

34%

Watched full-lengthin most recent session

57%

43%

66%

27%

7%

71%

29%



US Online Video Viewers Who Watch ShortProfessional Online Clips,* by Age and Gender, April2009 (% of respondents in each group)

Male Female

12-17 55% 59%

18-24 61% 48%

25-34 64% 46%

35-44 52% 45%

45-54 43% 39%

55-64 49% 39%

Note: n=1,279; *includes music videos, news stories, moviepreviews/trailers/clips, short films, TV show clips, TV previews, weather,video game content and TV commercialsSource: Frank N. Magid Associates, "Magid Media Futures 2009:Opportunities in Online Video" sponsored by Metacafe, July 7, 2009


Amount of Attention Paid to Different Types of VideoContent According to US Online Video Viewers,January 2009 (scale of 1-10*)

Video clip shared by a friend 5.5

News highlights 5.4

Film trailer online 4.8

Video clip found on a video-sharing site 4.6

Music videos online 4.0

User-generated content online 3.9

Online ad 3.0

Note: ages 16-64; *1=low attention and 10=very high attentionSource: "Global Web Index" conducted by Lightspeed Research, May 28,2009




18


“If you’re talking about running prerolls in along list of user-generated content, then youwant to know what you’re running in,because there is a very large range of topics,of editorial content for that and you don’tnecessarily want to be associated witheverything in there.” —Sharon Gallacher, globalmedia director and managing director, Neo@Ogilvy,in an interview with eMarketer, July 2009

The habit of online video snacking—a short clip here, another onethere—creates basic problems for marketers wanting to reach thegeneral video audience. Since most of the streams people vieware from 1 to 5 minutes long, that limits the space for commercialmessages.And many short videos are not suited for mostadvertisers, since user-generated content—nearly all of it short—is a far less reliable vehicle for supporting brand messages.

That said, one study shows marketers shifting their perspectiveabout user-generated video.While 73% of advertisers told CollectiveMedia in 2008 that they would not run ads on user-generated video,that figure dropped by 10 percentage points in 2009.

103727

Where People Watch (and When)

Where people watch videos on the Web says a lot

about where marketers might want to place their

online video ad dollars.

The number of video viewers at any single Website is one indicatorof popularity. Counting heads for online video echoes the traditionalratings that Nielsen gives TV shows. However, as marketers havelearned time and again, online traffic alone is a highly imperfectgauge of marketing effectiveness.A smaller, focused audience thatspends time with professionally created videos is a better target forbrand marketers (who value engagement) than a broader audience(that views many user-generated snippets).

That said, 2009 data from both Nielsen Online and comScoreshows that far more Internet users visited YouTube each monththan any other video destination. However, as is often the case,the figures between the two research firms vary because ofdiffering methodologies. Nielsen, for example, includesprogressive downloads but excludes video advertising, whilecomScore includes users at university locations, as well as homeand work, in its samples.

Such variances lead to further disparities. For example, whileYahoo! is the second-largest video destination according toNielsen’s research, it falls behind Microsoft and Fox InteractiveMedia (mainly MySpace) in comScore’s data.

Types of Online Content on Which US Advertisers andthe Clients of US Agencies Will Not Run Ads, 2008 &2009 (% of respondents)

User-generated video

73%

63%

Blogs

39%

35%

Social media

27%

24%

News

13%

10%

Other

0%

23%

2008 2009

Note: 2008 n=200; 2009 n=417Source: Collective Media, "2009 Ad Network Study," provided to eMarketer,May 5, 2009




19


Perhaps more striking is the steady growth of the total onlinevideo audience: 6.6% from February to July 2009, according toNielsen, or 9.1% according to comScore.

106660

106648

Not all data is disparate—comScore and Nielsen agree that in July2009, for instance, about 76% of unique video viewers visitedYouTube (Google sites is nearly all YouTube, at this point).

Beyond the largest video site,however,data from the two firms barelymatches up when measuring share of total audience.But most of thesame sites appear on both companies’ monthly top 10 lists:Hulu,Turner,Yahoo!,Microsoft,Fox and MTV (Viacom).That consistencyindicates that despite the differing metrics, the mainstream onlinevideo audience can be readily found by marketers.

106661

106649

Unique Viewers at Top 10* Online Video Properties inthe US, February-July 2009 (millions and % change**)

Feb Mar Apr May Jun Jul % change**

Google sites 99.4 100.4 107.9 103.9 112.6 120.9 21.6%

Microsoft sites 28.5 32.2 32.0 38.1 72.9 64.5 126.5%

Fox Interactive Media 53.8 55.2 58.8 57.1 59.5 51.8 -3.7%

Yahoo! sites 41.7 42.5 45.4 44.2 50.3 47.4 13.6%

Viacom Digital 22.1 27.2 34.6 25.0 40.1 42.4 91.9%

Hulu 34.7 41.6 40.1 40.2 36.9 38.1 9.9%

Turner Network 20.1 19.6 34.2 28.7 52.1 35.7 77.5%

CBS Interactive 24.6 35.4 37.1 40.3 41.7 30.7 24.9%

AOL 22.8 22.3 23.7 26.9 24.1 24.2 6.0%

Facebook - - 13.3 18.0 17.0 20.5 -

Disney Online 12.7 13.0 - - - - -

Total video viewers 145.2 149.4 151.7 150.3 157.1 158.4 9.1%

Note: home, work and university locations; numbers do not add up to totalbecause Internet users view videos at more than one site; *sites includedin the top 10 and their rankings may vary by month; **% change fromFebruary to July 2009Source: comScore Video Metrix as cited in press releases, March-August2009; eMarketer calculations, September 2009


Share of Unique Viewers at Top 10* Online Video Sitesin the US, February-July 2009 (% of total)

YouTube

Yahoo!

MSN/Windows Live/Bing

Fox Interactive Media

CNN Digital Network

Hulu

Turner Sports andEntertainment Digital

Nickelodeon Kids and FamilyNetwork

MTV Networks Music

Blinkx

ABC.com

CBS Entertainment Network

Feb

69.1%

18.9%

9.6%

11.3%

6.6%

7.4%

4.7%

-

3.5%

-

5.3%

-

Mar

68.7%

19.0%

9.3%

11.3%

6.9%

6.8%

4.5%

4.9%

4.9%

-

5.3%

-

Apr

71.7%

19.1%

10.1%

12.1%

7.8%

6.4%

5.0%

4.0%

5.7%

-

4.3%

-

May

71.3%

18.9%

9.3%

12.0%

-

7.6%

5.1%

4.3%

5.0%

-

4.7%

-

Jun

73.7%

20.3%

12.0%

11.9%

8.1%

8.0%

5.7%

4.8%

6.9%

-

2.5%

5.0%

Jul

76.9%

22.0%

13.1%

10.8%

8.6%

7.6%

5.8%

5.3%

5.1%

0.2%

-

-

Note: includes progressive downloads and excludes video advertising;monthly shares total more than 100% because Internet users view videosat more than one site; *sites included in the top 10 and their rankings mayvary by monthSource: Nielsen Online, "VideoCensus" as cited in press releases,March-August 2009; eMarketer calculations, September 2009


Share of Unique Viewers at Top 10* Online VideoProperties in the US, February-July 2009 (% of total)

Feb Mar Apr May Jun Jul

Google sites 68.5% 67.2% 71.1% 69.1% 71.7% 76.3%

Microsoft sites 19.6% 21.6% 21.1% 25.3% 46.4% 40.7%

Fox Interactive Media 37.1% 36.9% 38.8% 38.0% 37.9% 32.7%

Yahoo! sites 28.7% 28.4% 29.9% 29.4% 32.0% 29.9%

Viacom Digital 15.2% 18.2% 22.8% 16.6% 25.5% 26.8%

Hulu 23.9% 27.8% 26.4% 26.7% 23.5% 24.1%

Turner Network 13.8% 13.1% 22.5% 19.1% 33.2% 22.5%

CBS Interactive 16.9% 23.7% 24.5% 26.8% 26.5% 19.4%

AOL 15.7% 14.9% 15.6% 17.9% 15.3% 15.3%

Facebook - - 8.8% 12.0% 10.8% 13.0%

Disney Online 8.7% 8.7% - - - -

Note: home, work and university locations; monthly shares total more than100% because Internet users view videos at more than one site; *sitesincluded in the top 10 and their rankings may vary by monthSource: comScore Video Metrix as cited in press releases, March-August2009; eMarketer calculations, September 2009


Unique Viewers at Top 10* Online Video Sites in theUS, February-July 2009 (millions and % change**)

YouTube

Yahoo!



CNN Digital Network

Hulu


Nickelodeon Kids andFamily Network

MTV Networks Music

Blinkx

ABC.com

CBS EntertainmentNetwork

Total video viewers

Feb

88.1

24.1

12.2

14.4

8.4

9.5

6.0

-

4.4

-

6.7

-

127.6

Mar

89.4

24.8

12.1

14.7

9.0

8.9

5.8

6.4

6.3

-

6.9

-

130.1

Apr

83.7

22.3

11.8

14.1

9.1

7.5

5.9

4.7

6.6

-

5.0

-

116.7

May

95.4

25.2

12.5

16.0

-

10.1

6.8

5.7

6.7

-

6.3

6.6

133.8

Jun

99.0

27.2

16.1

16.0

10.9

10.7

7.7

6.4

9.3

-

3.3

-

134.3

Jul

104.5

29.9

17.8

14.7

11.7

10.3

7.9

7.3

6.9

0.3

-

-

136.0

%change**

18.6%

24.3%

45.8%

2.2%

39.4%

8.7%

33.0%

-

5.7%

-

-

-

6.6%

Note: includes progressive downloads and excludes video advertising;numbers do not add up to total because Internet users view videos atmore than one site; *sites included in the top 10 and their rankings mayvary by month; **% change from February to July 2009Source: Nielsen Online, "VideoCensus" as cited in press releases,March-August 2009; eMarketer calculations, September 2009



20


How might the audience reach for a single popular show match upwith overall online video audience figures? In December 2008,1.4 million Internet users viewed ABC’s “Lost,” the top-rated TVnetwork show on the Web that month according to Nielsen Online.In fact, Nielsen found that for the same month,ABC.com rankedlast in the top 10 list, with 3.6 million unique viewers.That meansone show alone attracted 39% of the site’s visitors.

The power of the TV program as a brand is an important markerfor companies looking toward online video advertising.

101708

The number of videos streamed at each site is more useful thansheer traffic, since each video potentially gives marketers anopportunity to advertise.

Notably, the number of video streams at YouTube grew by 35.9%from February to July 2009, according to Nielsen, or 10 percentagepoints more than the 25.9% growth for total streams. It’s rarewhen the largest company has the greatest growth rate, too.

What’s more, the strong overall growth rate points to how onlinevideo is still an emerging market with a broad base, as the numberof videos streamed at sites outside the monthly top 10 alsoincreased by over 21%.

106656

The comScore data paints a vastly different picture, with nearly21.4 billion total streams in July 2009 versus Nielsen’s 11.2 billionfigure.The total growth rate for all US video sites is also far larger,at 63.5% from February to July—as are nearly all the individualvideo site growth rates.

Top 10 US Broadcast TV Network EntertainmentPrograms Viewed Online, Ranked by Unique Viewers,December 2008 (thousands)

1. Lost (ABC) 1,425

2. Saturday Night Live (NBC) 1,111

3. Grey's Anatomy (ABC) 879

4. Desperate Housewives (ABC) 723

5. Heroes (NBC) 685

6. Ugly Betty (ABC) 631

7. Samantha Who? (ABC) 560

8. Scrubs (ABC) 519

9. Survivor (CBS) 496

10. True Beauty (ABC) 462

Note: reflects video content on ABC.com, CBS Television, CWTV.com, FOXBroadcasting and NBC.com, including their respective embedded videoplayers; programs are client-definedSource: Nielsen Online, "VideoCensus" as cited in press release, February12, 2009


Total Videos Streamed at Top 10* Online Video Sites inthe US, February & July 2009 (millions and %change**)

YouTube

Hulu

Yahoo!


Nickelodeon Kids and Family Network

Turner Sports and Entertainment Digital


MTV Networks Music

CNN Digital Network

Blinkx

ABC.com

CBS Entertainment Network

Other sites

Total streams

February

5,158.7

308.8

250.4

162.9

209.5

125.1

194.3

100.1

99.8

-

187.1

-

2,101.2

8,897.9

July

7,010.8

383.8

265.6

188.0

179.7

155.1

151.6

119.1

109.2

86.4

-

-

2,550.9

11,200.1

%change**

35.9%

24.3%

6.0%

15.4%

-14.2%

24.0%

-21.9%

19.0%

9.4%

-

-

-

21.4%

25.9%

Note: includes progressive downloads and excludes video advertising;*sites included in the top 10 and their rankings may vary by month; **%change from February to July 2009Source: Nielsen Online, "VideoCensus" as cited in press releases,March-August 2009; eMarketer calculations, September 2009



21


Note that the comScore data combines multiple sites into acorporate property, such as Viacom Digital, which grew by over227%. In contrast, Nielsen breaks down the individual sites, suchas Nickelodeon, MTV Music and so forth.

106653

Among all the top video sites,Hulu’s growth is particularlynoteworthy—not just for its usage but more for the audienceawareness.As the largest third-party site for professionally madevideo,Hulu promises to become an even bigger player for onlinevideo advertising.Ranked by number of videos viewed, it is fifth inthe comScore data for July 2009 and second in the Nielsen research.

Further, the audience’s response in the Ipsos survey showed asharp increase in the video site’s own brand awareness.Whileonly 9% of respondents in the ongoing study had heard of Hulu inSeptember 2008, by April 2009 that figure had soared to 41%.

106713

Still, when it comes to viewing professionally made content, 63%of Internet users who accessed full TV content in 2008 wentdirectly to the network sites, according to Knowledge Networks.Only 28% used third-party sites such as Hulu.

101751

When you derive the average number of videos each visitor viewson any site—calculated from total videos and unique trafficfigures—you begin to get the clearest picture yet of how usefulthat site might be for video advertising.

Blinkx, a video search engine, is an anomaly in July 2009.And whileYouTube continues at a high level of usage, the four sites below itin the Nielsen data represent a sweet spot for advertisers:professional content only, blended with an active audience.

Among all four of those sites—Hulu, Nickelodeon,Turner Sportsand MTV Music—marketers can find both long-form and shortvideos, a wide range of ages and genders, and varied types ofcontent to appeal to different audiences.

Total Videos Viewed at Top 10* Online VideoProperties in the US, February & July 2009 (millionsand % change**)

February July % change**

Google sites 5,348.6 8,953.9 67.4%

Viacom Digital 248.1 812.3 227.4%

Microsoft sites 259.0 630.6 143.5%

Fox Interactive Media 462.6 558.5 20.7%

Hulu 332.5 457.0 37.4%

Turner Network 179.5 390.8 130.6%

Yahoo! sites 353.5 374.7 6.0%

Disney Online 116.1 169.8 46.2%

CBS Interactive 111.8 150.2 34.3%

AOL 117.1 - -

Blinkx - - -

ABC.com - 137.8 -

Others 5,553.4 8,735.6 57.3%

Total videos viewed 13,072.2 21,371.3 63.5%

Note: home, work and university locations; *sites included in the top 10and their rankings may vary by month; **% change from February to July2009Source: comScore Video Metrix as cited in press releases, March-August2009; eMarketer calculations, September 2009


US Online Video Viewers’ Awareness and Usage* ofSelect Digital Video Providers, September 2008 &April 2009 (% of respondents)

Awareness Usage*

September 2008April 2009September 2008April 2009

YouTube 83% 87% 49% 56%

iTunes Store 46% 49% 8% 10%

MySpace 54% 46% 13% 9%

Hulu 9% 41% 3% 14%

Note: September 2008 n=1,102; April 2009 n=939; ages 12+; *in past 30daysSource: Ipsos MediaCT, "MOTION study," September 9, 2009


Methods Used by US Internet Users to Stream TVNetwork Content, by Age, 2006 & 2008 (% ofrespondents*)

13-54 13-17 18-34 35-49 50-54

2006

Use TV networkWebsites to viewvideo

61% 56% 61% 65% 53%

Use third-partyvideo sites toview networkvideo

16% 27% 15% 12% 8%

2008

Use TV networkWebsites to viewvideo

63% 51% 68% 64% 53%

Use third-partyvideo sites toview networkvideo

28% 26% 30% 26% 32%

Note: *among those who access full TV programs onlineSource: Knowledge Networks Inc., "How People Use TV's WebConnections" as cited in press release, February 12, 2009



22


Notable, too, is how the growth rates for Microsoft,Yahoo! andMySpace (Fox Interactive Media) showed declining usage over thesix-month period.As the online video audience grows morefamiliar with offerings, it appears to be gravitating directly towardvarious TV network sites, as well as Hulu.

106665

Based on average videos per viewer, the rankings from comScoreare essentially in accord with Nielsen’s.That is,YouTube gets themost usage per person, followed by a mix of Viacom (MTV andNickelodeon), Hulu and Turner.

106650

When People WatchFor companies looking for specific markets, daypart targeting forvideo advertising appears to make sense. Data from comScoreshows the peak viewing time in June 2009 was weekday daytime,with 27.3% of the total online video audience.

106679

Average Streams per Viewer at Top 10* Online VideoSites in the US, February-July 2009

Blinkx

YouTube

Hulu

Nickelodeon Kids andFamily Network


MTV Networks Music



CNN Digital Network

Yahoo!

ABC.com

CBS EntertainmentNetwork

Total streams perviewer

Feb

-

58.5

32.6

-

21.0

22.7

13.4

13.5

11.9

10.4

27.9

-

69.7

Mar

-

61.3

39.3

30.7

23.6

19.5

14.0

14.1

11.5

9.4

25.7

-

74.4

Apr

-

65.6

50.1

37.5

22.3

21.7

14.0

14.3

12.4

9.1

29.6

-

81.0

May

-

63.4

37.8

26.9

20.6

18.9

11.9

11.9

-

8.3

27.8

14.8

75.1

Jun

-

62.0

32.6

24.4

19.5

14.8

12.1

10.4

9.5

8.6

23.8

-

75.1

Jul

322.4

67.1

37.3

24.8

19.6

17.3

10.6

10.3

9.4

8.9

-

-

82.4

%change**

-

14.6%

14.3%

-

-6.8%

-23.8%

-20.9%

-23.6%

-21.5%

-14.7%

-

-

18.1%



Average Videos per Viewer at Top 10* Online VideoProperties in the US, February-July 2009

Google sites

Viacom Digital

Hulu

Turner Network


Microsoft sites

Yahoo! sites

AOL

CBS Interactive

Facebook

Disney Online

Total videos perviewer

Feb

53.8

11.3

9.6

8.4

8.6

9.1

8.5

5.1

4.5

-

9.1

90.0

Mar

59.0

10.2

9.1

8.5

7.9

9.0

7.9

4.7

5.4

-

9.7

96.8

Apr

63.3

9.1

9.9

8.0

8.7

9.0

7.8

5.1

5.5

3.3

-

110.7

May

63.6

11.2

10.2

9.1

10.1

8.1

7.1

5.9

5.2

4.0

-

109.8

Jun

67.8

19.3

10.1

9.5

10.1

9.5

7.4

5.6

4.7

4.0

-

124.1

Jul

74.1

19.2

12.0

11.0

10.8

9.8

7.9

5.2

4.9

4.0

-

134.9

% change**

37.7%

69.9%

25.0%

31.0%

25.6%

7.7%

-7.1%

2.0%

8.9%

-

-

49.9%

Note: home, work and university locations; *sites included in the top 10and their rankings may vary by month; **% change from February to July2009Source: comScore Video Metrix as cited in press releases, March-August2009; eMarketer calculations, September 2009


Share of Time that US Internet Users Spend ViewingOnline Video, by Time of Day/Week, June 2008 vs. June2009 (% of total and % change)

June 2008 June 2009 % change

Early morning (M-F, 7am-10am) 6.0% 7.1% 18.8%

Daytime (M-F, 10am-5pm) 28.0% 27.3% -2.3%

Early evening (M-F, 5pm-8pm) 12.0% 11.3% -5.6%

Evening (M-F, 8pm-11pm) 11.8% 11.2% -5.5%

Late evening (M-F, 11pm-1am) 6.3% 7.3% 14.8%

Late night (M-F, 1am-7am) 7.6% 10.2% 33.7%

Weekday, all day 71.7% 74.4% 3.7%

Morning, weekend 5.4% 4.7% -12.8%

Afternoon, weekend 8.1% 7.6% -5.2%

Evening, weekend 7.8% 5.8% -25.2%

Late evening, weekend 3.1% 2.9% -4.8%

Late night, weekend 4.0% 4.5% 14.0%

Weekend, all day 28.3% 25.6% -9.5%

Total 100.0% 100.0% 0.0%

Source: comScore as cited by Advertising Age, September 2009; eMarketercalculations, September 2009



23


However, the greatest growth in time spent watching was also onweekdays, but during late night (at 205%), early morning (171%)and late evening (162%).

The night and evening slots tend to be more relaxed times formost individuals, which potentially means the growth of long-formprofessional content online can be accompanied by ads daypart-targeted to those time slots.

106678

Who Watches (and How)

The 144 million people in the US who will watch

online video in 2009 represent 72% of Internet

users and 47% of the total population. By 2013,

those numbers will shift to 188 million, or 85%

and 59%, respectively. By any standard, the

audience has not only reached critical mass, but

is still sharply on the rise.

105370


Research from Deloitte matches up with common perceptions, too,with a higher percentage of younger Internet users inSeptember–October 2008 watching videos weekly than older users.

105582

Amount of Time that US Internet Users SpendViewing Online Video, by Time of Day/Week, June2008 vs. June 2009 (billions of minutes and % change)

June 2008 June 2009 % change

Early morning (M-F, 7am-10am) 1.86 5.05 171%

Daytime (M-F, 10am-5pm) 8.72 19.45 123%

Early evening (M-F, 5pm-8pm) 3.73 8.05 116%

Evening (M-F, 8pm-11pm) 3.69 7.96 116%

Late evening (M-F, 11pm-1am) 1.97 5.17 162%

Late night (M-F, 1am-7am) 2.38 7.28 205%

Weekday, all day 22.35 52.96 137%

Morning, weekend 1.69 3.36 99%

Afternoon, weekend 2.51 5.44 116%

Evening, weekend 2.43 4.15 71%

Late evening, weekend 0.96 2.08 117%

Late night, weekend 1.23 3.21 160%

Weekend, all day 8.82 18.24 107%

Total 31.17 71.19 128%

Source: comScore as cited by Advertising Age, September 2009; eMarketercalculations, September 2009


US Online Video Viewers and Penetration, 2008-20132008 2009 2010 2011 2012 2013

Online video viewers (millions) 135 144 154 165 177 188

% of Internet users 70% 72% 75% 78% 82% 85%

% of population 44% 47% 50% 53% 56% 59%

Note: individuals who download or stream video at least once per monthSource: eMarketer, July 2009


Weekly Online Video Watching Activities of USInternet Users, by Generation, October 2008 (% ofrespondents in each group)

Millennials (14-25)

67%

49%

Generation X (26-42)

62%

47%

Boomers (43-61)

40%

27%

Matures (62-75)

26%

18%

Watch movies and other videos Watch TV shows

Note: use frequently/occasionallySource: Deloitte, “State of the Media Democracy Third Edition," provided toeMarketer, January 12, 2009



24


However, the common perception that active viewing of onlinevideo is almost only among a younger audience is off-target.Anongoing study of this audience by Ipsos MediaCT found that in June2008, 60% of those ages 35 to 54 had watched video in the previous30 days.The only other group with a greater share was teens.

099157

Another perspective, this from the Veoh/Forrester study, breaksdown video viewers not as a percentage of each age, but whatages make up the entire active online video audience. In this study,“active” refers to those who watch more than 1 hour of onlinevideo each week.

By that standard, the age skew is lower, as common wisdomwould have it, with 57% of the total ages 13 to 34.

098732

But just because the numbers are there does not make it easy formarketers to reach the online video audience. It is not even easyto accurately gauge the size of the audience.

In a comparative estimate of how many Internet users watchvideo or how often, the higher numbers range from Nielsen’s 76%who watched video on the Internet in Q4 2008, to eMarketer’s 72%share, to the 62% in January 2009 who told Lightspeed Researchthat they watched one or more videos in the previous week.

At the lower end, 23% watched online video (MediaMarkResearch) or viewed longer video content (Ofcom/Synovate),down to the 12% of respondents in the Magid/Metacafe surveywho watched videos daily in April 2009.

With comparative estimate percentages as in this chart,definitions are central. For example, is the researcher measuringInternet users in general, video watchers in particular orconsumers overall? Or is the researcher gauging particular typesof video viewing, such as just TV shows and other professionalcontent? Further, the varied timeframes—from September 2008through 2009—also contribute to the broad spread of numbers.

Demographic Profile of US Online Video Viewers*,December 2007 & June 2008 (% of respondents ineach group)

December 2007 June 2008

Gender

Male 55% 58%

Female 45% 54%

Age

12-17 73% 71%

18-24 59% 58%

25-34 52% 60%

35-54 49% 60%

55+ 32% 38%

Note: ages 12+; *in the past 30 daysSource: Ipsos MediaCT, "Motion Study" as cited in press release, October27, 2008


Demographic Profile of US Active* Online VideoViewers, by Segment, August 2008 (% of respondents)

Watchers Controllers Connectors Total

Gender

Female 50% 51% 53% 51%

Male 50% 49% 47% 49%

Age

13-34 55% 58% 60% 57%

35-54 31% 30% 28% 30%

55+ 15% 12% 13% 13%

Income

<$75,000 63% 66% 71% 66%

$75,000+ 37% 34% 29% 34%




25


Nevertheless, the wide span of percentages indicates howdetermining the size and usage patterns of this audience is still ata toddler’s level.


Frequency and TimeThere is a sharp division between regular video viewers andsporadic or non-users.According to Magid/Metacafe, 43% ofInternet users watch online video once per week or more often,while 48% watch less than once per month or never.

This split will likely narrow as the increase of professional videocontent available on the Web attracts a portion of that lower-usesegment and ups usage among those already watching online video.

105278

No matter what the age or gender, people spent more timeviewing video on the Web in April 2009 than only six monthsearlier, according to Nielsen Online. In fact, the greatest increase inaverage minutes viewed per person, at 29%, came in the 35 to 49age bracket.

When you consider this group grew up during the time that cable TV became a major force among media, it is not surprising that they are increasingly gravitating to online’s video-on-demand environment.

104094

Frequency with Which US Internet Users WatchOnline Video, April 2009 (% of respondents)

Daily 12%

Several times per week 19%

Once per week 12%

Once per month or more 10%

Less than once per month 15%

Never 33%

Note: n=1,927 ages 12-64; numbers may not add up to 100% due toroundingSource: Frank N. Magid Associates, "Magid Media Futures 2009:Opportunities in Online Video" sponsored by Metacafe, July 7, 2009


Average Time Spent Viewing Online Video Among USInternet Users, by Age and Gender, November 2008 &April 2009 (minutes per viewer and % change)

November 2008

April 2009

% change

Gender

Male 209 249 19%

Female 151 170 12%

Age

2-11 113 116 3%

12-17 178 190 7%

18-24 303 349 15%

25-34 253 296 17%

35-49 187 243 29%

50-64 122 139 14%

65+ 67 81 21%

Total 178 206 16%

Note: home and work locationsSource: Nielsen Online, "VideoCensus" as cited in press release, May 14,2009


Comparative Estimates: Frequency/Usage of OnlineVideo Among US Internet Users, 2008 & 2009 (% ofrespondents)

The Nielsen Company,February 2009

eMarketer, July2009

Lightspeed Research,May 2009

Ipsos MediaCT,December 2008

Deloitte, January2009

Metacafe & Frank N.Magid Associates,July 2009


Arbitron & EdisonMedia Research,April 2009

MediamarkResearch,November 2008

Ofcom-UK &Synovate,November 2008


Metacafe & Frank N.Magid Associates, July 2009

Frequency/usage

Watched videoon Internet

Download orstream video atleast once permonth

Viewed one ormore videosin past week

Streamed videoin the past 30 days

Watch movies andother videos weekly

Watch videosweekly

Watch TV showsweekly

Viewed videoin the past week

Watched onlinevideo in past 30days

Watch/downloadlonger video content,e.g., feature filmsor complete TVprograms

Downloaded videofile in past 30 days

Watch videosdaily

% ofrespon-dents

76%*

72%**

62%

57%

52%

43%

37%

27%***

23%

23%

22%

12%

Age

2+

allages

16-64

12+

14-75

12-64

14-75

12+

18+

18-64

12+

12-64

Surveydate

Q4 2008

2009

January2009

September2008

2008

April2009

2008

2009

September2007-

October 2008

October2008

September2008

July2009

Note: *not survey data; eMarketer calculations based on rounded figuresfrom Nielsen; **not survey data; eMarketer estimate; ***among total USconsumers, not just Internet usersSource: eMarketer, July 2009; various, as noted, 2008 & 2009



26


Further Nielsen data reinforces that trend toward overall higherusage.While the number of unique viewers grew by only 6.6%from February to July 2009, consider these other growth rates forthe same period:

■ 25.9% for total streams viewed

■ 25.0% for average time per viewer

■ 33.2% for total time spent watching videos among all unique viewers

106668

Even with the large online gains, the amount of time people spendwatching video on the Web pales compared with TV time.

When the total time spent watching video is divvied up among thethree screens—television, Internet and mobile—online video (notto mention mobile) seems like a rounding error.As recently as Q22009, according to Nielsen, only 1% of total video time occurred ata computer, while 98.5% was spent in front of a TV.

Note, though, that this data looks at all US consumers, not justInternet users. But for marketers, all consumers are potentialtargets for their products—and for broad reach alone, they are stillbest reached via TV.

106716

“If people actually watched more online video,then it would be bigger.And they’re notwatching that much of it—it’s just not thatbig.” —Brian Wieser,SVP global director of forecasting,MAGNA, in an interview with eMarketer, June 2009

At this point, a significant minority of people—20% according tothe Magid/Metacafe study—are reducing their TV-watching habitsdue to the availability of online video.

105286

US Online Video Viewer Metrics, February-July 2009

Unique viewers (millions)

Total streams (billions)

Average streams perviewer

Average time spent perviewer (minutes)

Average time per stream(minutes)

Total time spent (millionsof hours)

Feb

127.6

8.90

69.7

169.3

2.43

360.1

Mar

130.1

9.67

74.4

190.7

2.56

413.4

Apr

116.7

9.45

81.0

205.7

2.54

400.1

May

133.8

10.04

75.1

188.7

2.51

420.8

Jun

134.3

10.08

75.1

188.8

2.52

422.6

Jul

136.0

11.20

82.4

211.6

2.57

479.5

%change**

6.6%

25.9%

18.2%

25.0%

5.8%

33.2%



Share of Total Time US Consumers Spend per MonthViewing Video on TV, Internet and Mobile Phone, Q12008-Q2 2009 (% of total)

Watching TV in the home

Watching timeshifted TV

Watching video on Internet

Mobile subscribers watchingvideo on a mobile phone

Q12008

99.3%

0.8%

0.5%

-

Q22008

99.0%

1.0%

0.7%

0.1%

Q32008

98.9%

1.1%

0.8%

0.1%

Q42008

98.8%

1.2%

0.8%

0.1%

Q12009

98.6%

1.5%

0.9%

0.1%

Q22009

98.5%

1.5%

1.0%

0.1%

Note: total time calculated by multiplying reach for each medium byaverage hours per month; total is greater than 100% becauseapproximately 75% of timeshifted TV is also counted under TV in the homeSource: The Nielsen Company, "A2/M2 Three Screen Report," February,May and September 2009; eMarketer calculations, September 2009


Change in Time Spent Watching TV due to OnlineVideo According to US Online Video Viewers, April2009 (% of respondents)

More 9%

The same amount 64%

Less 20%

Don't spend any time doing that activity 7%

Note: n=1,279 ages 12-64Source: Frank N. Magid Associates, "Magid Media Futures 2009:Opportunities in Online Video" sponsored by Metacafe, July 7, 2009




27


Even as online video consumes a bare percentage of total time,Nielsen found that the annual growth rate hit 62.7% in Q2 2009, asimilar figure to its conceptual cousin, timeshifted TV through aDVR (both allow the audience to watch when they want). Incontrast, watching TV in the home inched up by only 2.4% duringthat same quarter.

106718

However, the relatively low growth rate for time spent watching TVin the home belies the actual hourly increase. From Q2 2008 to Q22009, the total hours US consumers spent on that activityincreased by 951 million. Even with its rapid growth, the totalhours for online video increased by only 164 million.

With such changes, it is hard to picture online video topplingtelevision usage anytime soon.

106715

Conclusions

The potential audience for online video

advertising does not much trust the ads—and

only about one-third actively enjoys them. So

what’s the ad industry to do?

For some the solution is same old, same old—simply pushing videocommercials much like on TV, and trusting that the combination offrequency and captive audience will get the message across. For asubstantial slice of advertisers, that method will actually berelatively effective in the interactive space—if perhaps costly.

However, marketers should consider these takeaways:

■ Only 1% of total video viewing time goes to Web video. Toreach the online video audience, marketers will need to payattention not only to sheer audience numbers but to how peopleinteract with video.

■ There are multiple points of disconnect between the TVaudience and the online video audience. They are not oneand the same—the Internet audience does not necessarilyrespond to the same ads in the same way they would afterviewing them on TV.Therefore, developing ad creative just forInternet video will give marketers a competitive edge.

■ The audience has not displayed a definite preference fortypes of video ads. Marketers, then, should continue toexperiment with different video ad formats or combine themwith other types of ads.

■ The audience is mixed about free-or-fee revenue models.There is no absolute consensus on ad-supported video contentversus some form of payment to avoid advertising altogether oreven just reduce it. If the value of video content is high enough,marketers and content providers together can attract a largeaudience through a fusion of fewer ads (less intrusive) and lowerfees (less prohibitive).

For more information about online video advertising,see eMarketer’s previous report in this series,“Digital Video Advertising: Where’s the Money?”

Total Time US Consumers Spend per Month ViewingVideo on TV, Internet and Mobile Phone, Q4 2008-Q22009 (% change*)



Using the Internet


Mobile subscribers watching videoon a mobile phone

Total

Q42008

5.0%

82.4%

7.0%

-

-

6.2%

Q12009

3.1%

92.3%

8.0%

73.9%

-

4.2%

Q22009

2.4%

57.9%

18.4%

62.7%

52.4%

4.5%

Note: total time calculated by multiplying reach for each medium byaverage hours per month; *change from average month in quarter vs.average month in same quarter of prior yearSource: The Nielsen Company, "A2/M2 Three Screen Report," February,May and September 2009; eMarketer calculations, September 2009


Total Time US Consumers Spend per Month ViewingVideo on TV, Internet and Mobile Phone, Q1 2008-Q22009 (millions of hours)



Using the Internet


Mobile subscriberswatching video on amobile phone

Total

Q12008

42,344

340

4,416

226

-

47,071

Q22008

39,163

379

4,237

262

33

43,789

Q32008

39,746

436

4,370

303

37

44,565

Q42008

43,097

531

4,372

355

41

47,998

Q12009

43,668

653

4,771

393

49

49,044

Q22009

40,114

598

5,015

426

50

45,754

Note: total time calculated by multiplying reach for each medium byaverage hours per month; total line is less than sum of individualcategories because approximately 75% of timeshifted TV is also countedunder TV in the homeSource: The Nielsen Company, "A2/M2 Three Screen Report," February,May and September 2009; eMarketer calculations, September 2009




28

Endnotes

Endnote numbers correspond to the unique

six-digit identifier in the lower left-hand corner

of each chart. The charts from the report are

repeated before their respective endnotes.

098734 | 098735

098734

098735

Extended Note: "Watchers" spend a little more than an hourwatching video each week but besides showing up to watch, donot engage the experience deeply by controlling playback orsharing videos. "Controllers" are younger viewers who take anactive role in controlling their video experiences and feel thatonline video is important in their lives. "Connectors" are typicallyeven younger and, representing just 7% of online viewers,consume 20% of all online video and are responsible for 42% of allthe online video-sharing.

Ad Responses of US Active* Online Video ViewersDuring Their Most Recent Online Video Site Visit, bySegment, August 2008 (% of respondents)

Noticed the product or brand that was advertised during myvideo viewing experience

48%

34%

34%

Clicked on an ad and registered for a sweepstakes or giveaway

12%

3%

4%

Clicked on an ad that was part of or next to a video I waswatching and learned more about the product or service

9%

3%

5%

Clicked on an ad and signed up for more information

6%

3%

2%

Purchased a product

4%

2%

2%




Attitudes Toward In-Video Ads Among US Active*Online Video Viewers, by Segment, August 2008 (% ofrespondents)

In-video ads are annoying

61%

59%

58%

In-video ads are fair as they help pay for online video services

41%

37%

34%

I avoid Websites that use in-video advertising

19%

21%

23%

In-video ads are useful when they feature something I aminterested in

28%

19%

19%

In-video ads can be entertaining

23%

17%

10%





29

Endnotes

099240

099240

Extended Note: "Connected consumers" are Internet users withaccess to broadband who spent $200 online in the past year(travel, Netflix, tickets,Amazon gifts, etc.); visited a "communitysite" (MySpace,YouTube, Facebook, Classmates,Wikipedia, etc.);and consumed or created some form of digital media, such asphotos, videos, music or news.

105370

105370

Extended Note: eMarketer defines an Internet user as a personof any age who uses the Internet from any location at least onceper month. Population estimates used to calculate penetrationrate are based on US Census Bureau data.

105581

105581

Citation: Arbitron and Edison Media Research, "The Infinite Dial2009" sponsored by TargetSpot,April 16, 2009; Deloitte, "State ofthe Media Democracy Third Edition," provided to eMarketer,January 12, 2009; Frank N. Magid Associates, "Magid Media Futures2009: Opportunities in Online Video" sponsored by Metacafe, July 7,2009; Ipsos MediaCT, "Motion" as cited in press release, December18, 2008; "Global Web Index" conducted by Lightspeed Research,May 28, 2009 with eMarketer calculations; Mediamark ResearchInc. (MRI), "Survey of the American Consumer, Fall 2008" as cited inpress release, November 20, 2008; Office of Communications(Ofcom) - UK, "The International Communications Market 2008”conducted by Synovate, November 20, 2008;The Nielsen Company,"A2/M2 Three Screen Report - 4th Quarter 2008," February 23, 2009with eMarketer calculations

Online Advertising Formats Preferred by USConnected Consumers, June 2008 (% of respondents)

Banner ads on same page 44.44%

Preroll 33.63%

Interstitials 13.10%

Tickers 8.83%

Source: Razorfish, "Digital Consumer Behavior Study" as cited in "FEED:The Razorfish Consumer Experience Report," October 2008


US Online Video Viewers and Penetration, 2008-20132008 2009 2010 2011 2012 2013

Online video viewers (millions) 135 144 154 165 177 188

% of Internet users 70% 72% 75% 78% 82% 85%

% of population 44% 47% 50% 53% 56% 59%

Note: individuals who download or stream video at least once per monthSource: eMarketer, July 2009


Comparative Estimates: Frequency/Usage of OnlineVideo Among US Internet Users, 2008 & 2009 (% ofrespondents)

The Nielsen Company,February 2009

eMarketer, July2009

Lightspeed Research,May 2009



Metacafe & Frank N.Magid Associates,July 2009


Arbitron & EdisonMedia Research,April 2009

MediamarkResearch,November 2008

Ofcom-UK &Synovate,November 2008


Metacafe & Frank N.Magid Associates, July 2009

Frequency/usage

Watched videoon Internet

Download orstream video atleast once permonth

Viewed one ormore videosin past week

Streamed videoin the past 30 days

Watch movies andother videos weekly

Watch videosweekly

Watch TV showsweekly

Viewed videoin the past week

Watched onlinevideo in past 30days

Watch/downloadlonger video content,e.g., feature filmsor complete TVprograms

Downloaded videofile in past 30 days

Watch videosdaily

% ofrespon-dents

76%*

72%**

62%

57%

52%

43%

37%

27%***

23%

23%

22%

12%

Age

2+

allages

16-64

12+

14-75

12-64

14-75

12+

18+

18-64

12+

12-64

Surveydate

Q4 2008

2009

January2009

September2008

2008

April2009

2008

2009

September2007-

October 2008

October2008

September2008

July2009

Note: *not survey data; eMarketer calculations based on rounded figuresfrom Nielsen; **not survey data; eMarketer estimate; ***among total USconsumers, not just Internet usersSource: eMarketer, July 2009; various, as noted, 2008 & 2009



30

Endnotes

106674 | 106676 | 106669

106674

106676

106669

Extended Note: In the campaigns, there were three types of ads:preroll (5-second and 30-second), lower 1/3 (semitransparentoverlay that launches from the bottom third of the video screenapproximately 10 seconds after video content begins), andsideloader (custom unit that slides out of the right-hand side of thevideo screen approximately 10 seconds after video contentbegins).Video ads were from three companies (movie studio,beauty care, CPG) and were measured over 50 million impressionsacross five sites (MTV,VH1, CMT, Comedy Central and Nick Jr).

106682

106682

Extended Note: Dwell time (time spent) is defined as theaverage number of seconds consumers spend engaging with anonline ad, which includes the amount of time the mouse was overan ad, user-initiated video duration, user-initiated expansionduration and other user-initiated custom interaction duration.Unintentional dwell instances of less than 1 second are excluded.

US Internet Users’ Attitudes Regarding Online VideoAdvertising Campaigns on MTV Networks, by Type ofAd, January-April 2009 (% of respondents*)

More intrusive than other ad formats

Liked that it didn’t interrupt the video

An ad I would stop and look at

The way the ad was delivered was cool

An ad I would click on or interact with

Length of the ad was appropriate

Ad was short and let me get to myvideo quickly

An ad I’d like to see again

5-secondpreroll

+lower 1/3

24.7%

64.5%

35.0%

44.1%

23.4%

61.0%

68.4%

24.6%

5-secondpreroll

+sideloader

29.2%

57.9%

35.2%

43.2%

25.1%

60.5%

64.2%

26.5%

30-secondpreroll

30.0%

61.8%

42.2%

45.7%

25.1%

63.1%

58.6%

29.8%



US Internet Users Who Liked Elements of OnlineVideo Advertising Campaigns on MTV Networks, byType of Ad, January-April 2009 (% of respondents*)

Animation or motion of the ad

Movement of ad across the screen

How the ad appears and then disappears

Ability to expand the ad

Ability to click a link in the ad

Ability to see a longer video ad with aclick

5-secondpreroll

+lower 1/3

38.8%

32.4%

43.9%

31.7%

36.4%

34.2%

5-secondpreroll

+sideloader

39.8%

37.9%

47.9%

33.9%

38.8%

36.3%

30-secondpreroll

40.4%

30.9%

38.5%

29.4%

30.9%

29.5%



Brand Metrics for US Online Video AdvertisingCampaigns on MTV Networks, by Type of Ad,January-April 2009 (% of respondents)

Control 5-second preroll

+ lower 1/3

5-second preroll

+ sideloader

30-second preroll

Unaided awareness 12.6% 22.9% 17.6% 19.8%

Aided awareness 72.9% 81.0% 79.4% 81.7%

Online ad awareness 29.2% 56.2% 56.1% 47.5%

Purchase intent 41.4% 44.2% 36.3% 37.9%

Note: n=2,306 (control); n=403 (5-second preroll + lower 1/3); n=427(5-second preroll + sideloader); n=353 (30-second preroll)Source: MTV Networks, "Project Inform" conducted by InsightExpress, July2009; eMarketer calculations, September 2009


Average Time Spent on Online Advertising AmongInternet Users in the US and Canada, 2008

Dwell rate* Average dwell time(seconds)

With video 7.53% 85.16

No video 5.56% 32.73

General 6.54% 57.58

Note: excludes messenger ads; *cumulative instances of where a userinteracted with the ad divided by served impressionsSource: Eyeblaster, "Trends of Time and Attention in Online Advertising,"provided to eMarketer, July 22, 2009



31

Related Information and Links

Related LinkscomScorehttp://www.comscore.com

Knowledge Networkshttp://www.knowledgenetworks.com

Metacafehttp://www.metacafe.com

Nielsen Onlinehttp://en-us.nielsen.com

Veoh Networkshttp://www.veoh.com

ContacteMarketer, Inc. Toll-Free: 800-405-084475 Broad Street Outside the US: 212-763-601032nd floor Fax: 212-763-6020New York, NY 10004 [email protected]

Report ContributorsSusan Reiter Managing EditorJoanne DiCamillo Production ArtistDana Hill Production ArtistJared Jenks Numbers EditorJames Ku Data Entry Associate

and Production ArtistNicole Perrin Senior EditorAllison Smith Director of ChartsTracy Tang Senior Researcher

About eMarketer

eMarketer is a business service unlike any

other. eMarketer does not conduct research. It

aggregates and analyzes all the available

research, surveys and data on a given topic. So

you see far more than a single source could ever

provide. eMarketer doesn’t deliver one point of

view, it gives you thousands. Providing overviews

and insights available nowhere else, eMarketerreports on the reports, analyzes the analysts and

dissects the data.

A Trusted Resource eMarketer serves as a trusted, third-party resource, cuttingthrough the clutter and hype–helping businesses make sense ofthe numbers and trends. eMarketer's products and services helpcompanies make better, more informed business decisions by:

■ Streamlining research sources and reducing costs

■ Eliminating critical data gaps

■ Providing an objective, bird’s eye view of the entire landscape

■ Better deploying and sharing information across the company

■ Building solid business cases backed up by hard data

■ Reducing business risk

■ Saving valuable time

To learn more about subscriptions to eMarketer,call 800-405-0844 (outside the U.S. and Canada,call 001-212-763-6010), or e-mail [email protected].

2009.10.05 --e marketer---marketing-to-online-video-audience

Business

Transcript of 2009.10.05 --e marketer---marketing-to-online-video-audience