“Kill Your Television:” A Video Advertising White Paper from Specific Media


ADOTAS – With online video viewing exploding internationally and across multiple demographics, advertisers and brands are looking for ways to reach this growing, eager audience. But the lingering question is: What actually works? How do you speak to the user experience — and how is the online user experience different from or similar to watching TV? How will people respond to ads in this format, where they’re often looking for a specific video that they want to watch immediately — will they respond favorably? What about whether they’re watching what could be considered “branded content?”

Actually, those are a whole bunch of lingering questions. Hoping to answer them and more, Specific Media recently launched the metric VITAMIN (Video Testing and Measurement Insights), which aimed to look at the way viewers responded, in an emotional or psychological sense, to being shown video ads. After testing the insights in the U.K. during the fall of 2011, the company assembled a white paper called Kill Your Television. Specific Media has shared the paper with ADOTAS — you can read the text below.



Introduction – A snapshot of video advertising in the UK

Without doubt video has been the growth story within digital over the past 18 months, and it shows no sign of slowing down. Cisco forecast that inside two years video traffic will account for 90% of all internet traffic in the UK and in the US, Netflix² alone accounts for some 33% of peak downstream traffic. Across the EU-7 some 226 million people watch an average of 20 hours of online video each.

The explosion of high speed broadband, affordability and evolution of devices and increasingly fragmented media landscape, allied with an increasingly time poor consumer, are among the factors that have colluded to generate such growth in video and alter the way the public consume content.

More anecdotal evidence of this shift in consumer consumption can be seen in retailers such as John Lewis, where headphone sales are up 80% versus last year, with a pair sold every 82 seconds as media consumption (audio and visual) becomes more personal. On tablets, people who watch video on these devices do so for 28% longer than those who view on a desktop.

Also the small, but significant, fall for the first time in 20 years in the number of US households that own a TV set, down from 98.9% to 96.7%.

Importantly, the video market is not the preserve of early adopters or young men – maturity can be seen in the fact that men under 35 make up some 19.6% of video viewers in Britain vs 44.9%⁷ in a less mature market such as Turkey.

Not surprisingly, given the increase in viewing amongst a broad spread of consumer segments, advertisers are not missing out either – video advertising spend in the UK is up 91% in the past 12 months and is forecasted to continue growing.

The viewer experience

The proliferation in high quality audio visual content online has transformed the media landscape, with newcomers such as Hulu, Vevo and Daily Motion sitting happily alongside established media names like the BBC, Daily Telegraph and Sky.

It has also transformed the viewer experience – just as TV is not radio with pictures, online video is not TV on a smaller screen. Research conducted by Toluna on behalf of Specific Media amongst a group of over 2,000 consumers across the UK, Germany and France highlights the differences between online video and its older, more established sibling.Online video is a ‘lean forward’ medium, and consumers are far more likely to consume it actively compared to the more passive consumption of television, which is habitual and ‘lean back’. Over two thirds of respondents in the study said that they consume online video actively compared to two thirds who view television passively. This is largely driven by the more active triggers for online video – consumers choose content that they want to watch at a time and place convenient to them, on a screen or device that suits them. Consumers are embracing the array of utilities unique to digital versions of content – one in four research respondents has commented on content, shared content via social media and forwarded content to others.

Video works

As online video heads towards (if indeed it hasn’t already) becoming a mainstream communication channel for marketers, there is an increasing need to answer the many questions that exist about what works and how, its position in the digital eco system and how marketers can best maximise their investment in the channel.

Will new creative or repurposed TV content generate the best impact?

How do people respond to original branded content?

Does the length of content have any bearing on advertising effectiveness?

What is the perception of advertising formats – how well do they fit with the content?

Introducing VITAMIN

To help marketers and the industry at large answer some of these questions, we have built a new insight engine, called Specific Media VITAMIN – Video Testing and Measurement Insights. VITAMIN is a bespoke tool that allows us to measure any number of content and creative combinations amongst key consumer segments in a range of simulated online environments, measuring both advertising effectiveness metrics and format evaluation metrics.

The VITAMIN approach

First use of VITAMIN took place in November 2011, measuring 12 different content/advertising cells amongst over 2,400 respondents. Principally, it tested advertising effectiveness for 30 second advertising spots served within short form environments (professional quality content, c3 minutes long), long form environments (c10+ minutes long), user-generated content and original brand programming.

Consumers were profiled, with demographic information collected, on and offline viewing habits measured and their ownership and propensity to purchase different products and services collected. They were also asked to select their favourite content genres, e.g. entertainment, sport, news etc. and to select a clip they would like to watch.

Prior to the clip being shown their anticipation is measured using an Anticipation Index scale, and post-viewing their enjoyment level measured using an Enjoyment Index scale. Respondents then view the clip within a pre-determined environment and predetermined sequence of advertisement(s). Finally, they are asked a series of questions relating to the brands advertised and the advertising formats used. A follow up set of questions are asked two weeks later to measure latent advertising effect. An unexposed control group is also measured.

What has VITAMIN taught us?

Video length has no bearing on advertising effectiveness

This example shows the percentage of people who have a favourable opinion of a brand they have seen advertised in VITAMIN, compared to a control group. Advertising prior to both short and long form increased brand favourability (video works) but long form content is not more effective than short form (in fact short form performs marginally better in this instance). (57% of short-form viewers, 55% of long-form viewers and 48% of the control group said they had a positive opinion.)

Original brand content was the most effective advertising format measured

This illustration shows the percentage of people recalling, unaided, a brand they have been exposed to advertising from, split out by the content format the advertising was served around. Again, there is little difference between short and long form content, but the benefits of using original brand programming can clearly be seen – with unaided brand recall twice that of all other formats tested. (25% recalled the brand through original branded content, 14% through short form, 13% through long form, 11% through user generated content.)

Enjoyment and Anticipation have a direct impact on advertising effectiveness

Using the VITAMIN tool we can also examine advertising effectiveness data within the context of format evaluation data. For example, respondents are asked whether they enjoyed the content they had just watched. When they do, this has a direct impact on brand metrics, in this instance brand favourability. When they haven’t enjoyed the content there is a negative halo on advertising effectiveness.

VITAMIN vs. the click

More and more advertisers and brands are embracing the opportunity that online video presents; namely to connect and engage with consumers in a one to one environment, combining the data led targeting of online with the audio visual power of television. So should we measure the success of this exciting new medium in the same way we do with digital display – namely the click?

Whilst this is the default option – it is not a particularly insightful one. A highly engaged consumer is hardly likely to click away from content they have chosen to watch, just as they are unlikely to change direction if walking past a relevant poster. The click through rate for display has been in steady decline for some time, a reflection not only of the novelty factor wearing off, but that the internet has evolved from a task orientated medium to an entertainment destination where the consumer is empowered.

The early digital philosophy was that the internet allows consumer to act instantly, therefore we should measure that instant action  through clicks. Unfortunately this has become the key measure of success or failure, despite a vast body of evidence existing that illustrates the click’s flaws.

Only 16% of the population account for all clicks. Further heavy clickers are even more prominent – some 8% of clickers account for 85% of all clicks, according to comScore.

CTR also runs the risk of damaging value for brands if they optimise their campaigns to perform on CTR. The vast majority of people do not click on ads.

Furthermore, users who have clicked in the past are twice as likely to click again in the future and nearly 20% of clicks come from ad impressions that were clicked on more than once, suggesting that many of these clicks were not intended.

When looking at who actually clicks a picture of an audience is painted that is unlikely to be attractive to most advertisers. Clickers tend to be lower income, older and late adopters of new technology.

For example, users with incomes under £25K click 30% more often than users with incomes over £130k, and those with only ‘fair’ credit scores click 20% more often than users with ‘excellent’ credit scores. And users who are late adopters of new technologies click 50% more often than early adopters¹.

Many clicked ads don’t materialise in terms of action or brand lift – there is no discernable correlation between CTR and post-impression action. Furthermore, the highest performing CTR campaigns examined (top 20%) have a 150% higher CTR but an 8% lower postimpression action rate².

Similarly, a preliminary study of 100 campaigns shows no correlation between CTR and brand lift and purchase intent as measured by post-impression surveys. Therefore, optimisation of campaigns to achieve higher CTR may in fact be reducing brand ROI.Measuring CTRs against Experian MOSAIC profiles sees a very similar pattern, with those in less desirable MOSAIC categories generating the highest CTR³.Finally, to compound this, brand marketers are not entirely interested in the click as a measure of success.

A Bain survey of brand marketers highlights the measures they want, compared to the measures they get. In summary, the click is lazy and irrelevant as a measure.

Rather than focus on CTR, advertisers should instead concentrate on targeting a relevant audience and communicating with them in a premium advertising environment. In terms of measuring and evaluating success of advertising, the full suite of available tools should be utilised – for example VITAMIN, brand studies and incremental reach.

The measurement ecosystem

In addition to using VITAMIN a range of measurement tools and techniques exist to understand the effect of the medium.

Bespoke brand studies examine in detail the effects of a live campaign in real time. We measured a recent campaign for Barclays to assess the impact display, video and TV pre-roll advertising had on advertising awareness amongst the core target audience, and compared this to an unexposed control. In this instance, video nearly doubled advertising awareness.

Incremental reach studies measure, analyse and model the additional reach (and efficiency of) video running with a TV campaign. The below example from Nielsen highlights the role video can play. Combining an original TV campaign (c£600,000 spend) with an online campaign (£60,000 spend) results in the overall campaign reaching 60.31% of the UK 15+ population or 30.01 million people. Therefore, the online campaign results in an incremental reach of 1.95 million UK people (a 6.95% increase in reach) on the original TV campaign.

The TV+Online campaign increases the average number of exposures from 2.41 “TV-only” to 2.49. As this is lower than the “TV+TV” campaign average (2.60) but also achieves 150 GRPs, it suggests new people being exposed to the combined campaign at lower, arguably more efficient, frequencies.

Analysis of network data also sheds light on the performance of video, and how it combines with display advertising. In a recent campaign for a major clothing retailer running a display campaign, video advertising was used for a two week period. During this period the CPA for digital activity fell 46% – video advertising had a direct impact on digital performance, irrespective of user interaction with the video advertising itself.

Concluding thoughts on video

1. Consumption of video is growing at a phenomenal rate and changing viewing habits amongst consumers

2. Video is its own unique medium, not television on a smaller screen

3. It represents a ‘silver bullet’ opportunity for advertisers – the emotional engagement of audio visual, combined with the data-led targeting of online

4. Measuring the right thing is critical to understanding and evaluating success, the click is a misleading tool5. Video advertising works, regardless of content length – Short Form is as effective as TV Catch Up

6. Original branded content is incredibly effective as an advertising tool

7. Thirty seconds is just a random number online – testing and experimenting are key to success



Please enter your comment!
Please enter your name here