How Viewers Engage With Television

It Ain't What You View, It's The Way That You View It

Engagement is the new black. As with all new fashions, it's been the source of much hand wringing, eyebrow furrowing and brain storming. We know there is more to television viewing than just exposure, but that is primarily what we measure. Research often asks people about that exposure after the event and out of context but pretty much all we know about what goes on at the point where viewers are being exposed to commercials is what we learn from the ratings; that they were present in the room whilst the set was on. Following Thinkbox's pioneering study into what goes on inside the 'black box' of television viewing, we are happy to tell you that viewers can, and do engage with television commercials in a clearly observable way.

Our study consisted of an innovative mix of video ethnography, lab-based memory study and an in-depth quantitative survey of 3,000 people. It cost nearly $1 million and it took almost a year to complete. It had to. Any study of engagement has to consider our rapidly changing understanding of how television advertising works on the viewer. The challenge is that the theories of low attention processing, the role of emotion in advertising and implicit memory do not sit well with a media research tradition which asks direct questions and takes rational measures of campaign response. Thinkbox's stakeholders decided a fresh approach was required and the Engagement Study was born.

The Methodology

Our research partners (ACB) installed video cameras within the TV sets of 22 households, totalling 74 individuals. They represented a range of demographics, household types and technographics. The only way in which our sample differed significantly from the population at large was that half of them owned a PVR, which meant a higher than average number of them also had digital multi-channel TV services such as Sky, Virgin or Freeview. This was to 'future-proof' the research against emerging trends in programme access as well as allowing us to thoroughly test the probability that PVR ownership would dramatically impact on live TV spot audiences.

We monitored their viewing for 6 weeks and took samples of viewing (to commercial breaks only) after 2 weeks, once their viewing had settled. Altogether, we had 4 hours of viewing to commercial breaks for each household, giving us a data set of almost 16,000 commercial exposures. Each of these exposures was coded for behaviours as well as commercial content and viewing context factors - a total of 170 different codes for each commercial exposure.

This was followed by a lab-based memory study, where we tested the impact of 6 test ads in different attention contexts - full attention, no attention and two partial attention groups. We attempted to measure implicit as well as explicit memory of ads, as well as test the approach for the larger scale quantitative survey.

Finally, we conducted a full quantitative survey amongst 3,000 respondents, using an online broadband sample (pre-recruited). This allowed us to measure viewing patterns and contexts, as well as test the impact of a wider range (20) of different ad types to test some of our theories from the earlier stages of the study.

The Results

TV Still Plays a Central Role in Most People's Lives

We made things as hard for ourselves as we possibly could. We didn't ask respondents anything about their TV viewing or attitude to ads before recruiting them, so we could have had 22 TV-hating ad avoiders for all we knew. On top of that, fieldwork was conducted during the fiercest heat-wave on record, right after yet another disappointing England World Cup campaign (again!), slap bang in the middle of the summer programme schedules - explaining why much of the footage shows people in varying stages of undress!

And yet TV was a significant in-home pastime for every one of those households. There were 3 distinct types of TV viewing.

Our time: shared viewing; social occasion; early peak and weekends

  • In between time: often when doing other activities; flit between sets; weekday daytimes
  • My time: 'treat' viewing; often alone; mainly women; lunchtimes and late peak

Exit interviews confirmed the importance of new TV technology in their lives ("Sky is the best thing that ever happened to us") and the wide range of roles that TV plays in their lives (entertainment, information, relaxation, mood lifter etc.), all of which were replicated in responses to the quantitative survey.

Engagement Can Be Identified Through a Wide Range of Behaviours

We admit to being nervous when we first took on this project. We worried at the prospect of lots of footage of people sitting stock still, hopefully looking at the screen, but otherwise not doing much. How naive I was! Across those 15,000 exposures, around two thirds (68 per cent) recorded some observable ad-related behaviour, according to ACB's team of professional ethnographers. What's more, positive behaviours outnumbered negative by almost two to one (43 per cent compared to 25 per cent).

The range of behaviours was fascinating. Music is important - a cue for viewers to sing along, whistle, clap and even dance, which they did on many occasions. People will stop the PVR and play back an ad at normal speed. They laugh out loud, openly express wonder at the creative execution, or refer the creative to things that are important in their daily lives (which we call self-referencing). They mimic the voice-overs affectionately and regularly play 'guess the ad' games. They can also tut, shout at the telly, roll their eyes and disparage the ad to each other. Just not nearly so often.

Engagement Is Not The Same As Attention

Although engagement is more likely to occur when attention is more focussed on the TV, there were numerous examples of high engagement springing from a very low attention base. People would interrupt deep conversation for a shared favourite and mimic an ad even when they had their backs to the screen. We recorded many examples of strong engagement occurring even when people were in a different room.

The main driver of attention is concurrent behaviours, so it is not surprising that engagement is higher when there are few distracting activities. However, we need to look closer at the types of activity being conducted; there are many common activities that are 'heads-in TV' activities - where the activity is created around the TV set (eating and drinking, exercise, some forms of housework) where engagement is often higher than it is when viewing is their sole activity.

The Power of the Shared Experience

In this era of the importance of word of mouth and the power of the viral message, it is reassuring to know that TV offers the most viral medium of them all - where ads are often discussed and shared with each other in the same physical space. It is something unique to television, and part of the power of the medium to get advertisers' messages out there.

We saw lots of examples of people discussing the ads, the brands, previous experience of the products and the people in the ads. They love to share the humour, often taking 'ownership' of favourite ads and pointing them out whenever they watch with family or friends. It is a powerful characteristic of the medium.

Perhaps that is why the optimum number of people in a room to create engagement is two. It gives people the opportunity to share their experiences of the ads, but not so many it can be distracting.

The good news is that the majority of TV viewing (around 70 per cent) in the UK is shared viewing, and it appears to be growing as more viewing comes back to the main set (where most of the new TV technology sits).

The Importance of Emotion in Advertising

When we classified the ads by their primary content factors (affective, cognitive or sensory) it was the affective (i.e. emotional) ads that performed by far the best in terms of eyes-on-screen or positive engagement behaviour. In particular, ads that elicited nostalgia, humour, excitement or personal identification (via self-referencing) seemed to do well. Interestingly, the second most positive response was seen for highly cognitive ads - the opposite end of the spectrum - suggesting ads should not try to mix styles. They should either be highly emotive OR information-packed, but not both. Worryingly, more than 20 per cent of ads could not be classified as strongly cognitive, affective OR sensory, which made us wonder why they had been made. Needless to say, these ads performed particularly poorly in our analysis.

When we tested the 20 ads in the quantitative stages of this study, we found that the strongly emotional or affective ads performed far better than the informational ones and even better against ads that attempted to mix emotion and information too much. These ads seemed to work more intensely at an implicit level, which is not really covered by many research methodologies. Emotional ads seemed to work particularly well against ad liking, which we found to be the most important metric of them all. All of this supports the recent IPA findings, following their meta-analysis of more than 800 IPA Ad Effectiveness Awards entries that showed campaigns that aim to strike an emotional connection with consumers performed much better than those that attempt to impart information about the brand.

How To Create An Engaging Ad

When we looked at the secondary content factors, the standout factor was music. It is an overlooked part of the power of television. It was highly successful in switching attention towards the TV - an important consideration in these days of multi-tasking - and in engaging viewers who sing along, dance, clap and whistle to the music. It was interesting to me that the most popular music for many of the kids in our sample was 'Hey Good Looking' from the ASDA ads - a country & western tune written in the 1950s, whilst one of the favourites for the older adults was Groove Armada's 'I See You Baby' from the Renault Megane commercials. The world is truly turning upside down!

When we compared the performance of different ad types across all three stages of this study, it was encouraging to see that the M&S campaign, which won the IPA Ad Effectiveness Award in 2006, was the standout across all 3 stages of this study. Even though the ads were quite traditional creatively, the tones of voice, inspiring product demonstrations, clever use of celebrity and synergy with established perceptions of the brand were all appreciated by our respondents.

Ad Avoidance and The PVR

One of the topics we looked at with special interest was the whole issue of ad-avoidance, not least because half our sample have the technology to avoid ads altogether, via the PVR.

In fact, one overwhelming conclusion from this study is that the PVR is used to enhance programme viewing, not to avoid commercials. As with a range of studies conducted last year, we found that PVR owners actually watch more commercials at normal speed as a result of owning the PVR. This is because only 12 per cent of their viewing was to time-shifted material (much less for younger respondents) and around 40 per cent of the commercial breaks were watched without fast forwarding. At the same time, owning a PVR appears to increase broadcast TV viewing by around 15 per cent, mainly to commercial channels, resulting in an increase in the number of commercials viewed. It is counter intuitive, but it is true, and has been supported by studies from BARB, Sky and the London Business School. The PVR is our friend!

'Zipping', or fast forwarding, does occur, but even the households who claim to zip most of the time only zipped ads for a small proportion of their TV viewing. Whether it is because they are aware of the short length of some breaks, they just forget they are viewing in time-shifted mode, or because they want to see the ads (quite often people fast forwarding would rewind to watch an ad that caught their eye) there seems to be no massive desire to edit out the ads.

Decades before the PVR was a glint in the eye of the electronics companies, people would often claim to leave the room when the ad break started, to make a cup of tea or attend to some vital chore. In our study, several respondents claimed to always get up and do something else as soon as the programme credits rolled. Again, the reality does not match their claims. Even the most adamant only left the room two or three times during the four hours of their that breaks we recorded and, in total, less than 5 per cent of breaks suffered.

Mood Congruence

Viewers watch TV for a variety of (mainly emotional) reasons.

It performs a major role as a mood moderator and we had many examples of respondents responding negatively to an ad because it did not match their mood at the time of viewing. Consequently, in contrast to findings from other studies, we found ad engagement was enhanced if the tone of the ad reflected the tone of the surrounding programming.

There was a definite relationship between type of programme viewed and positive engagement. There appeared to be three overall types of congruence - humour (with comedy-based programming); drama and excitation (with drama, reality, news and sports programming); and relaxation/escapism (with gentler dramas, soaps, travel shows, talk shows etc.)

Although respondents seemed to 'switch' mindsets temporarily going into commercial breaks, they also appeared to move back into the mindset as the break progressed. They openly objected to ads that clearly disrupted that mindset and appeared to engage best with ads around the relevant genres. As the report stated;

"Ad mood congruency is important in order to match the mood of the ad with that of the viewer via the attaching programme. For example, there is the potential for positive misattribution of positive mood, in a state of excitement under the influence of increased adrenaline, such as may occur whilst watching sporting events. Equally, viewers not desiring excitation, who may be relaxed in a state of homeostasis, may react with increased negativity to having their mood affected by a misplaced ad."

Break Ecology

What did we learn about the ecology of commercial breaks, one of the toughest things to measure via conventional research techniques? With this study, we were able to observe what happens within a wide range of breaks as well as analyse the differences between them.

As we might have expected, centre breaks do seem to elicit more attention and positively engaging behaviours than end breaks. The audience is obviously more settled and more likely to be attentive to the screen. They are also less likely to engage in any of the ad avoidance behaviours mentioned above.

Less expected was the finding that attention and engagement rise continuously during the break. First in break has always been thought of as a premium spot but this study would suggest it should be later in the break that deserves the premium. This appears to be because many viewers have a short 'wind down' period at the end of each viewing 'session' and then begin to build up their 'heads-in-TV' mindset for the beginning of the next segment.

The Metrics Don't Work

Finally, a few comments on the impact of TV ads and how we measure their effects. When we tested the 20 ads during the quantitative phase, we measured a number of things to do with response to the ad itself as well as the brand being advertised. We then ran regression analyses to see what drove the brand performance the most. Correlations between 'relevance', 'creativity' and 'ad liking' were extremely strong, which suggests consumers will like ads that are seen to be relevant to them but, more importantly, they think are creative. They are more likely to recall an ad if they have used the brand before, but other than that, there is little relationship between the brand performance and awareness, recall, recognition or correct brand attribution. Interestingly, 'ad liking' did correlate well with both brand favorability and future purchase intention. The more they liked the ad, the more likely they were to claim to buy the product, even if they did not remember which product the ad represented. This is potentially one of the key findings of this study; how we evaluate TV campaigns appears to be pretty much unrelated to their impact on the brand. Indeed, our current metrics appear to dilute creativity in advertising, and yet creativity is seen as the cornerstone of a campaign's impact on the consumer.

Conclusions

This study is the sort of thing television needs to do to demonstrate just how our viewers engage with the medium and just what a great advertising opportunity it provides.

The results outlined above are all significant and relevant to today's business. They question the ways in which we currently evaluate TV advertising and they offer opportunities for creative and media planning disciplines to work together to maximize engagement via the best combinations of content, audience and context coming together at the right times. Most importantly, they show us exactly what goes on behind the 'black box' of television viewing and how the ways people interact with their TV commercials can amplify the message and 'hard wire' it into their brains in ways that no other advertising media can match.

The frustrating aspect is that there are still so many insights we never have time to share. For example, that PVR owners appear to be able to recognize commercial messages based on much less information than their non-PVR owning equivalents. Or the finding that respondents in the memory study appeared to be better able to recall ads for food and drink products when they were eating and drinking in front of the TV. Or that the advertising averse segment of our audience appears to show the greatest uplift in purchase intent for ads that they allow through their 'cognitive guard'. The list is virtually endless.

Meanwhile, we have been presenting the results to the advertising industry here in the UK and the response has been both positive and reflective. Positive because it has shown through rigorous research in a time-sensitive and naturalistic setting just why TV has been successfully creating and building over the decades; reflective because the findings cover so much ground, they require a complete rethink of how the medium is used. To this end we have created an 'Engagement Community' amongst media and creative planners in the UK and they are already starting to provide much-needed input into how we can apply all these findings to build a useable engagement planning tool. That is what will keep us up at night for the next twelve months at least. The destination is in sight, but we'll always remember the journey to get there.



Engagement Study

Thinkbox sought to identify what realty goes on inside the 'black box' of television viewing. How do people engage with TV advertising? What tunes them in, turns them on and ultimately, how does engagement impact upon brand takeouts? This groundbreaking study encompassed real people watching real TV in their own homes and provides fascinating insight into how TV advertising actually works.