Mike Follett, Managing Director, Lumen, responds to Facebook’s Ian Edwards and Harry Davison and argues that the smartest media planners are talking about ‘attention strategies’, utilising different ’shapes of attention’ provided by different media to achieve different aims.

Ian Edwards and Harry Davison of Facebook wrote an extremely interesting article, challenging some of the implications of Lumen’s research into the reality of attention to advertising. I don’t agree with everything they said, but I don’t want to spend time rebutting their arguments. The only thing more boring than listening to a report about someone else’s argument is listening to a report about someone else’s dream. Instead, I want to pick up on two really important insights highlighted by Ian and Harry, which I think have major implications for the advertising industry.

The two most important ideas that Ian and Harry put forward are these:

  • Advertising should be judged by its outcomes, not its inputs
  • Different ad campaigns have different objectives, and using a single metric such as visual attention flattens these differences – a bit like entering a F1 car into a tractor pulling competition

Both these points are profoundly important, and have significant implications for where attention measurement needs to go next. 

Firstly, the value of advertising should be judged by its consequences. As the great Jeremy Bullmore once said, 'Art is; advertising does’ – and what advertising does is sell things. Attention is the active ingredient that helps advertising do its job, but it is the means, not the end.

The problem is that a sale is the result of a witch’s brew of interconnecting factors: the brand and the offer, the message and the creative execution, the media space and the targeting. Media owners – even media owners as sophisticated as Facebook – are only in charge of part of this whole.

Using average attention metrics such as the ones proposed by Lumen can help isolate the contribution of individual factors. The point of the attention data is to help identify which ingredients in the advertising gumbo are contributing what. The data we presented averages out the effects of creative and targeting (let alone the impact of pre-existing mental structures or the specifics of particular offers). Further research needs to abstract the effect of media and then isolate the contribution of these extra factors. But while it is important to identify and isolate the impact of each of these ingredients, we would do well to remember that they are never experienced on their own, and it’s their combined effect that produces a sale.

Secondly, there are horses for courses. Sometimes you need to tell a story, sometimes you need to highlight a pertinent offer, other times you just need a reminder at the right time and in the right place. Or, as Ian and Harry have it, sometimes you need a F1 car, sometimes a tractor.

The different business objectives suggests the need for different ‘attention strategies’. More attention is not always required to get the job done, and paying for more attention is not always the smartest deployment of client resources. There are times when even a relatively short amount of attention can result in significant business impact: Sophie MacIntyre and Nicolas Arrivé of Facebook presented some really interesting findings at the asi conference in October 2020 about the relationship between exposure to ads on Facebook and subsequent sales uplifts, showing that the ads work even if the viewable time is quite short.

At Lumen, we know that around 80% Facebook ads that appear on the screen get some visual attention – even if not all of them reach the MRC’s rather arbitrary viewability thresholds (100% of the pixels for 2 seconds or more). Counter intuitively, there’s often more actual viewing than there is technical viewability on social media!

We also know that ads can be recognised and recalled in quite a short amount of time. This matches some of Facebook’s own research presented at the ARF Audience x Science conference in April 2019, and further bolstered by the work presented at the same conference by Claire Charron at Google. More attention is usually better when it comes to increasing recall, but all attention is worth something. Better known brands, utilising distinctive brand assets, can sometimes get a brand awareness job done in less than a second; less well known brands, or less well branded ads, need longer.

Given these realities, the smartest media planners are beginning to talk of ‘attention strategies’, utilising different ’shapes of attention’ provided by different media to achieve different aims: this is the premise of the Dentsu Attention Economy project, which we are working on at the moment, and should be reporting on in the summer. To use a slightly different (and slightly blasphemous) analogy, to everything there is a season, and a time for every purpose. There is time for sowing (perhaps TV?) and a time for reaping (perhaps Facebook?); a time for mourning (or at least being serious: perhaps quality journalism?) and a time for dancing (er, TikTok?). There are different time lengths, and different time uses, for different brand purposes. I think we can all say amen to that.

Both the points that Ian and Harry raise are, therefore, really important and have moved the debate about attention on considerably. For that I am really grateful. There is still so much to learn about the reality of attention to advertising, and how attention contributes to the real goal that we are all striving for: creating the most successful marketing possible for our clients.