Lumen's Mike Follett on the Evolution of the Attention Challenge and the Future of Advertising Metrics

Ahead of the ExchangeWire LIVE Attention event, CEO at Lumen, Mike Follett, sheds light on the evolving landscape of attention, the significance of data-driven advertising metrics, compliance with privacy legislation, and the innovative possibilities that clients are uncovering.

How has the attention challenge changed recently, and why is capturing people's attention now harder?

Deep down, people’s attention patterns haven’t changed over time: we’re just as good at ignoring ads today as we were in years gone by. But the media we consume has changed, and so the volume of attention given to ads in aggregate has changed too. The biggest shift has been away from ad supported linear TV and towards subscription CTV. Time was when people would spend three hours in front of ad supported TV shows a night - which means roughly 20 mins of ads an evening. Even if you only watched half of those - which is what TVision’s data suggests - that is still a good 10 mins of pure commercial engagement per evening.

Nowadays, people watch just as much TV as they ever did, but much of it is on non-ad supported platforms like Amazon and Netflix or BBC iPlayer. People have literally paid not to watch the ads. This means that the largest commercial window into people’s lives has closed - and until the streamers start making their ad-supported offering more attractive for advertisers and audiences, then it will remain shut.

Ads on other platforms and media do get noticed and do work hard, but they are not generating the really significant slices of attention required to tell stories and build deep memory structures.They are more like posters or billboards. In fact, they are brilliant, engaging, perfectly targeted billboards. But they’re still billboards rather than TV ads.

Are some ad formats better able to capture consumers’ attention? In what ways do other formats struggle?

There is a clear hierarchy of media attention, with cinema sitting at the top of the tree. The conditions for attending to the ads are literally perfect. You get locked in a big black box with a massive screen and amazing surround systems. You are ordered to put your phone away, sit back and relax. The advertisers rise to the occasion and usually put something worth watching on the screen in front of you. It’s astonishing, but it is also a rare and unusual experience for most of us.

Compared to cinema, every other media is a little bit broken - there are distractions, or the sound is off, or the scroll speed is too fast, or we’re concentrating on the road ahead rather than the posters on the sidewalk. But while these other media are a bit cracked, they are plentiful, ubiquitous, and repeated.
The trick is to stop comparing other media to cinema, and embrace the fleeting and transitory nature of many other media. Designing your ads for these busy and bustling environments can make them stand out from the crowd. Understanding what they can realistically achieve, especially through repetition, is essential for success. Instead of being disappointed that other media aren’t quite cinema, I think we should celebrate their unique strengths.

Which metrics gauge ad campaign attention and their effectiveness?

If you are interested in human attention, then you have to start by actually measuring human attention. Companies like Lumen, Amplified Intelligence, Playground and TVision go to the trouble of actually collecting ground truth data on how people actually look at ads. All of us use some form of eye tracking or head tracking technology. This has its limitations (we have ears as well as eyes, after all), but I think that this is a good start. Once you have the ground truth data, you can start building predictive models that estimate attention given an ad’s viewability characteristics. If you can tell me how much of an ad was in view and for how long, the model will predict how likely it was that the ad was seen, and how long it was looked at for.

These impression-level attention predictions are then linked to impression-level outcomes: clicks and conversions, or brand lift questionnaires or sales lift studies. This helps us understand the relationship between attention and outcomes - and how it differs from media- to-media and campaign-to-campaign. It’s this insight that allows us to build ad tech tools to help brands buy the right level of attention, and get the job done right.

The system is not perfect and, like good scientists, we’re all still learning as we go along. But it’s evidence-based, and the models are based on actual data rather than a bunch of assumptions. There’s no black box, just the facts.

What's being done to enhance attention measurement while adhering to stricter privacy laws?

Everything we do at Lumen is GDPR compliant: all the ground truth data is fully consented, the models are anonymised and we don't collect any cookies on live campaigns. In a sense, the ‘tightening privacy legislation’ that you mention is the reason that we are doing so well. The end of surveillance capitalism and the death of the cookie mean that contextual targeting solutions like Lumen are set to inherit the earth. Margrethe Vesthanger, the EU ICO, and inventor of GDPR, is our patron saint.

What are you looking forward to at the ExchangeWireLIVE Attention event?

I’m really interested in hearing about all the cool new uses that our clients are inventing for our data. Almost all the good ideas we’ve had at Lumen have actually come from our clients - people like Katie Hartley at dentsu, Jon Waite at Havas or Erfan Djazmi at MediaHub. We just keep our ears open (or perhaps it should be eyes open?) and try to keep abreast of their creativity. That’s what I am hoping for from ExchangeWire LIVE - new and interesting perspectives on attention and outcomes.