Mike Follett's lengthy career in media planning now helps LUMEN perfect its media attention measurement. He chatted with Realeyes about the importance of combining creative and media attention, why A/B tests are the best way to optimize for attention, and how attention helps to create bigger, better memories.
Why is attention important for marketers and advertisers today?
Attention is the very essence of advertising. There's a clue in the word: advertising comes from 'advertere', which means to turn towards. The point of attention metrics is to give some measure to this essence of advertising, where an ad can draw the eye or ear and get noticed within a busy environment. So, if advertising is all about attention, we finally have a way of measuring the most important metric we've all been seeking to optimize for these years.
If advertising is all about attention, we've finally got a way of measuring the most important metric that we've all been seeking to optimize to for these years. |
There's a lot of talk about creative attention measurement and media attention measurement. Do you think they should be viewed together or should they be measured separately?
You can't have one without the other. At Lumen, we do lots of eye-tracking studies to try and understand the reality of the media, like how ads in general on Facebook gain attention versus ads in general on TikTok or on the New York Times. So, we can produce really good models of the media.
But there's always the possibility that we overestimate or underestimate [the effect] because we don't have the detail of the attention to the creative. So, the two things really do have to work together. At Lumen, we're really good at measuring the box on the screen as it goes up and down, but what's inside the box is important too. The actual creative is tremendously important. So, it's not one or the either. It has to be both.
At Lumen, we're really good at measuring the box on the screen as it goes up and down, but what's inside the box is important too. The actual creative is tremendously important. So, it's not one or the other – it has to be both. |
Where do you see attention going in the next couple of years?
Things are growing and leaps and bounds, not just commercially but also in our understanding and fact. By running lots and lots of commercial tests, we can learn more about the fundamental drivers of attention.
You know it's very much a sort of learning by doing approach. And so I see a really swift and precipitous increase in our understanding with will go absolutely hand in hand with a commercial application of our technology. |
At Lumen, we pay people in the US and UK to download our software and passively track everything they notice when they're on their phones or computers. It was used to power predictive models of attention. But the predictions are constantly being validated and demonstrated by application in the real world. So, we must carry on collecting data and modelling this, but we learn tons by applying it. It's very much a sort of learning-by-doing approach. And I see a swift and precipitous increase in our understanding that will go absolutely hand in hand with the commercial application of the technology.
For advertising leaders trying to dip their toes into attention, or learn more about it, What's the best advice for them?
Any journey starts with the first step, so you should try a small project first. Just dip your toe in the water and be conscious that you're trying it. But then also look for the outcomes of this stuff. There's a lot of attention data here and in predictive models; the crucial thing is to remember why we're doing this: to create bigger, better memories. We tell our clients to start with an A/B test and have an outcome like more clicks and conversions or increases in brand lift.
They should try this out as a commercial tool and see if it works on a small scale, and then increase their knowledge by doing bigger experiments. They will learn much more about attention [by doing] than if they're trying to understand the theory.
(This interview transcript has been edited for clarity and length.)