Lumen is on a mission to move the ad industry from ‘OTS’ to ‘S’. Currently, we use viewability data to measure the ‘opportunity to see’ (OTS) of advertising. But OTS has little to say about how many ads actually get looked at. Humans are very good at ignoring things – even things that are staring them in the face. We have to be: we don’t have the cognitive resources to look at everything that has the ‘opportunity to be seen’. Ads are just a special case of things that we are particularly good at ignoring. Lumen’s eye tracking research has shown that, at the extreme, only 18% of viewable desktop display ads actually get looked at.

And attention matters to brands, because as numerous studies from ourselves and the other founding members of The Attention Council have shown, attention is the active ingredient in advertising success. Dentsu’s Attention Economy study (which combines TV attention data from TVision with Lumen’s eye tracking data for mobile and desktop) showed that attention was 700% more powerful than mere viewability in predicting changes in brand choice. Our work with Mediacom and British Gas has shown similarly impressive links between attention and online sales.The relationships are not simple and not linear, but they are consistent. When it comes to attention, more is more.

Let’s start taking the ‘S’ in ‘OTS’ more seriously.

Lumen has been at the forefront of the Attention Economy debate for over 8 years. Founded in 2013, we have developed a technology that turns the webcams on panelists phones and desktop computers into high quality eye tracking cameras. This allows us to measure what people could see, and what they do, in fact end up looking at. The technology has been investigated and approved for use by researchers at Oxford University and Imperial College London, as well as being thoroughly audited by Dentsu as part of their pitch process.

We deploy our technology amongst nationally representative panels of consumers: 1000 respondents in the UK and 1000 in the US give us attention data every day, making our dataset orders of magnitude bigger than anything else in the market. We collect data on attention to ads across social media, open web and BVOD services: if you can look at it on your phone, we can measure it. We also have the ability to set up shorter term panels or even conduct ad hoc creative tests, and have recently completed studies in Canada, Mexico, Brazil, Norway, Italy, Indonesia and Australia. All our data collection is fullyinformed and consented and we adhere to the highest GDPR and MRS standards.

But what do we do with this data?

The first thing to investigate is the drivers of attention. Can we build a predictive model of what causes people to look at ads? Some of these factors are pretty easy to quantify. We’ve been able to isolate the impact of media factors such as time in view and size of ad on the likelihood people have to look at ads.

Others are more complex: the context within which an ad is seen can have a big impact on how much attention it will generate. Uniting Lumen’s predictive models with contextual targeting data can reveal powerful new insights for media planning.

But other factors are almost impossible to build a universal predictive model for. Creative is massively important – and enormously variable. A simple, elegant and well branded ad can have a much bigger effect – in a much shorter time – than a bigger, busier or more boring ad that placed in a better position.

Targeting also makes a huge difference. It is hard to predict what people will look at unless you know what they are looking for. The intentions and aims of the readers of websites have a big impact on the attention they give to the accompanying ads. Similarly, if you are not in-market for a particular brand’s product or service, you have every incentive to ignore their ads. But as soon as you are in-market, the same ads can be positively useful.

Understanding this complexity is important. It suggests that attention isn’t just ‘selective’ – choosing between different options in a visual field – but, in a sense, ‘constructive’. We look at things because they have meaning and significance to us right now. But by understanding that complexity, we can do something about it.

From insight to action

At Lumen, we created a series of tools that allow brands and publishers to do something about attention. We can help with planning, reporting buying and optimising attention.

Some agencies and auditors use our data within their planning and valuation tools. You can apply our ‘attention discount factors’ onto your reach and frequency estimates to understand the actual attention your campaign will generate – and how much it will cost to ‘buy attention’ in one media or on one platform versus the rest.

Other partners use our LAMP tag, a simple piece of javascript that can be appended to each impression served, that records the viewability characteristics of an ad and then estimates how much attention it is likely to have generated. This attention data can then be matched to CPM data to give clients live insight into the ‘cost of attention’; clicks and conversions data to help understand the ‘performance value of attention’; and brand uplift study data, via partners such as On Device Research, to give us insight into the ‘brand value of attention’. Crucially, we can ingest viewability data from the Walled Gardens and give clients consistent attention data for ads across Facebook, Instagram, YouTube, TikTok, Snap and the open web.

Reporting is one thing, but many clients want to act on attention in realtime. So we have integrated our predictive model into DSPs to allow brands to direct investment towards inventory that is most likely to get looked at, and away from that that, though viewable, are unlikely to be viewed. Our integration with Google’s DV360 system is already creating enormous value for clients in the US, UK and across Europe: we hope to be able to share some case studies in the coming months.

Finally, as mentioned above, the most sophisticated clients are well aware that creative and targeting matters. So we have tools to test and optimise creative and use this data to create brand-specific attention models that can be used by the LAMP tag or the DSP plug ins. One brand may have creative that is so efficient that they can get away with buying cheaper media; another brand may have an ad that works powerfully, but only if it gets 5 seconds of attention. Knowing this allows us to tweak both the buying strategies and the reporting models. This understanding has allowed Dentsu to create it’s proprietary ‘Effective Attention’ model, which takes the basic LAMP system a step forward for all their clients.

Don’t wait. Act.

All this is happening right now. The most advanced publishers, format owners, agencies and brands are all already using these tools to reduce wastage and drive value. As one of the original ‘attentioneers’, it is gratifying to see that theory is turning into practice.

But practice is what we need, as that is what is going to make perfect. There is still so much to learn and so much value to create. But the best education is ‘learning by doing’. Your brands will have unique challenges and unique opportunities. You will need to understand the ‘shape of attention’ that is right for you. We can start small – deploying the LAMP tag to understand the lay of the land, or running some creative tests to understand the performance of your ads – but we have to start somewhere. This is just the beginning of the journey.

Also Published In: MarTech Series

Powered By: Audio Harvest