Attentive seconds: One currency to rule them all?

Why can’t we measure all media the same way? We live in an alphabet soup of different measurement systems: GRPs, OTS, LTS, ROTS, viewable (and non-viewable) impressions. They are all sort of talking about the same thing – did anyone see my ad – but then again, they are all slightly different and incommensurate. Close, but no cigar.

This confusion is obviously a thorn in the side of advertisers, who find it hard to judge exactly how much a TV ad is worth compared to, say, a digital ad, and feel that they are probably getting ripped off by publishers as a result. But it doesn’t benefit the publishers much either. If buyers don’t trust the market to reliably set prices, then they can’t be sure that they are getting a good deal when they are offered one. This distrust lowers the price for all the sellers: the good ones and the bad ones. You don’t have to be a Nobel Laureate to understand that it would be helpful to have a reliable common currency of attention to avoid this ‘market for lemons’.

And this is where Lumen can come in to help. Attention can be the common currency across all media (well, everything except radio. Sorry, radio).

Ads may come in many shapes and sizes, but visual attention is pretty much the same all the time. You’re either looking at something, or you’re not. And looking at ads is important. Visual attention isn’t the be all and end all, but you are much more likely to remember ads that you look at, and similarly, much more likely to buy from ads that you look at. Which shouldn’t really need to be said, but that’s the world we live in now.

Lumen measures visual attention across media, applying the same standards of viewability and viewing to everything. And this consistency of approach allows us to rate formats and media against one another, using the currency of ‘attentive seconds’.

At last week’s JCDecaux upfonts, we explained our approach via a worked example. Imagine you have bought 1000 desktop MPU impressions. How much actual attention do you really get? Well, the first thing to do understand how many of those 1000 ads could have been seen. A 57% viewability rate means that only 570 of your 1000 ads were available to be seen at all.

But Lumen can go further. It turns out that only 8% of your MPUs are likely to have been looked at. Only 80 ads of the 1000 you bought are getting any attention. And, on average these ads are likely to have got 1.3 seconds of attention each. With this information, you can create a composite metric that takes into account how likely people are to look at your ads, and how long they look: 1000 ads x 8% chance of being seen x 1.3 seconds of attention = 110 ‘attentive seconds’.

Now, the beauty of this approach is that it can be applied consistently across media. We can apply it to press advertising, and mobile ads, and OOH and even TV. Every media has a ‘viewability rate’: how many people actually open a page of a newspaper, or have a clear view of a poster as they are walking down a high street, or are in the room when a TV ad is playing. It’s the same principle for every media.

And every media has a ‘likelihood to see’ rate: how likely you are to actually look at an ad, if it is viewable. Sure, you were reading the article – but did you actually look at the ad? Sure, you could have seen the poster – but did you? Sure, you were in the room when the TV ad was playing – but were you looking at your phone instead?

And finally, every media has a ‘view time’ rate. So, you looked at the ad, but for how long? A second or a minute?

And when you collect all this data, as we have (with a little help on the TV side from our friends at TVision), you can come up with a common view of the actual attention that different media formats actually deliver. We took some representative formats from each media and compared the ‘attentive seconds’ that each produce.

Perhaps unsurprisingly, TV ads do best of all. According to TVision, about 1/3 of TV ads play out to empty rooms: the TV equivalent of non-viewable inventory. Of the rest, about half the time people are doing something else when the TV ad is playing: the equivalent of an ad being viewable by not viewed. But, TV ads are long, and if people view them, they view them for much longer than they do for press ads or posters or mobile pop ups. So, even if the likelihood to look at a TV ad isn’t actually that much greater than other media, the viewing time is substantially longer. As a result, 1000 TV ad impressions will get you about 6000 attentive seconds.

Applying the same approach to press ads means that you get about 1400 seconds of attention per 1000 quarter page ads; 838 seconds for typical digital 6$s; about the same for mobile MPUs; and just 110 seconds for 1000 desktop MPUs.

And now you can see why advertisers pay more for TV ads: they get a lot more bang for their buck. But looking at this data, are they paying too much to TV? Sure, it’s 4 times better than press – but only 4 times. If you’re paying 10 time more on a CPM basis, then it’s not a good deal. And mobile and DOOH: are you paying the same rates for these media? From what I hear about price differentials, DOOH seems like a bargain according to these numbers. But perhaps you have your own pricing information and can come to your own conclusions.

‘Attentive seconds’ are important because they allow us to do these comparisons, and make smarter, joined-up decisions about how to invest our media money. We’d love to help you do just that.