Commentary
Why Aren't We Buying On Attention?
- by Ishan Bhaumik , Columnist, April 3, 2018
Reach, frequency, delivery and ratings. For decades, the advertising industry has primarily traded against these time-honored metrics. But in essence, marketers use these metrics as proxies for something else more valuable — attention. In this era of big data, media fragmentation, ad avoidance and consumer messaging overload, it’s time to transact directly against attention, the metric that we all really care about.
Attention as a measurement metric has been gaining, well … real attention in media circles. Joe Marchese, president of advertising revenue at Fox Networks Group, speaking at a recent CIMM conference, advanced the idea, especially as it affects accurate attribution. “Attention is a currency. There is no attribution without attention,” he concluded.
Attention Measurement Challenges
If we are to transact against attention, then we need to accurately measure it first. And in this area, it has not been easy to find common ground.
In TV advertising, measuring attention has many criteria to consider. For years, Nielsen’s length-of-tune has been used as a proxy for the degree of attention paid to content. But the TV can be on, accruing length-of-tune, while we are otherwise un-engaged or absent. With the advent of smart TV data, the ability to determine actual television attention has become more precise. And more recently, eye-tracking technology, such as that from TVision Insights, has shown the ability to directly record TV viewership.
In digital advertising, publishers sometimes turn to metrics that indicate “time spent on page” via mouse or other keyboard movements. This makes the entire ad inventory supply chain vulnerable, as non-human fraud becomes all too easy to introduce. Recently, I wrote about several companies that are leading the way here while reaping the benefits of real-time attention.
Attention Measurement Solutions
The unique ways that newer companies are collecting certain types of real-time data are shedding light on how much attention is actually being paid to content and ads.
For example, in mobile, we have created a new way to measure attention: processing phone motion sensors to verify when an actual person is viewing the ad creative in real-time. This always-on phone sensor data can also indicate what activity is taking place and pinpoint motions that correlate to greater attention and ad delivery.
It’s time to stop transacting only against proxies for attention — such as reach — and get to the heart of the matter. The creation of a verifiable attention measurement metric is close at hand. With insightful uses of new and emerging datasets, marketers may find that attention is not only quantifiable, but can also become part of our established measurement currency. It’s time for digital advertising to “pay” attention.
Even better, attribute verified, viewable advertising to incremental business actions - which only humans can transact on - and you are, by default, optimizing to advertising attention.
Ishan, just out of curiosity, how do "phone motion sensors"determine that the user's eyes were fixed on the ad?
Yes Lee, but that is mostly applicable to online advertising. How does one measure the influence offline exposure had on online or offline transactions? Cinema audiences, for example, are paying more attention to an advertisement than at just about any other time in their life. There's no denying it's engagement but how is that captured at the transaction level?
Hi Ed, good question. Velocity has found that phone motion sensors (accelerometer and gyroscope) can predict if users are looking at their phone in a receptive state (and when they're not). For example: users handling their phone while standing in line, in the back of a car, or relaxing on a couch at home. When we find these moments of attention, we then serve an ad in real-time, and we see a drastic increase in measured ad enagement - everything from CTR, an install, or even a mobile purchase. Thus, we know that the user is paying attention to the ad.
Thanks for that reply, Ishan. However, I believe that a much tighter definition of attentiveness would be needed---especially across media platforms. You are probably right that the way a phone is handled may suggest a greater or lesser degree of potential attentiveness but one would need some form of ad recall---verified with message playback----to put a value on various handling configurations. Even so, there is no way that this kind of thing can be projected to other forms of digital media usage--- let alone to TV.
My personal opinion is that while it might be very desirable to have a valid method of measuring ad attention for every media placement, this is most probably a bridge too far, hence it is impractible. Such considerations---guided by whatever research is available plus common sense----should play more of a role in the media planning stage, rather than the vastly more granular aspects of media buying. In other words, the planners should tell the buyers what media genres are best suited for garnering attention, perhaps with weighting factors---but then let the buyers make the best deals they can under these guidelines.
Ed I would contend that the dramatic increase referred to was because the prior ads were being served to a phone that was switched on and unattended - just like mine is now, and beeping away.
While I am all for attention measurement (and it is WAY harder and more expensive to measure), the issue in media is that basically whoever 'owns' the content pays for its measurement. The TV networks pay for the TV ratings. Newspaper and magazine publishers pay for their readership data. Radio broadcasters pay for their radio ratings. Out-of-home site-owners pay for their 'eyes-on' measurement. The list goes on ...
But what I think was raised by Ishan and what is being discussed is attention to the ads (and this discussion is nothing new).
But who owns the ads? The advertiser. So pause and ask the questions, how many advertisers re willing to pay to measure the attention to their ads across each of the media they use.
Further the attention to the ad is affected by the placement and environment they are presented in. We have a good idea which sections of a magazine or newspaper work best. We know certain dayparts and content works better for certain brands on TV or radio. We know which billboards work. We know which websites and applications work best. But these are overall unquantified learnings. The kicker is we also know that putting a 'bad ad' into a 'good environment' won't work, as good media placement won't save a bad ad.
So I look forward to the day that advertisers stump up the finances to measure THEIR content. Mind you, I have been waiting over 20 years for that day.
John, I think that your wait is only just beginning. Moreover, it's a dream that the media will pay for anything that does not promote more ad spending. Attentiveness data---even if it can be defined and measured--- is certain to show that some ads work while others don't. Can you imagine a TV network telling any of its advertisers that not many viewers bothered to watch their messages? The problem---which I seem to be the only one articulating though it has long been evident----is that most advertisers think that their positioning strategies and "creative" are the entire ball game and that a "good" commercial will find its "audience" rather than crediting media targeting as being very important. This has long been the case with TV. Another problem is the idea, well established in digital media, that attentiveness can be measured by time- on- screen or other electronic metrics. While these can be directionally useful, they simply don't cut the mustard in terms that many advertisers can accept.
In short, you simply cant quantify everything----and one of the questions that has consistently defied such efforts is whether the "audience" was "attentive" to every ad placement, to say nothing of what exact effect each exposure had on sales. We have to rely on common sense and judgements founded on past experience once in a while.
Hear, hear.
Famously Einstein had a sign in his office ... "Not everything that matters can be measured, and not everything that is measured matters". He would have been a good media guy as well ... relatively speaking (chuckle, chuckle).
Lee, I agree, and thats why for mobile, we see viewable ads as not enough; A human transaction is the equalizer. In fact, at Velocity we were suprised to find up to 12% of all "100% viewable" mobile ad impressions as wasted as the device is unlocked but on a table or otherwise unattended.
The mobile phone is unique in that its always with us and is used for dozens of other uses, and ad interaction is usually at the bottom of the list. So its more important than ever to find optimal moments in the day that an ad is welcome, and advertisers will be rewarded with meaningful brand engagement.
Hi Jeff, I see all advertising becoming "online", or interactive in some way. E.G. there will come a time (pretty soon IMHO) when one will be able to interact with every ad, including the ones in the cinema. Perhaps through an app or theater innovation. Then dollars will flow towards ads with the right "incremental business interactions", like Lee mentioned, that are important for the advertiser.