If a tree falls in the woods and no one is around to hear it – it does, in fact, make a sound. Similarly, if an ad influences a consumer’s purchase decision, even if it doesn’t fit
neatly into an existing attribution system – it has, in fact, made an impact.
Unfortunately, the RTB ecosystem, with its fundamental need for measurement and attribution, could be holding
marketers back from fully realizing the potential of their overall media plans. By continually reinforcing the notion that spend should be focused on channels that have the most measurable or
attributable performance in the context of the consumer purchase path, brands risk fundamentally misaligning their spend with their goals.
It’s incorrect to assume that media
that can’t be measured isn’t effective. But the greater and subtler risk is believing that just because a form of media can be measured, that the attribution actually measures the efficacy
of that media in isolation – or even with any accuracy. To take the most obvious example, the measurement gaps in traditional Nielsen-based TV ad effectiveness measurement are plain to even the
most casual observer. But we know that TV has been and continues to be a highly effective medium for advertising. Biasing media plans against TV because it’s not precisely measured would be
absurd. On the other hand, a marketer’s spend on Google Adwords – or an active radio campaign – may have more to do with a display advertising campaign’s success than any RTB
platform’s technology, but this won’t show up in the average marketer’s data.
advertisement
advertisement
Most impressions of banner ads are never actually seen. Most users never, ever click on banners.
And yet, studies show that non-clicking users shown an abundance of banner ads – relative to those in a control group – will convert more often. So what exactly is attribution technology
measuring, and what is RTB technology optimizing on – if they don’t know which ads are actually seen (this is not the same as viewability), nor do they do know which ones actually have an
effect?
Unfortunately, the answer is often the last-clicked ad, the last absolute ad seen – or some vague, proprietary black-box algorithm that also doesn’t know or use the salient
data points just discussed. Of course, this completely ignores the impact of any other channel, such as out-of-home or search. Yet RTB has been hailed for the precise valuation of impressions based on
sophisticated algorithms. The reality is that it’s using a rough approximation based on a number of dubious data points (such as clicks), while ignoring a litany of meaningful unknowns (such as
which ads were actually seen).
The fact remains, however, that RTB and attribution technologies are actually measuring, optimizing and reporting on something. Marketers receive weekly reports
with numbers, graphs and pivot tables. This is reassuring and likely has a directional relationship to the truth. But it also leads brands to prioritize partnering with people that can deliver the
same. After all, the marketing of RTB and attribution companies is completely tailored to emphasize both the quality and importance of their efforts.
The temptation of digital is to create
causal relationships. Yet we simply do not know when ads are seen, let alone effective. That’s true for media that can technically enable traceable conversion attribution, like banner
advertising, as well as media that cannot, like radio. But just because we can do it, doesn’t mean we should. Instead of focusing on holding vendors only to the conversions we can directly
attribute, we should – at the very least – be measuring lift between control groups and exposed groups. The vendor would ideally get credit for all lift over the control, regardless
whether ads were clicked or not.
Opportunistic technology vendors try to bias brands toward banners and away from spending on TV, radio and mobile because those channels don’t fit into
their attribution models. While most brands aren’t necessarily going to fall into this trap, we need to remain vigilant about shaping the debate to ensure that certain channels
aren’t underrated, and that we really move the industry toward a marketing paradigm that acknowledges both the knowns and unknowns of marketing efficacy.