Viewability and marketing attribution continue to be two big topics in 2015. Although they share some common ground, which is to help optimize display impressions, they ultimately play different roles in the optimization journey and tackle different problems. If they are coordinated incorrectly or used in the wrong way, marketers can end up compromising ROI without realizing it.
Viewability is about understanding which sites, publishers and Private Marketplaces (PMPs) are providing quality inventory space and the probability that the impressions they serve will be seen by a user. The IAB has been working on guidelines to outline what is meant by ‘viewable’ and defined it as “more than 50 percent of pixels shown in the browser view for more than one second.” However, this metric typically does not take into account the cost of the inventory or the quality of the audience reached.
Attribution, on the other hand, focuses on how each media touchpoint across all channels in a market contributes to an organization’s overall marketing performance and provides each touch point with its fair share of the credit. Attribution will also provide actionable budget recommendations, which can be put into market to help reduce CPA and improve ROI.
While viewability is important, there are several reasons why marketers should take care when bringing the resulting data from such a metric into an attribution model.
First, when attribution modelling is implemented correctly, it naturally measures the impact of viewed and non-viewed ads on conversions and enables marketers to reallocate budget as a result. If an ad impression isn’t viewed, it will have little impact on a user’s likelihood to convert and therefore will naturally receive less credit. The general consensus, however, is that non-viewable impressions should be stripped of their weight and given zero credit. But who is to say that an impression definitely wasn’t viewed? Viewability technology is not 100 percent accurate today, and I have seen instances where the viewability status was “unknown” for upwards of 25 percent of impressions – particularly when the technology was unable to penetrate embedded iFrames or certain device browsers.
To arbitrarily take credit away from a touch point that is not viewable or where viewability is unknown would mean skewing the attribution model and the recommendations it delivers. What if the viewability technology deemed an impression as non-viewable, but the attribution model says that it had a significant impact on driving a conversion? Is the model wrong, or the viewability metric? Furthermore, a number of companies claim to measure viewability, each using a slightly different approach. Facebook, for example, recently released its own viewability pricing model using a different definition of viewable than the IAB. All these questions and variations add uncertainty and may ultimately lead to incomplete, inconsistent and/or inaccurate information on which to base optimization strategies.
Secondly, attribution is about improving ROI by making recommendations on how best to allocate budgets across all channels. Yet delivering budget allocation recommendations based solely on viewable impressions is fruitless, since it’s not possible to only buy viewable impressions today (viewable impressions should not to be confused with above-the-fold impressions). Once technology exists that enables marketers to buy viewable impressions only, this will become a moot point – if all impressions that are purchased are viewable, then viewability data won’t need to be incorporated into an attribution model.
So if viewability doesn’t matter when it comes to attribution modelling, is it still useful? Absolutely! First, it is a great tool during negotiations with publishers and programmatic marketing platform providers. Use it to get them to improve viewability and performance across their site or network, or to negotiate cheaper prices. You can even try and get them to move to a viewable cost per thousand (vCPM) pricing model. Another great use of viewability is to validate that the publishers are providing what they promise (though any two measurements of viewability will give you different results, so do take that into account). If you are paying premium prices for viewable inventory, use viewability to make sure you’re getting what you pay for.
Overall, both viewability and attribution technologies are powerful and will help you drive better ROI and reduce CPA across your business. They just need to be used correctly and with a bit of consideration.