Given the recent focus on transparency and brand safety scandals, it’s safe to say attribution may have slipped off the radar for some in the adtech industry.
It was refreshing, therefore, to get a recap on this equally as important corner of adtech at a recent IAB UK’s Back to Basics Attribution session.
It was a full house and the lineup of speakers included executives from Quantcast, Microsoft, Affilinet, Verve, and Fifty-five. The agenda saw them tackle attribution in display advertising, the attribution journey, attribution within the affiliate customer journey, online to offline attribution, and finally connecting offline sales to online paths.
As part of a global PR agency, specialising in technology PR, we are constantly evolving to keep up with adtech’s latest buzzwords – I’m looking at you PWA (Progressive Web Apps) and wrapper tags. So it was nice to ‘get back to basics’ with this session and think about one of the key factors in ensuring a successful campaign.
The four pillars of attribution
Data collection: Capturing the relevant data is the first hurdle to attribution, gaining a holistic view can be hard to do with the continuing dominance of walled gardens
User matching: Once you have the data, you then need to connect the dots. This can be anything from deterministic or probabilistic matching to estimations
Attribution modeling: The next step is for touchpoints to be weighted to form the attribution model – data driven, algorithmic, rules based, etc.
Insights: These can then be derived from the data and the analytical layer
Key takeaways – attribution for a successful campaign
When setting up your attribution model, the first step is to work out your data sources. From there you can determine your blind spots and understand where you need to bring in additional partners to build out a successful model.
When we look at location-based mobile attribution, and tracking both online and offline attribution, our data sources grow. This is great, however it’s important to also consider how to manage this data and ‘connect the dots’ to derive meaningful insights.
Ultimately, it sometimes takes concrete data on a campaign to move the advertiser away from their perceived path to success and point them in the right direction.
We are all aware that a holistic view of the campaign is the ultimate target for an advertiser. When we apply this goal to attribution, this means assigning credit to each and every touchpoint with the aim of rewarding correlation instead of causation. Often you will see campaign touchpoints weighted equally, meaning credit is often misdirected so you won’t see the difference between prospecting and retargeting.
This easy route of misattribution has become habit so the industry needs to ensure they are measuring the right insights from the data. They need to start asking questions to their attribution vendors and making sure that all members of the team can use the tools, not just statisticians and analysts. These steps will help marketers to move away from outdated last-click attribution models.
For me the session reinforced that we cannot afford to drop attribution from the ongoing discussions in adtech communities – regardless of what other stories may be hitting the headlines.
The processes used and insights gained are still evolving but if we get attribution spot on – and include our knowledge of viewability and brand risk factors – then we are well on our way to achieving that elusive holistic view of a successful programmatic campaign.
Read about GingerMay’s international PR services here.