How can we move beyond the last ad attribution model?
The minute that marketers use multiple media buying channels for advertising campaigns, they’re faced with an attribution conundrum. At a basic level, attribution is the analytics process of determining how effective each media buying channel is at producing the desired advertising outcome. It drives most of the campaign optimization decisions, such as budget allocation and tuning of campaign tactics.
At first blush, the challenge of attribution may seem solvable by simply assigning tags, specifying conversion events, and then letting ad servers report the performances on all the tags. But like most things in digital advertising, it’s not that easy, and in a media landscape of multiple media buying channels with real-time bidding strategies, many marketers are left wondering if their less-than-perfect attribution model is providing sub-optimal performance or even wrong decisions.
The current attribution model is the "last-ad wins" model. Basically, the last ad impression before the user conversion event gets 100 percent of the credit. Savvy marketers know that this instant gratification model oversimplifies the consumer decision process. Let’s pretend you see a great ad while watching the World Cup finals online. Chances are that you’re not going to rush to your online store to make the purchase right away. However, the next time you shop online or in stores, the influence of that brand is very likely at work. Attributing this type of delayed influences gives marketers visibility into what combination and sequence of ad messaging leads to conversions. It also provides the needed dials to optimize during dialogs with their consumers.
Even though the last-ad attribution model succeeds in its simplicity, it’s been under scrutiny for years, most famously called out by Microsoft’s Atlas Institute. The institute’s explanation as to why the current attribution model is broken mostly focuses on the concept of the purchase funnel. For example, the "last-ad" model would overestimate the performance of search ads because they sit so low in the funnel, while display ads are underestimated because they’re much higher up. To avoid flame-throwers from SEM (define) enthusiasts, let me limit the discussion on display ads.
In the new ecosystem of real-time bidding and audience buying, the flaws of today’s attribution model are magnified. The last-ad model narrows the focus on the last impression, ignoring everything that precedes the last touch. This myopia makes all sorts of craziness possible: disproportional focus on short-term gains, such as those consumers who are likely to convert with or without additional ad impressions; and cookie-spraying games where cookies are placed on as many people’s browsers as possible by purchasing cheap, and sometimes inappropriate inventory. Needless to say, neither tactic benefits the advertiser.
Buying audiences instead of content shifts the focus to multiple touch points of users’ experiences. How effective are different touch points in getting the desired user conversion events? With real-time bidding, marketers have more precise control of who, when, and how much to bid. It enables a user touch points-based bidding strategy to be executed. For example, if the advertiser’s goal is to influence as many consumers as possible with the given budget, we should stop buying impressions for users who are about to make a purchase anyway, and instead buy impressions of new users who are in the early stage of the decision funnel. It may sound counterintuitive, but it’s an important change of perspective.
As real-time bidding and audience optimization become the core strategy of more campaigns, there’s a greater need to move beyond the last-ad attribution model. The question remains - how do advertisers and agencies solve this? Here are three steps to take:
For example, you may find that the last touch point is only significant if it meets particular quality/engagement conditions. Or, maybe the first high-quality touch point is of almost equal importance for your campaign. The key to finding the right attribution model lies in the data, and although the most viable solution may require a lot of number-crunching work, it’s worth it in the end to have an attribution model that’s just right for your campaign.
As Chief Technology Officer at Turn, Xuhui Shao focuses on the power of optimization, machine learning, and advanced analytics solutions in driving new business models, products, and services across all industries. Xuhui is responsible for architecting the machine learning and optimization technology to deliver the most effective data-driven digital advertising in the world. He is passionate about the dynamic online advertising community and works closely with industry leaders developing data transparency and consumer privacy protection.
For the last 12 years, Xuhui has practiced research and development in machine learning, statistical theory, and computational intelligence for Fortune 100 companies in various industries from banking, finance, online retailing, healthcare, insurance, marketing, and online advertising. As the lead inventor and co-inventor of three awarded patents in the areas of advanced analytics and optimization, Xuhui is a recognized expert in harnessing data and transforming analytics into actionable insights and optimization strategies.
He earned his bachelor's and master's of science degrees from Tsinghua University, Beijing, and his Ph.D. in electrical engineering from the University of Minnesota.
2015 Holiday Email Guide
The holidays are just around the corner. Download this whitepaper to find out how to create successful holiday email campaigns that drive engagement and revenue.
Three Ways to Make Your Big Data More Valuable
Big data holds a lot of promise for marketers, but are marketers ready to make the most of it to drive better business decisions and improve ROI? This study looks at the hidden challenges modern marketers face when trying to put big data to use.
December 2, 2015
1pm ET/ 10am PT
Wednesday, December 9, 2015
5pm HKT / 5am ET