Attribution modelling, which has the goal to determine the value of a single media channel or ad element, has become a hot topic ever since site-analytics platforms have allowed tracking detailed web traffic and conversions around 2006/ 2007.
However, while there are many good reasons for using attribution modelling, we still need to battle some basic advertising myths in this regard: according to the Experian 2015 Digital Marketer Survey, 60% of marketers employing media attribution keep using simple rules to allocate which ad obtains credit for conversions. In particular, the last-touch attribution – which dictates that the last event before a conversion gets 100% of the credit – is still prevalent among agencies, trading desks, ad-ops teams, and clients.
But does this heuristic really do justice to all the other possible advertising touchpoints that may have contributed to driving a consumer closer to a brand?
For example, a customer first sees a couple of ads on news pages, then clicks on a referral on Facebook, sees more retargeting ads in the following days, and then searches the brand online before buying from their homepage. In this case, it would be quite short-sighted to assume that only the final search entry made the consumer purchase from this brand.
So why is the last-touch measure still so popular despite bloggers, trade magazines, and academics pointing out for years that this this approach is flawed and unfair?
To understand the persistence of “last touch” better, one needs to have a closer look at the key stakeholders and how the ad-serving ecosystem works.
From a practical perspective, the last touch method is easy to understand and execute. In fact, it used to be the only attribution standard implemented in Google Analytics in the beginning. While other methods exist and have been applied more recently (e.g. algorithmic multi-touchpoint models), not every CMO may have the skills in their team to run statistical models or the budget to buy a third-party solution.
And it is easy to see why some companies are hesitant to invest too much yet. Since the whole area revolving around attribution is fairly new, there is little evidence revealing the monetary benefits of using more sophisticated methods versus free tools (also, these days site-analytics providers offer more flexible allocation rules that consider all touchpoints, such as equal weighting of all channels). No surprise that clients ask themselves: are the efficiency gains worth the extra efforts and investment in advanced analytics?
The best way to prove the power of more robust attribution methods is to take action based on a model’s recommendations and then track KPIs and financial performance. However, this can be quite challenging because other aspects may also be behind the ongoing use of last-touch attribution among certain stakeholders. For example, some trading desks, DSPs, and media buyers may have their own motives to assign credit according to “last touch”, because it is possible to game the system for their advantage. The trick is to find cost-efficient ways that place an ad at the last position before someone converts and thus to “steal” the credit.
How can this be achieved?
One possibility is to buy high volumes of very cheap inventory (below-the-fold or hidden, remnant impressions) and spread these all over the internet to make sure everyone is covered by these ads. This approach works particularly well for attribution stacks that only look at converters and neglect the long tail of non-converting consumers who have been exposed to the same inefficient ads.
As an alternative to this strategy of mass prospecting, one can exploit extreme retargeting with the goal to “throw your ad” at customers who are about to convert. Ever wondered why retargeting has nearly always one of the lowest CPAs? One contributing factor is that many consumers may have converted anyway, but by piggy-backing an ad to the process it looks like the ad was driving this behaviour. Hence, whilst retargeting can be very effective when done smartly for the right customer, it may not be as strong a tactic as often assumed and it can also be exploited.
Finally, it should be noted that one of the biggest winners of last- touch attribution typically is the SEM industry. The reason for SEM being so strong here is that consumers often search the desired brand or product online when they intend to return to the website and convert. Yet, whenever multi-touch methods are used to assess marketing tactics, “search” tends to experience the greatest losses in comparison to the last-touch approach. Because many organisations spend a huge amount of their budgets on SEM optimisation (guided by the traditional paradigm that this is the only true efficient channel), the message that paid search ads are often much less effective than they thought initially is not easy to communicate.
In sum, independent of the possible motives for using last-touch attribution, publishers of premium content may not receive the credit they deserve for quality, high-impact ads. After nearly 10 years of media attribution and the recently increasing discussion on transparency in the media business, everyone should focus on providing maximum value to their clients. As a result, it should be neither in the interest of professional publishers nor advertisers to rely on last-touch attribution models.
By Nico Neumann
Senior research analyst at University of South Australia
Did you catch this last opinion piece by Nico, when he was lead statistician - product development at Anomaly Research and Analytics (part of IPG Mediabrands)? Check it out here: Why attribution modelling is such a hot topic.