The Attribution Question
Like many of you, I look at traffic and conversion reports every day for various ad campaigns from different digital traffic sources for various advertised offers.
Accordingly, the ads placed, whether SEM/paid search, display, social, or otherwise get optimized both manually and systematically on a regular basis to extract the best possible results.
There are typically target CPA or ROAS goals baked into every campaign purchased, with pixels, tracking URL’s, cookies and various other technical methodologies to generate reports showing tangible results which validate success or show where improvements can be made….or both.
Simple enough, right? Attribution for digital media buys in your bag of marketing tricks (aka layers of the funnel) comes down to assigning credit where credit is due, in terms of which impressions, clicks and resulting actions such as sales conversions were generated by which ad/source/campaign/ad creative used/audience/platform/media vendor, and even by which person or agency handles that conversion generating media buy for you.
But, I have to say it, this notion of attribution and some of the existing standard practices related to measurement, is a very subjective piece of the science of cost accountable digital media. Especially now, in an era where attribution judgments (did the campaign work? Was it profitable?) are being made by media buyers with ad sources driving view throughs, a measured event viewing of ads, such as display & video ads, but which did not yield a click/conversion immediately, yet did so within a defined window of time afterwards).
Here is a (very common) scenario and thus my beef with how the digital media industry standards for this can be biased: a user who visited your eCommerce site on Monday but does not convert, then sees (views) one of your retargeted display ads on a major news site on Tuesday, creating the view through which can be attributed to the retargeting campaign, yet does not click through and visit your eCommerce site or take any other actions at that moment. The same user ultimately comes back to your site and converts to sale on Thursday (always a beautiful thing if profitable), two days later. Many retargeting colleagues have said that this view through on Tuesday from the retargeting campaign should get the attribution credit for the sale.
But…between that Tuesday view through and a purchase completed on Thursday by that user, was that user exposed to any other media/campaigns related to your site, brand or product, whether the additional media or ads were earned, owned or paid for? Perhaps the user eyeballed some information elsewhere in the vastness of the Web and Social networking about your brand, your product lines, specific products, etc? Maybe you got unexpected and positive news coverage on Wednesday that impacted the user’s viewpoint?
Further, how about the cloud or CDN network enhancements that kicked in in this scenario Tuesday night or Wednesday, which made the UX on your site on Thursday faster, better or otherwise more enjoyable for your user to complete the purchase.
Did one of your brilliant design artisans adjust the shade of blue in the background on the product page that this user saw, between visit 2 and visit 3 to the site?
Perhaps your merchandisers tweaked the category and PDP (product) pages a bit to reflect more product information than had been previously known, or added a product bundle for purchase which made our user want to convert even more on their Thursday visit.
Why do these variables matter for trying to cost account the success of a given digital ad buy or campaign? Plain and simple. If we are to be truly transparent and even holistic in approach to understanding true cost of a customer acquisition from digital media campaigns, then we need to get real in terms of factoring in the wide range of other elements, changes, nuances or events which impacted that user’s behavior, aka customer/consumer journey.
I believe strongly that digital media attribution standards should allow for shared credit, perhaps via a points system or some other method that a great mathematician can help devise. Not just shared credit among the 3-5 digital media sources and ads potentially seen by the user before they converted to sale, but also shared credit for site or vendor-internal changes made during the user’s view through window after their first ad exposure on Monday which may not always get accounted for, unless they were part of an intentional A/B test.
Even the subtlest events which could impact the user from Monday’s visit to your site turning into a conversion on Thursday deserve credit for their part in the process of making a sale happen; adopting at least some elements of this view can allow all of us involved in the digital media and eCommerce ecosystem to have ultimate transparency with factoring how much it really cost to generate the conversion as well as deeper understanding of what actually caused the conversion to happen.
Rant completed. Have a great weekend!