Paths Are the Way

Written by Brook Schaaf

So far in this series, we’ve covered misbehavior by affiliates, centered around Phia’s forced clicks (aka stuffed cookies aka auto drop), followed by unfair advice from the program platform ShopMy, along with a reminder that tracking can easily break on the merchant side. The point is that every player in the ecosystem can disadvantage counterparties, deliberately or accidentally. Affiliates can manipulate tracking. Merchants can foul up tracking. Platforms can express their own conflicts of interest.

This is bad for business. It creates confusion and distrust inside and outside of the affiliate ecosystem. As one reviewer wrote about Phia’s browser extension, “Another cookie-stuffing extension you.” Never mind that the issue was with the mobile extension, not the desktop browser extension; this attitude is going to keep users from engaging. This mistrust is bad for the entire affiliate channel because consumers will bypass these properties, much as they already do coupon sites because the codes they find never seem to work. (Not from FMTC, of course.) 

There are changes we can make to improve this situation.

Before we get to the proposed solution, there’s a widespread argument being made that should be countered. ShopMy’s since-erased statement justified voiding commissions on competing networks because “ShopMy provides a more direct brand-to-creator connection.” Although he didn’t call for possibly violating contractual terms with other networks, Paul Archer, CEO of referral network Duel (billed as a “brand advocacy platform”), laid out the same self-serving argument in a Performance Marketing World interview: “[m]isattributed affiliate credit has impacted performance marketing for years” because it “creates an immense problem, where the person who genuinely created awareness, intent, or advocacy doesn’t get financially endorsed for their efforts.” Ali Kriegsman offers up this same specious argument alongside helpful explanations of both Phia and ShopMy. 

There are two claims here. The first is that last-click attribution is imperfect. Of course it is. The second is that creator content therefore deserves the credit that last-click gives someone else. That does not follow and actually begs the question, i.e., it presumes as true or settled the premise that creator content is what “genuinely created awareness, intent, or advocacy.” 

Maybe. Maybe not. If there is a single touchpoint that deserves credit for a transaction, you never know which it really is, because you don’t know what foreknowledge or intent any given user had prior to her appearance in the funnel. Suppose I have already decided to buy from a merchant, Google its name, and click the merchant’s paid search ad. Google may get 100% of the observable credit while having contributed approximately 0% of my intent. 

So is what a creator posts more valuable than navigational, commerce, or coupon content? Possibly, but not necessarily. Different users may consciously assign their own value to touchpoints. As you may recall, my wife is obsessed with Rakuten Rewards. Her primary shopping intent is often to basically get points there. At the same time, she has bought plenty of stuff she’s seen on Instagram ads and was watching what was basically a YouTube product placement video for a family digital calendar called Skylight. 

As Kriegsman notes, “a user often has many touch-points in their path to purchase.” This is correct, and surely there is value in each touchpoint, but another statement calls for more scrutiny: “Last-click hands the entire commission to whoever happened to be standing there at checkout.” 

This is not necessarily true because attribution and compensation are separate, if overlapping, things; there is often non-commission compensation. In fact, hybrid deals are probably the norm, in part because the click or coupon tracked through a third party is a reliable touchpoint. And if there are only creators in the click stream, one will win at the expense of all the others. So paid placements and impression buys make sense. 

Moving away from measurable performance doesn’t solve the problem; it replaces attribution uncertainty with pricing uncertainty. Per “Creator industry admits that fee pricing is out of control, but can’t agree on a fix,” in Digiday last week, “Half of marketers misprice creator fees, and 40% of them feel like they overpaid, according to 1,000 marketing and procurement leaders surveyed by Billion Dollar Boy.” 

This kind of distrust pushes advertisers back toward performance compensation, including commissions, which takes us back to the contentious, unresolvable issue and a proposed solution: the advertiser creates multiple pathways and advises each publisher which pathway they are on. The merchant’s marketing team must decide what pathways exist, who they want to work with, and negotiate around remuneration, but they should also provide as much clarity as possible around how the tracking works. 

This solution largely already exists. The giant walled gardens are usually already their own pathways with their own integrations with shopping carts, which they can “optimize to” (some might argue the optimization is more for the walled gardens than for the advertisers). So if you use, say, Alphabet, Meta, Open Web Programmatic, and two affiliate networks, you might have five pathways, each conceivably getting credit for the same sale. Within one network, you might tell each affiliate they are on plan X, Y, or Z. Again, this largely already exists. CJ was the first network I was aware of to show affiliates who else might have won a commission when they were in the same clickpath. 

As I recall, CJ was also an early provider of information such as whether a merchant was Mobile Certified, Cross Device Enabled, Cookieless Tracking Enabled, or Content Certified. Others offer similar data, but none are enough. 

Each program description should state how tracking works (e.g., batch/server-side vs. pixel), and this information should be available by API so it can easily be retrieved and updated. Then affiliates could better observe correlative patterns and merchants would have more of an incentive to implement best practices. Ideally, there might also be some kind of ancillary information like the percentage of clicks that would have been transaction-eligible. This information is not overly sensitive, and merchants are probably already giving it to the walled gardens, anyway.

Rules and enforcement still matter. That’s why the PMA’s Proposed Industry Toolbar Standard is welcome (especially the part about removing bad coupons), even if it’s largely a consolidation of existing standards and practices. At this point, improving trust is more about information sharing than policies because affiliates, platforms, and merchants can all be negligent or worse, and this is hardly limited to the affiliate channel. Consider what came up in Google’s antitrust trials and was allegedly repeated by Amazon. 

This kind of behavior can never be eliminated entirely. Neither can uncertainty around attribution. The industry doesn’t need to agree on who metaphysically “deserves” every sale. It needs merchants to establish the rules beforehand, disclose them clearly, and apply them consistently.

We may never know which attribution is truly correct. We can at least know which attribution model is being applied.