The news
Martech.org ran a piece from Awin's North America president arguing that affiliate marketing — despite averaging 12-15x return on spend — keeps underperforming because companies run it as a technology platform instead of a managed partnership portfolio. The argument: AI will accelerate execution, but it won't replace the strategic judgment that makes affiliate programs actually compound.
Our take
dAIs agrees with the conclusion but wants to sharpen the diagnosis — because this isn't really an affiliate problem. It's the same pattern showing up across every GTM channel teams try to hand off to software.
The failure mode goes like this: a team buys a platform, connects the integrations, sets up the rules, and then mentally checks out. The tool runs. The reports populate. Nobody's actually reading the signal. That's not AI over-reach — that's abdication dressed up as automation.
Affiliate just makes this failure unusually visible because the economics are so outcome-dependent. When your paid search underperforms, you might not notice for a quarter. When your affiliate program is running on autopilot, you're leaving partner relationships to atrophy in real time and you'll see it in revenue.
The piece correctly points out that Gartner data shows companies using less than half their martech stack capabilities on average. But the instinct to solve that with more AI — smarter dashboards, better recommendations — misses the point. The capability gap isn't technical. It's operational. Teams don't have documented processes for what "good" looks like in their programs, so there's nothing for any tool, AI or otherwise, to optimize against.
This is where it hits a wall: the automation is ready before the operator is. The team wants to deploy AI to manage partner tiers or personalize outreach, but they can't articulate what a high-value partner looks like or what the escalation path is when performance drops. That's not a workflow problem. That's a process documentation problem, and no amount of AI capability closes it.
The so-what
The channels that return the most — affiliate, strategic partnerships, high-touch ABM — are exactly the ones that break fastest when you try to fully automate them. That's not a reason to avoid AI in those workflows. It's a reason to be specific about where AI accelerates human judgment versus where it's being asked to replace it.
Before you wire AI into any high-stakes GTM motion, ask whether you could hand that process to a smart new hire and have them run it successfully. If the answer is no, the process isn't ready for automation — and adding AI will just make the gaps harder to see. The teams pulling away right now aren't automating more. They're documenting better, then automating fast.
---