My Completely Objective Ranking of iPX Announcements

Written by Brook Schaaf

impact.com romped through a tour de force of product announcements at iPX last week. I will do my best to give a fair rundown, going from least to most meaningful in my own judgment. Taken together, the announcements amounted to feeling out tracking alternatives in LLM land, a big wet kiss to the creator economy, and an advance on the big play announced last year. 

AI-generated storefronts and AI agents seem neat, but the latter is unproven, albeit exciting, and the former has heavy competition with existing widgetry tools like Affilize (by Affinity), Lasso, and even GeniusLinks’ Choice Pages, not to mention many sub- and creator networks, all of which seem anchored to the Amazon ecosystem. That said, it could be a very nice long-tail play, especially when the monetization pages can be generated with the ease of a prompt, as for impact’s newly-named “Software Creators” (aka vibe coders) category. 

On-platform AI-query capabilities and off-platform Model Context Protocol (MCP) connectors were also announced, though it’s unclear to me if these represent any particular advances over competitors’. For example, Everflow just announced an AI Playbook.

Clickless tracking in the zero-click era was much discussed, with one slide noting that “Every generation of marketers has faced a measurement shift,” followed by some adaptation. A framework of “Black box” vs. “Evidence chain” with trust based on “evidence, not authority” was proposed, and I wondered if this wasn’t a knock against walled gardens, though no further hint was granted. 

What followed might be characterized as a series of signal trial balloons (various forms of observation, inference, calculation, and, yes, traditional tracking) and program management levers (compensation based on “AI Equity,” audience engagement, and even impressions). What I mean by trial balloons is impact.com feeling out which signals are accepted and which are rejected as a basis for what kind of compensation. One imagines the mix will be on a per-advertiser basis, which presumably suits them just fine if they can customize everything. One also imagines this will spur the creation of more advanced affiliate-side signals to ensure they are maximizing their own compensation. This, in turn, would support approaches like the Geodesix licensing model of semantic sharing. (Disappointingly, General Manager Hanan Maayan did not literally teleport onto the stage for his presentation.)

The good news is that there are lots of solutions to the problems. The bad news is that it will probably take years to resolve, especially if it means more work for lazy affiliate managers. My prediction: future discussions over up-funnel compensation, especially of a probabilistic nature, will dwarf past wrangling over last-click attribution, which will seem petty in hindsight.

The last two announcements are tied for most impactful (ha), the first likely winning over the short term, the second over the long term. The first is deeper integrations with social media platforms. An employee of Instagram came on stage to explain their integration, which was a rather nice encore to April’s news of impact’s inclusion among 17 API partners to access YouTube creator information. The social media platforms seem to finally be following YouTube’s years-long lead and waking up to the power of supporting affiliate monetization, making these platforms stickier for creators, which in turn means more content to attract and retain audiences. This is the new expectation.

Second, we have the unfortunately named “retail media” functionality, which I think may make for the most dollars and greatest competitive advantage. As a quick aside, it’s not retail media the way that Amazon is retail media, a term itself which has been partially succeeded by the term “commerce media,” which floats among others like PayPal’s own “Financial Media Network.” Perhaps the accepted term will eventually be “Publisher-That’s-an-Advertiser-or-at-Least-not-Normally-a-Publisher Network.”

Anyway, this consists of two parts. The first is what we normally call a Seller Network, like Levanta, PartnerBoost, or Amazon’s own Creator Connections. The second part is product boosting, paid by the manufacturer brands, which are called vendors. You may recall this was announced last year and stirred great excitement in me. It has apparently come to fruition. At the time, I found a 2023 reference to $75 billion in annual co-op dollars. This year, the number was over $200 billion; I suspect the delta is mostly global vs. domestic. This is perfect for performance marketing and may well bring in massive additional dollars to our space.

Finally, as Mike McNerney shared in his newsletter, Bryce Widelitz opened the show with an unusually entertaining corporate skit. Now he just needs to pair the video content with a storefront.