/Sweepstakes casino bans: what operators and affiliates faced/Stake sponsorship case: Everton, UFC, F1 and Kick/Sportsbook marketing: what DraftKings and FanDuel spend/SaaS Affiliate Programs: A Blogger's $680K Semrush Case/Robux gambling ruling: why affiliates are exposed/PrizePicks: how DFS pick'em became a $2.5bn US business/Popunder Ads Case: $30K in MENA, 1,536 iGaming FTDs/Player lifetime value case: $9,125 in CPA, $61,376 GGR
Episode from The Affiliate Marketing Podcast. Summarised by the AffHouse newsroom from the full episode. How we work →
Aired Published Updated
Affiliate managers are short of time, and Everflow CMO Michael Cole says AI is most useful for the small, personal jobs no product was ever going to be built for. On The Affiliate Marketing Podcast he explains what works, what breaks, and where a program manager should start.
In this episode
Cole says AI lets teams build dashboards and automations for a few people without developers; Everflow built a resource hub that tracks its content the way you would track affiliate links (3:34, 4:43).
He says "AI is very bad at judgment": his content team produces three to four times as much content, but humans still focus on polishing and choosing (7:42, 7:50).
The most useful cases he sees with customers are back-end admin, reporting and data analysis, tedious tasks that used to take 5 to 10 hours (12:26, 14:41).
Everflow has lots of customers in a closed beta of its MCP server, so their agents can pull data from the platform (17:16).
On guardrails, he wants a secure data layer; vibe coding suits small projects that are not interconnected, while developer-built systems must never break in ways that take reporting down or stop affiliate payouts (20:44, 21:22, 21:48).
To start, he suggests asking AI for affiliates that performed in the past but not in the last year, and starting small with a spreadsheet and AI-written app scripts (23:33, 24:00, 29:16).
Episode from The Affiliate Marketing Podcast by Affiverse, published on YouTube.
