The Partner Roster You Haven't Re-Checked in a Year

Vetting is treated as a moment, but risk is a duration

Ask an affiliate team how they manage compliance risk and most will describe onboarding: the checks they run before a partner goes live. That is the moment everyone focuses on. It is also the wrong place to look for the biggest exposure. A partner who was clean at signup can change tactics, add markets, refresh creative, or start promoting offshore brands months later, and almost no program goes back to look.

The roster you approved last year is not the roster you have today. Partners drift, and because the operator carries strict liability for what they publish, that drift is your liability accumulating quietly in the background while attention stays fixed on new applicants.

What drift looks like in practice

Drift is rarely dramatic. It is small changes that compound.

• A partner adds a new language and starts reaching a market you are not licensed in.
• A clean comparison page gets a new banner with a non-compliant claim baked into the image.
• Bonus terms that were visible get pushed below the fold in a redesign.
• A dormant page gets repurposed and starts promoting a competitor's offshore brand alongside yours.
• An old promotion lingers in cached and syndicated copies long after it expired.

None of these trigger an alert in a program that only checks at onboarding. All of them are live evidence if a regulator looks.

Why this lands on the operator

The strict-liability principle does not have an expiry date. A regulator examining an affiliate page does not ask when you approved the partner, it asks what is on the page now and whether you had a reasonable system to catch it. 'We vetted them in 2025' is not a defense for what they published this week. The only posture that holds up is continuous, because the surface you are responsible for changes continuously.

Why this is a visual problem, not a text problem

Re-checking a roster by hand is hopeless precisely because the drift hides in the visual layer. A re-crawl that scans text will miss the new banner claim, the child-appeal imagery, the prohibited market in a screenshot, and the spoken promise in a refreshed video. To re-check a roster meaningfully you have to see the pages as they render now, not diff the source code.

Build a re-check cadence

The fix for drift is rhythm. Set a standing cadence so no partner goes more than a defined window without a fresh look at their live, rendered footprint, and make high-traffic and high-drift partners cycle faster than the long tail. The cadence does not have to be uniform: tie frequency to risk, so the partners most likely to change get seen most often. What matters is that re-checking is a scheduled property of the program rather than something that happens only when a regulator or a complaint forces it. Drift is continuous, so the watching has to be continuous too.

Where kaspero fits

kaspero keeps the whole roster under continuous watch rather than checking once and forgetting. It renders every partner's current footprint across websites, social, and video, judges it against the right market's rulebook, and surfaces what has changed, with a timestamped visual record of when it changed. Because it runs continuously on public data and is agentic, re-checking the roster stops being an annual fire drill and becomes a standing background process you can also query on demand.

Three moves worth running this week

1. Export your full partner list and sort by approval date. Everything older than twelve months is unverified against today's content and today's rules.
2. Spot-check five long-tail partners you have not looked at since onboarding. Long-dormant pages are a common home for stale or repurposed content.
3. Check whether any partner has added a language or market since they joined, and confirm you are licensed everywhere they now reach.

The takeaway

Onboarding tells you who a partner was on day one; only continuous monitoring tells you who they are today, and today is what you are liable for.