--- title: "Every Number Your Vendor Reports Is a Self-Report" url: "https://cooinsider.com/insight/every-number-your-vendor-reports-is-a-self-report/" author: "Ankush Gupta" published: "2026-09-25" updated: "2026-09-25" --- # Every Number Your Vendor Reports Is a Self-Report A vendor we were vetting looked clean. The traffic profile matched the asking price, the audience looked like the audience our clients want to reach, and the figure came from the third-party estimation tool everyone in our industry quotes at each other. I run marketing operations at FameNinja, and most of what we buy on a client's behalf is somebody else's audience. That audience figure is not a line item on the invoice. It is the product. It was manufactured. Self-refreshing scripts and bot sessions aimed at their own site, running long enough to move what the estimation tool displayed. The part worth sitting with is that nobody lied to us in a way a procurement checklist would catch. The vendor sent a real screenshot of a real tool showing a real figure, and every step of that paper trail held up. What had been faked was the measurement itself, upstream of anything we were inspecting. Internal process gets audited. External process gets a dashboard. Most operations leaders I speak to have serious rigour pointed inward. Approval thresholds, sign-offs, a QA step before anything ships, written procedures, somebody who reviews the reviewer. The same organisation will accept a vendor's monthly report as the record of what that vendor did, read the summary line and file it. This happens because vendor relationships get sorted as commercial rather than operational. There is a price, a contract, a renewal date, a person in procurement who owns the negotiation. It feels like a purchasing question. But the moment a vendor's output becomes an input to your process, their reporting is a control operating inside your business, and it is a control you did not design, cannot inspect and have never tested. In the cases we run, the vendor failures that actually cost something have all had the same shape. None of them were missed deliverables. Each one arrived on time, matched the brief and did nothing. ### **The second number has to come from your side of the wall** The fix is unglamorous. For anything a vendor reports, you need one independent measurement that you generate, even when it is cruder than theirs. We run a publishing network of our own, so for media work we can check the things a vendor's report tends not to cover. Is the placement live when you open it logged out of everything? Is the page indexed? Did it send a single referral session? None of that is sophisticated measurement, and all of it is ours, which is the only property that matters. Your own systems deserve the same suspicion. Our automation stack once reported healthy runs at every visible step while a backlog of stuck executions sat behind them, and the dashboard was not wrong about any individual node. It was answering a narrower question than the one we thought we had asked. A number you did not produce is a claim. Treat it as one. ### **Ask how the figure is produced, not what it says** The most useful question in vendor vetting is procedural rather than numerical. How is this figure generated, by which system, and what would it look like in a month where the work had not been done? A metric with no failure state is decoration. Strong vendors answer that in a couple of minutes, sometimes with mild irritation, because they built the pipeline and can describe it from memory. Weaker ones explain why the question is unfair. The one we caught did something more particular. They answered a different question, confidently and at length, and the confidence was the tell. Settle the measurement source before signing. Arguing about what counts as evidence in the middle of a performance dispute is a position you lose from the start, because by then the vendor has months of their own reporting and you have a feeling. ### **What this changes about planning** If vendor reporting is a control, it belongs on the same review schedule as every other control you own, with a named internal owner. Not a quarterly business review where the vendor presents their own scorecard and you ask questions about it. A check that runs on a date whether or not anyone is currently worried. Then budget for it properly. Verification costs hours from someone competent, and those hours almost never appear in the cost of a vendor, because the cost of a vendor gets recorded as the invoice. ### **The tradeoff** Checking vendors slows purchasing down, and it will occasionally insult people who deserve better than suspicion. Most vendors are honest. You will spend real hours confirming work that was already fine, and nobody will thank you for the ones that came back clean. That is the cost, and it is worth paying, because of what the alternative looks like. You do not find out at the point of the fraud. You find out later, once the number has been inside your reporting long enough that your forecasts, your budget and your promises to clients are all sitting on top of it. By then it is not a vendor problem. It is your number. --- Ankush Gupta is a Fractional CMO at [FameNinja](https://fameninja.com), where he works on online reputation management, digital PR and marketing automation.