The invoice from your affiliate network never has a line item for margin you would have kept anyway. That's the whole problem with brand bidding, and it's exactly why it survives inside programs that would never knowingly tolerate it.
By [Author name] · 18 September 2026 · 9 min read
Someone types your brand name into the search bar. Not a category term, not a competitor's name with your name as a comparison, your actual brand name, spelled correctly or close enough. That person has already decided who they're buying from. They are not shopping. They are closing.
An affiliate, or a hijacker with no relationship to you at all, buys a paid search ad on that exact term. If the shopper clicks the ad instead of your own listing, whoever placed that ad gets credit, and in the affiliate case, gets paid, for a sale that was going to happen with or without them.
That's the entire mechanism. It really is that simple. What isn't simple is getting a finance team to see a completed conversion as anything other than good news, because on the surface, that's exactly what it looks like.
Compare it to what a legitimate affiliate actually does: a comparison site writes an honest review, a coupon page catches someone who was already going to abandon their cart, a newsletter introduces your product to someone who'd never heard of it. All of that is demand the affiliate helped create or rescue. Brand bidding creates nothing. It sits in front of demand you already had and takes a fee to let it through.
There's a mechanical reason brand bidding is so hard to see in your own dashboards, beyond the invoice simply looking fine. Most affiliate programs pay on a last-click or last-touch basis inside a cookie window, often 30 days. A shopper can research your product for two weeks through organic search and your own emails, fully intending to buy, and still click a coupon-site ad on your brand name in the last five minutes before checkout out of pure habit, the way a lot of people check for a discount code before paying for anything.
That click resets attribution. The affiliate who contributed nothing to the previous two weeks of research now owns the sale, commission and all, because they happen to hold the last click inside the window. It's not fraud, exactly. It's a rule (last click wins) that was written for a world where the last click usually reflects real influence, applied here to a case where it plainly doesn't.
Here's where I think most brands get the accounting wrong, not out of carelessness, but because the affiliate channel's own reporting is built to make this invisible. A commission paid on an intercepted branded click shows up identically to commission paid on a sale the affiliate genuinely sourced. Same report, same column, same 8 to 12 percent (or whatever your program pays), same green checkmark next to "conversion."
Nobody's payout report has a field for "this one was already coming to you anyway." The network isn't going to flag it either. They get paid on volume regardless of where that volume came from, so their incentive to help you separate real affiliate-sourced demand from redirected demand is, at best, lukewarm. I don't think that's malicious on the network's part. It's just not their problem to solve, and expecting them to solve it for you is the first mistake I see brands make.
The honest way to model this, in my opinion, is to stop treating that commission as a cost of acquisition at all. It isn't acquiring anything. It's a discount you're handing to whoever happened to stand between the shopper and your checkout, funded entirely out of a margin line, not a growth line. Once you look at it that way, an 8 percent commission on a branded-search sale stops looking like a reasonable customer acquisition cost and starts looking like what it actually is: a toll.
There's a second cost that almost never makes it into anyone's analysis, and it's the one I find people most reluctant to believe until they've watched it happen on their own account. When an affiliate or a hijacker starts bidding on your brand name, they don't just risk stealing a click. They become another bidder in that keyword's auction.
More bidders on the same exact-match term pushes the floor of that auction up. Everyone still competing for position, including you, on your own name, now pays more per click to hold the spot you used to hold more cheaply. You can lose zero clicks to the interloper and still pay a higher CPC on your own brand name because they showed up in the auction at all.
I'll admit the caveat here: isolating this effect cleanly from ordinary CPC drift (seasonality, your own bid strategy changes, Google's own auction mechanics shifting quietly under everyone) is genuinely hard without a controlled experiment, and I haven't seen anyone run one publicly. Treat it as directionally real, something to watch for in your own CPC trend, not a number you can quote with confidence to a board.
We ran an early scan for a brand we monitor, a supplement company, across three countries: Hong Kong, Australia and the US. Two hundred search sweeps in that first pass. Those sweeps recorded 734 individual ad observations, meaning 734 separate instances of an ad showing up on a specific search at a specific time, before any grouping happened.
Grouped down, those 734 observations resolved into 40 distinct cases. Thirty-one of the forty were the brand's own affiliates, already inside the program, bidding on branded terms their agreement almost certainly prohibited. Nine were hijackers with no affiliate relationship at all.
That split doesn't surprise me. In most programs I've looked at, brand bidding by your own existing affiliates outnumbers outside hijacking by a wide margin, because your affiliates already have the tracking infrastructure, the account, and the financial incentive sitting right there. A hijacker has to build all of that from nothing just to take a shot at your name.
Scale matters here too, not just the count. One of the forty cases, on its own, was built from 88 separate observation rows off just 12 keywords over a two-day capture window. A "case" isn't one sighting of one ad. It's a pattern, captured repeatedly, before anyone's willing to call it a finding rather than a fluke.
Now the wrinkle, because I think it matters more than the headline count. Of those nine hijacker cases, four different advertiser names showed up in the public ad-transparency record. One of those four names sat behind five of the nine cases on its own, which reads to me like a single operator running several campaigns rather than five independent actors. One case disclosed no advertiser identity at all. And one, on closer review, turned out to be the brand's own official advertiser account, running under a name nobody had cross-checked against the roster yet.
I'm including that last one deliberately, not to undercut the number but because it's the actual failure mode of a first pass, any first pass. Automated detection produces candidates. It does not produce verdicts. Skip the human review step and you will, sooner or later, accuse your own media buyer of hijacking your own brand.
You don't need a tool to get a first, rough answer to whether this is happening to you. You need about an afternoon and access to reports you probably already have.
That last check is the manual version of what a continuous monitoring pass does on a schedule, at more search volume and across more regions than a person can realistically sustain by hand. Do the manual version first. It costs nothing, and if it comes back clean a few weeks running, you've bought yourself a genuine data point either way, not just a guess dressed up as one.
None of this requires taking my word for it. Pull the order report. Run the auction insight check. Do the incognito searches for a couple of weeks and actually write down what you see. If nothing turns up, you've lost an afternoon and gained a real answer. If something does turn up, you'll have found it the same way we did on that supplement brand: by looking properly, not by assuming a clause buried in an affiliate agreement was quietly doing the enforcement for you.
Tell us your domain and the keywords that matter. We'll scope the monitoring and quote based on your program size.
Get a quote