Twice a year a client forwards me the same article. A brand paused its branded search ads, nothing bad happened, and therefore branded search is a scam the platforms run on the innocent. The story recirculates because it is satisfying, and it is satisfying because it is sometimes true. eBay's economists made the canonical version of the case a decade ago, and I was buying media for eBay in that era, so I have lived inside both halves of this argument. The part worth taking seriously is that brand spend can be pure waste. The part worth rejecting is the ideology on either side, because whether brand bidding pays is not a belief. It is a measurable property of your specific search results page, and it changes without telling you.
Three inputs, no ideology
The math has three inputs. First, what does the organic result look like when the ad is off: are you the clean first result, or does a reseller, a marketplace, an affiliate, or a competitor's conquesting ad sit above you? This is not a question you answer from your own desk, where personalization shows you a flattering page; check it logged out, across geographies, on mobile, where the ad-free view above the fold may contain no organic result at all. Second, what is the click actually worth downstream, which for a retailer is a transaction and for a B2B firm might be a six-figure pipeline entry. Third, what does the brand click cost, which is usually very little, because quality scores on your own name are as good as quality scores get.
Where the organic slot is clean and unchallenged, pausing often costs almost nothing and the skeptics are right; you were paying a toll on traffic that was already yours. Where competitors camp on your name, or retail partners outrank you, or the query is really a navigation shortcut for a high-value customer, the cheap brand click is the best-priced insurance in the account. Both situations are common. Neither is permanent. The error is assuming yours is whichever one you read about last.
The only honest test
The only honest way to know which situation you are in is a holdout. Turn brand off in a set of geographies, leave it on elsewhere, run it long enough to matter, and count total brand-driven visits and revenue, paid plus organic together, treated as one pool. The design detail that decides whether the test means anything is that combined count: measure paid and organic separately and you will watch organic clicks rise where the ad went dark and conclude whatever you already believed. The question was never whether organic picks up some of the slack. It always does. The question is how much leaks to competitors, resellers, and abandonment in the handoff, and only the combined number answers it. Anything else is a dashboard grading its own homework, and brand campaigns are where self-grading flatters hardest, since the platform happily claims customers who typed your name on purpose. I have run this test for clients where the answer was cancel the spend, and for clients where the leak to conquesting competitors paid for the entire year's brand budget in one quarter. Same test, opposite answers, both correct.
Here is the piece the recurring scandal article always omits: your brand SERP is an auction other people are allowed to enter, which makes brand bidding a game, not a policy. Pause your ads and you have not ended the auction; you have lowered the price of the top slot for everyone who is not you. In competitive categories, your pause is your rival's arbitrage, and they will find it within weeks, because conquesting tools exist to find exactly that. The brands that pause safely are the ones whose name nobody else wants. Take a moment with what that implies before celebrating the savings.
Run the test annually, because every input moves. Competitors arrive and leave, SERP layouts change, AI answer modules push organic results further down the page, and the insurance that was pointless in March can be essential by October. The advertisers who get this wrong are not the ones bidding or the ones abstaining. They are the ones who decided once, years ago, and called the decision a principle.