.png)
Most people who bid on their own brand name have never run the test that would tell them whether they should. They just do it because a rep told them to, or because a competitor showed up above their listing once and gave someone a heart attack. Then they report a 900% ROAS on that campaign, everyone nods, and nobody asks the only question that matters: how many of those buyers would have clicked the free organic result sitting one line below the ad.
That question is the whole debate. Brand term bidding is buying ads on searches for your own company name, or for a competitor's. It sounds trivial and it isn't, because the money involved is real and the reported returns are almost always inflated by clicks you'd have gotten for nothing. I spent years managing accounts where the brand campaign was the crown jewel of the monthly report and, quietly, the least defensible line item in the account. I've since changed my mind on parts of this. I'll tell you where.
Here's the short version before I earn it: bidding on your own brand is worth it when someone is stealing your clicks or when the SERP is genuinely competitive, and it's mostly a waste when you already own the page and nobody's bidding against you. Conquesting competitors can pay, but the CPCs are ugly and the trademark rules are stricter than most people realise. And you cannot know which situation you're in without a pause test. The rest of this is how to actually run one, and how to read the number at the end without lying to yourself.
Start with the incrementality problem, because everything else follows from it. When someone Googles your exact brand name, they already know you exist. They typed you in on purpose. If you show a paid ad and they click it, Google charges you for a visitor who, in most cases, was going to reach your site anyway through the organic listing directly beneath. That's the tax nobody puts in the deck: you're paying a per-click fee to intercept your own free traffic. The reported ROAS looks spectacular precisely because those people were always going to convert. The ad didn't cause the sale. It just took credit for it.
So the real question isn't 'does the brand campaign have a good ROAS.' It always will. The question is: what percentage of those clicks are genuinely additional, and what percentage would have happened for free. That number decides whether you're investing or lighting money on fire with a very impressive-looking receipt.
There are three situations where brand defense is genuinely worth the spend, and they're specific. First, when a competitor is actively bidding on your name. If someone else's ad sits above your organic result, the top of the page belongs to them until you show up too. Second, when the SERP around your brand is crowded, review sites, comparison pages, marketplaces, aggregators, so your organic listing gets pushed down and the ad is the only thing guaranteeing you the top slot. Third, when you're running a promo or a specific offer you want controlling the message, because the ad headline and sitelinks let you say things your organic title tag can't. Outside those three, you're usually paying to be handed traffic you already had.
The flip side is the case almost nobody wants to hear. If you already dominate the SERP for your brand, meaning your organic listing is the first result, it takes up half the screen with sitelinks, and no competitor is bidding against you, the brand campaign is probably a rounding-error waste. You're paying for clicks that would land on the free result an inch below. I've seen accounts spending $2,000 to $4,000 a month on a brand campaign whose entire contribution was cannibalising organic. Pause it and total sales don't move. That's not a hypothetical. That's the most common result of the test I'm about to describe.
The uncomfortable part is that the brand campaign usually shows the best numbers in the whole account, so it's the last thing anyone wants to touch. I used to defend those campaigns to clients on exactly that logic. I was wrong. A high ROAS on brand terms tells you almost nothing about incrementality. It tells you people who search your name tend to buy, which you already knew.
The only honest way to settle this is to turn the campaign off and watch what happens to total conversions, not the campaign's own conversions. Here's the method I use. Pick a clean two-to-four week window with no promos, no seasonality spikes, no PR. Pause the brand campaign entirely. Track total brand-driven conversions across paid and organic combined, from both Search Console and your analytics, against the same window before you paused. If total conversions hold roughly flat, organic simply absorbed the traffic and the campaign was a tax. If total conversions drop, the paid clicks were genuinely incremental and the gap is the real value the campaign was adding. The number you care about is the delta in the combined total, never the ROAS the paused campaign was reporting.
If you operate in multiple regions, a geo split is cleaner and lower-risk than an all-or-nothing pause. Keep the brand campaign live in half your markets, pause it in a matched half, and compare total brand conversions per region over the same weeks. You get a control group without betting the entire account on one window. This is the same discipline I'd apply to testing ad copy: never trust a before-and-after when you can run a real control alongside it. The reason most people skip it is that it takes two to four weeks and produces an answer they might not like.
One practical warning on the pause test. If a competitor is bidding on your name, do not run a naked pause, because you'll hand them the top of the page for a month and your test will just measure how much traffic they steal. In that case the geo split is the only safe design, or you accept that the presence of a competitor already answers the question and you keep defending. The test is for the quiet SERPs, not the contested ones.
Takeaway: run the pause or geo test once a year on every brand campaign you own. The answer changes when a competitor enters or exits the auction, so it's not a one-time decision.
Conquesting is the other half of this, and it's where people get either too timid or too reckless. Bidding on a competitor's brand name is legal in the US and most markets. You're allowed to target their name as a keyword. What you're generally not allowed to do is use their trademark in your ad copy, in the headline, description, or display path. Google's policy is roughly: you can bid on the term, you can't write 'Better than [Competitor]' with their trademarked name in the ad, and if they file a trademark complaint Google will usually disallow their name appearing in your text. Reseller and comparison-site exceptions exist, but if you sell a directly competing product, assume you cannot put their name in the copy. Bid on the term, sell your own benefits, name nobody.
Here's the mechanic that makes conquesting expensive. Your landing page and ad have nothing to do with the competitor's name, so your ad relevance and expected click-through rate are both weak for that keyword. That tanks your Quality Score on the term, and a low Quality Score means you pay more per click for a worse position. So you're bidding on a search where the user was specifically looking for someone else, paying a premium because Google can tell your ad is a poor match, and converting a fraction of them because they wanted the other guy. The CPCs on competitor terms routinely run two to four times your own brand CPC for a conversion rate that's a fraction of it.
It pays in a narrow set of cases. When your product has a concrete, provable advantage the searcher would switch for, cheaper, faster, no contract, a feature they lack. When the competitor's customers churn often and are actively shopping. When the lifetime value is high enough that a brutal cost per acquisition still nets out, think B2B software or anything with a long retention tail. When it burns budget is the opposite: low-margin products, one-off purchases, or a market where you and the competitor are basically interchangeable and the searcher has no reason to defect. If you can't name the specific reason someone would abandon the brand they just typed in, don't bid on it.
Before you decide whether to defend, find out if you actually need to. Google's Auction Insights report, run on your brand campaign or a brand keyword, shows you exactly which domains are showing up in the same auctions and how often. That's your evidence. If Auction Insights shows no competitors with meaningful impression share on your name, that's a strong signal your defense budget is optional. If it shows a rival sitting on your terms 40% of the time, that's your answer too. Check it monthly, because competitors switch conquesting on and off and you want to know the day they start.
If you want something more continuous than a manual weekly check, brand-monitoring tools and rank trackers will alert you when a new advertiser appears on your terms. Useful, but honestly Auction Insights covers most of what a small or mid-size account needs for free. Don't buy a monitoring subscription to solve a problem a built-in report already shows you.
Takeaway: never carry a permanent brand defense budget on faith. Tie it to what Auction Insights shows, and let it scale up or down with the actual threat.
The deeper problem behind all of this is reporting. When your brand and non-brand campaigns get blended into one account-level ROAS, brand cannibalisation hides in the average. Brand terms convert at a huge rate and a low cost, so they drag the blended number up and make your genuinely acquiring campaigns, the non-brand ones doing the actual work of finding new customers, look worse by comparison. You end up cutting the campaigns that grow the business to protect the ones that just harvest people who already knew you. I've watched that exact mistake get made in a QBR, with everyone in the room congratulating themselves on it.
The fix is boring and non-negotiable: segment brand and non-brand into separate campaigns, report them separately, and set separate targets. Judge non-brand on cost per new customer. Judge brand on the incrementality delta from your pause test, not its face-value ROAS. Once they're split, the cannibalisation stops hiding and you can size each budget for what it actually does. A blended report is how agencies keep a cannibalising brand campaign in the account for years without anyone noticing.
This is genuinely one of the things automation does better than a human doing it by hand at 1am. The disclosure: I work on product at groas, so weigh that. But the mechanic is straightforward and it's the same logic I applied manually for years. A system trained on how brand and non-brand searches behave separates them structurally, caps brand spend to what's incremental rather than letting it swell to whatever volume the terms will absorb, and keeps the reporting split so the blended-average trap never forms. The groas engine is built to run those bid and budget calls continuously instead of when someone remembers to audit them, which is the part that matters, because brand cannibalisation isn't a one-time cleanup. It creeps back the moment you stop watching.
Do brand campaigns inflate my reported ROAS? Almost always, yes. Brand terms convert cheaply and at high rates because those people already decided to buy from you, so the campaign posts a ROAS that flatters the whole account. That's not fraud, it's just credit misattribution. The ad takes the sale that organic would have delivered for free. If your headline ROAS number looks great and you've never separated brand from non-brand, the number is partly fiction. Split them and look at the non-brand figure alone. That's the one telling you whether your advertising actually acquires customers.
Can competitors use my brand name in their ads? They can bid on it as a keyword, which is legal. They generally cannot put your trademarked name in their ad copy, and if they do, you can file a trademark complaint with Google to have it removed. Filing the complaint is free and usually the more effective move than escalating your own bids in a spite war you'll both lose money on. Check Auction Insights, see who's actually there, and deal with the copy violation through the complaint process before you throw budget at the problem.
So the whole thing comes down to one habit most accounts never build: refusing to trust the brand campaign's own numbers. Run the pause test. Read Auction Insights. Split the reporting. Decide with evidence instead of the reflex that a competitor might show up someday. Defend your name when someone's actually taking it, conquest only when you can name the reason a customer would switch, and the rest of the time keep the money. The brand campaign that survives all three of those checks is one you can finally stop apologising for in the monthly report.
