BIDDING ON COMPETITOR KEYWORDS IN GOOGLE ADS

I get asked whether bidding on competitor keywords google ads is worth the money more than almost any other tactical question, and for a long time I answered from theory. Then I decided to stop guessing and run it properly: 60 days, a real account, competitor terms isolated and measured against everything else. This article is what that test actually returned, not what the case studies promise. If you want the policy rules first, the cornerstone piece on whether you can use competitor brand keywords lays out exactly where Google draws the line; this page is the field report that sits on top of it. The short version up front: google bidding on competitor keywords works, but only as a disciplined, isolated experiment, and the average number will lie to you unless you split it apart.

The 60-day numbers, honestly

Here's what I saw. Across the full run, competitor keywords cost roughly 2.5x my brand-term CPC and about double my generic high-intent CPC, and they converted at close to a third of the rate of my own terms. On a blended basis the CPA looked bad enough that if I'd only glanced at the campaign summary, I'd have killed it on day 20 and written the whole idea off. That blended figure is exactly the trap.

When I broke the campaign apart by rival, the picture changed completely. I was bidding on four competitors, one ad group each. One competitor's terms were quietly profitable the entire time, sitting at a CPA I'd have been happy with on generic traffic. Three were burning money at a rate that only got worse as I gave them room to spend. Averaged together they cancelled into a mediocre-looking loss; separated, they told me precisely which door to keep open and which three to shut. That segmentation lesson is the single most valuable thing the test gave me, and it's why I now never run competitor terms as one undifferentiated bucket.

Quality score behaved exactly as the mechanism predicts. On a rival's brand name you are, almost by definition, less relevant than the rival, so their landing page and ad win the relevance signal on that query. My quality scores on competitor keywords sat low across the board, which is what pushes the CPC premium in the first place. If you want to understand why that gap is structural rather than fixable, the broader mechanics of running competitor keywords go into how intent and relevance interact on these terms.

When competitor bidding actually pays

The one profitable rival in my test wasn't luck, it fit a pattern I've since seen repeat. Competitor bidding pays when three things line up: the customer value is high enough to absorb a premium CPC, you have a genuinely stronger or clearly differentiated offer to convert the switch, and the searcher is comparison-shopping rather than loyalty-searching. The profitable competitor sat in a high-ticket category where a single won customer justified weeks of expensive clicks, and my client's offer beat theirs on a dimension buyers cared about. When even one of those three legs is missing, the math collapses fast.

The inference I keep landing on is that competitor terms reward patience and a sharp landing page far more than they reward budget. The mistake I see advertisers make is treating them as a growth lever you just pour money into. They aren't. They're a scalpel: precise, deliberate, and dangerous when swung like a hammer. If your lifetime value is thin or your offer isn't obviously better than the rival's, you're paying a premium to interrupt a decision that's already been made, and you'll lose that trade almost every time.

There's also a defensive case that's easy to overlook. In categories where everyone is already bidding on everyone, sitting out isn't neutral, it's ceding ground. Sometimes I run competitor terms not because they're efficient in isolation but because absence would cost more than the premium. That only holds when you've measured it, though, which is why the setup below matters as much as the decision to run.

The three real risks

First, trademark. The keyword bid itself is safe: Google's trademark policy explicitly states it will not restrict using trademarks as keywords. You can read the exact criteria on Google's own Trademarks policy page. The risk lives entirely in the ad text: put a rival's trademark in your headline or description as a direct competitor and you've handed them a valid complaint. Even then, enforcement isn't a hair-trigger, Google's policy states violations don't lead to immediate suspension without a prior warning issued at least seven days out. Across the whole test I never had a compliant keyword bid penalized. Keep the name out of the copy and this risk is manageable.

Second, the CPC premium. This one is guaranteed, not hypothetical. You will pay 2x to 3x your efficient CPC, and if competitor terms share a budget with your good campaigns, they'll quietly cannibalize spend from the efficient side. That's a budget-structure problem, and it's solvable, which is the whole point of isolating them.

Third, quality score on someone else's brand. It stays low no matter what you do, because relevance to a rival's name is something you can't fully win. Accept it as a fixed cost of entry rather than something to optimize away. Once you stop fighting it and price it in, competitor bidding becomes a clear-eyed economic decision instead of a frustrating one. If you're weighing this against defending your own turf, my read on bidding on your own branded keywords covers the other half of the equation.

How I set it up without burning budget

Isolation is the whole game. Competitor terms get their own campaign, their own budget cap, and a dedicated ad group per rival so I can read each one's economics independently, exactly the split that saved my test from a false verdict. I never fold them into an ad group with generic intent keywords, because their low quality score and high CPC drag the group's blended signal and hide the true cost inside an average that looks fine while it bleeds.

Match type is the next guardrail. I keep competitor terms on exact and phrase match only, never broad, because a rival's brand name on a loose match type reaches far too wide and racks up premium clicks on queries that have nothing to do with a real switch. Understanding how the match layer governs that reach is worth the detour into the match types explainer before you launch, since it's the difference between a controlled test and an expensive one.

On bidding, I start manual or with a tightly capped tROAS so the algorithm can't chase volume on terms I already know run expensive. The landing page is built to convert the switch, not a generic homepage, it has to answer "why leave them for you" in the first screen. And I watch it daily for the first two weeks, then rival by rival after that. If a competitor's ad group doesn't earn out, I pause that one ad group and leave the rest running, which is only possible because the structure kept them separate. That single discipline, isolate, measure per rival, cut the losers fast, is what turns competitor bidding from a budget leak into a controlled bet.

Put it all together and the answer to "is it worth it" stops being a yes or no. Bidding on competitor keywords in Google Ads is worth it exactly when the value, the offer, and the intent align, and only when you've built the structure to see that alignment clearly. Run it isolated, read it per rival, keep the winners on a short leash, defend your own brand at the same time, and the premium becomes a price you pay deliberately rather than a wound you don't notice until the month closes.