In short
Most stores should, but for defence and control rather than because the reported return looks good. Brand campaigns almost always show the highest ROAS in the account and most of that revenue would have arrived anyway, so the honest question is what share of it is genuinely incremental. Bid on your name when other sellers, marketplaces or competitors occupy the results page, when you need to control the message during a launch or a sale, or when Shopping listings from resellers sit above your own. Then test it with a holdout rather than trusting the ratio.
The reported return on brand traffic is not the return you are getting
People searching your name have already chosen you. When they click a paid listing at the top of the page, the ads account records a conversion at a very low cost, and the campaign reports a return that no other campaign can match. That number is real as a measurement and misleading as a decision input, because a share of those shoppers would have scrolled to the organic result and bought anyway.
What you are actually buying is the difference: the orders that happen because the paid listing was there and would not have happened otherwise. That difference is usually much smaller than the reported revenue, and it varies enormously between stores depending on who else is on the page.
This is why brand spend should be judged on its own and never pooled into an account level ROAS. A blended figure that includes brand looks healthy while the campaigns that actually acquire customers are underfunded, and it moves whenever brand demand moves, which has nothing to do with how well the advertising is working.
The cases where brand bidding clearly earns its budget
Bid when someone else is on the page. Competitors are allowed to bid on your brand name as a keyword in most markets, and so are the marketplaces and stockists who sell your product at a different price. If the top of the results carries three listings before your organic one, the paid listing is buying back a click you would otherwise share.
Bid when you sell through resellers. A shopper searching your name and seeing your product on a marketplace listing usually buys there, and the order arrives at wholesale margin instead of direct margin. The difference in contribution per order is frequently larger than the entire cost of the brand campaign.
Bid when you need control of the message. A sale, a launch, a stock recall, a shipping delay, a rebrand: in each case a paid listing lets you say something specific at the top of the page immediately, with sitelinks pointing at the exact pages, while organic results take weeks to catch up.
Bid on brand plus product queries in particular. Someone searching your brand with a product type is close to buying a specific thing, and a listing that takes them straight to that collection converts better than a homepage result. These are also the queries where a reseller listing does the most damage.
The cases where brand spend is mostly buying its own traffic
If nobody else appears on the page, the paid click is largely replacing a free one. A distinctive brand name with no competitor bidding, no marketplace presence and a strong organic result for the store's own name is the clearest example. The paid listing sits above an organic listing that would have collected the same shopper.
A very high organic share on brand queries points the same way. Search Console will show you the clicks and position for queries containing your name, and if you already hold the top result with a healthy click share, incremental value is limited to whatever the ad adds in message control.
Cost matters here too. Brand keywords are usually inexpensive because the landing page is highly relevant, so even low incrementality can be worth it. Run the arithmetic rather than the principle: at a small spend, buying insurance against a competitor appearing next month is a reasonable use of budget, and at a large spend it needs proof.
Test it with a holdout, and stop Performance Max absorbing the traffic
The only reliable test is turning it off in a controlled way. Split by geography rather than by time where you can: pause the brand campaign in a set of regions, keep it running in a comparable set, and compare total revenue from those regions, not campaign revenue. Give it long enough to cover a full purchase cycle and avoid running the test across a sale.
If a geographic split is not practical, alternate whole weeks on and off for a run of several weeks and compare total store revenue and total brand search demand in each state. It is a coarser test and it still beats reading the campaign ROAS.
Whichever you run, watch the search terms in the rest of the account first. Performance Max and broad match search campaigns will pick up brand queries unless you stop them, which means a brand campaign you paused can simply be replaced by brand traffic served from somewhere else, and the test measures nothing. Use brand exclusions and the account level negative keyword list to keep brand traffic in one place before you begin.
Write down the decision rule before the test starts: the size of the revenue difference that would make you keep the campaign, and the size that would make you cut it. Deciding afterwards is how tests get read to confirm whatever was already believed.
A regional holdout, laid out
- Test design
- Brand campaign paused in half the regions, on in the rest
- Duration
- Four weeks, no sale in the window
- Measured
- Total store revenue from each region group
- Also measured
- Brand query volume, to confirm demand was level
- Ignored
- Brand campaign ROAS in the ads account
- Decision rule
- Set before the test, in dollars of revenue difference
Illustrative design rather than results. The value of the method is that it measures the store, so it captures orders the paused regions still placed through organic search.
Related questions
Can competitors legally bid on my brand name?
Bidding on a brand name as a keyword is generally permitted, while using a trademark in the ad text is restricted and can be reported to Google through its trademark process. The practical response is usually to hold your own listing rather than to chase the complaint.
Should Performance Max be allowed to serve on brand searches?
Usually not, if you want to read the account clearly. Applying brand exclusions and account level negatives keeps brand demand in a campaign you control, stops it inflating the return on your acquisition campaigns, and makes any test of brand spend meaningful.
Does bidding on my own name raise what I pay per click?
Bidding on your own brand does not raise your own costs on its own. Prices rise when other advertisers enter the auction. Your landing page and ads are highly relevant to your brand query, which normally keeps your cost per click well below what an outside bidder pays for the same term.