Flight log / Transmission 036

Brand bidding: should you pay for your own name?

Paying Google for people who already typed your name feels like tipping the doorman at your own house. Then again, in Google's 2012 analysis of 390 ad pause studies, half the ad clicks that appeared next to a number 1 organic result were still incremental.

Brand bidding is one of the oldest arguments in paid search. One side says branded search campaigns are free money: cheap clicks, high conversion rates, beautiful ROAS. The other side says most of those people would have clicked your organic listing anyway, so you're paying for traffic you already own.

Both sides are right in some accounts. This post covers what the public research actually says, how competitor bidding changes the math, the brand controls Google gives you in Search and Performance Max, and a test that settles the question for your own account.

What is brand bidding?

Brand bidding means paying for ads on searches that contain your company or product name. "Acme," "Acme pricing," "Acme login" and "Acme vs. Globex" are all branded queries. Most advertisers isolate them in a dedicated branded search campaign so they can set separate budgets, bids and reporting.

The flip side is conquesting: bidding on a competitor's brand name. It's legal on Google and common in B2B. Google's trademark policy lists trademarks used as keywords among the things it will not restrict. What it can restrict, after a valid complaint from the trademark owner, is use of the trademark in the ad text itself.

That asymmetry is the heart of the debate. You can't stop competitors from appearing on your name. You can only decide whether you show up too.

Why are branded clicks so cheap?

Brand terms are usually the cheapest clicks in an account, and the reason is mechanical. Google says Ad Rank is calculated from many factors, including your bid and the quality of your ads and landing page. On your own name, your ad and page are about as relevant as an ad can be, and searchers click it at high rates. Competitors bidding on the same term start with a relevance handicap, so they pay more to show below you.

In a hypothetical B2B software account, the spread might look like this.

Cheap isn't the same as valuable. A $1.85 click that would have come through organic for $0 is still a $1.85 cost with no return. The CPC tells you the price. Only incrementality tells you whether you're buying anything.

Are brand search ads actually incremental?

Three bodies of public research frame the debate. None of them is recent, and none is B2B, so read them as directional.

Google's pause studies

In 2011, Google statisticians reported from their "Search Ads Pause" studies that, on average, 89% of the clicks from search ads were incremental, meaning they were not replaced by organic clicks when ads were paused. That figure covers search ads in general, not brand terms.

The 2012 follow-up is the one that matters for brand bidding. Across 390 pause studies, Google split the results by where the advertiser ranked organically for the same query. When the advertiser held the top organic result, which is the normal situation on your own brand name, 50% of ad clicks were still incremental.

Keep in mind who ran those studies, and that they measure clicks, not revenue or pipeline. Still, it's a large sample, and it says the obvious thing: the stronger your organic position, the less a paid ad adds.

The eBay experiment

The counterpoint is a large field experiment at eBay by Tom Blake, Chris Nosko and Steven Tadelis, released as an NBER working paper in 2014. They turned off brand keyword ads in a controlled test and concluded that "brand-keyword ads have no measurable short-term benefits." A later summary in VoxEU put it at 99.5% substitution: almost all the brand traffic simply arrived through organic links instead.

The Edmunds experiment

eBay is an extreme case: a household name that dominated its own results. A geo experiment on Edmunds.com, summarized in VoxEU in 2018 by Lorenzo Coviello, Uri Gneezy and Lorenz Goette, found a very different result. When branded ads were shut off in test markets, only about half of the branded search traffic still reached the site through organic links. The authors suggest much of the rest likely went to competitors who were bidding on the Edmunds name.

Put the three together and a pattern emerges. Brand ads add little when you own your results page outright. They add a lot when someone else is standing on it.

What happens when competitors bid on your name?

This is where most B2B brand bidding decisions are actually made. If a competitor's ad sits above your organic listing on your own name, pausing your brand ad hands them the first thing your prospect sees. Leave the maze empty and the ghosts move in.

You can see this directly by running auction insights on your brand campaign. Any domain with meaningful impression share on your name is conquesting you. In this illustrative example, a competitor ramps up on a brand term over a quarter.

When a competitor goes from showing on 1 in 10 of your brand searches to 4 in 10, the case for keeping your own brand ad live gets much stronger. When nobody else shows up, the case for testing a pause gets stronger. Watch this monthly, because conquesting tends to start quietly.

How do you test whether your brand ads are incremental?

The research gives you a prior. A holdout test gives you your own answer. The cleanest version for most advertisers is a geo split, the same design the Edmunds study used:

1. Pick matched regions. Choose test and control regions with similar historical brand search volume and conversion rates. Several smaller regions beat one big one.
2. Pause brand ads in the test regions only. Keep everything else, including non-brand campaigns, unchanged.
3. Measure total brand traffic, not paid traffic. Track paid brand clicks, organic brand clicks and the outcomes you care about, such as qualified leads or pipeline, in both groups.
4. Run long enough. Plan for at least 4 to 6 weeks, longer if your sales cycle is long. B2B leads are lumpy and a 2-week test will mostly measure noise.
5. Compare the gap, not the totals. The incremental effect is the difference between test and control, relative to their pre-test baseline.

Here's what a hypothetical test region might look like week by week.

In this illustrative case, organic picks up about 70% of the clicks the ads used to get, and total brand clicks fall by roughly 13%. Whether that's worth paying for depends on what those lost clicks were worth and what the brand campaign cost. If a competitor was conquesting the term, expect a bigger gap. For test design and reading results without fooling yourself, see our guide to incrementality testing.

How should you set up a branded search campaign?

If the answer is "keep bidding," set it up so brand traffic doesn't leak into the rest of the account and the rest of the account doesn't take credit for it.

The controls Google gives you, as of October 2026:

• Keep brand out of non-brand campaigns. Add your brand terms as negative keywords in non-brand Search campaigns, so your category campaigns are judged on new demand.
• Exclude your brand from Performance Max. Google lets you add your own brand to a brand exclusion list on a Performance Max campaign. There, exclusions apply to Search, Shopping and YouTube search inventory, and Google says they also block brand misspellings and other variants.
• Use exact match for core brand terms. Google says a Search campaign keyword that exactly matches the query, or is spell-corrected to it, is preferred over Performance Max.
• Consider brand inclusions if you want broad match on brand. Brand inclusions, formerly called brand restrictions, limit a Search campaign to queries about the brands you select. Google's June 2024 announcement says they're only available with the broad match campaign setting, and starting May 27, 2025 Google began moving brand settings for Search campaigns into AI Max.

Last, the landing page. Branded traffic is high intent but not one intent. "Acme pricing," "Acme login" and "Acme vs. Globex" want three different pages, and the homepage is rarely the best answer to any of them. Send each brand plus modifier group to the page that answers it, using the same discipline as any message match audit.

What we don't know yet

• The research is old and consumer-heavy. The main public studies date from 2011 to 2018 and come from marketplaces and consumer sites. The results page has changed a lot since, with more ad slots and AI-generated answers, and there's no comparable public study for B2B software.
• Clicks aren't pipeline. Google's studies measure clicks. A B2B brand test should measure qualified opportunities, which take weeks to show up and need larger samples.
• Short tests miss slow effects. The eBay finding is explicitly about short-term benefit. Whether a missing brand ad affects perception over months isn't something a 6-week test can see.
• Brand lists depend on Google's library. Exclusions and inclusions work from Google's brand library. Smaller brands may not be listed yet and need to request an addition.
• Competitors react. Pause your brand ads and a competitor's ad may get cheaper and more visible on your name. A test captures that only if it runs long enough for them to notice.

Brand bidding is one line item in a wider plan. For how it fits next to category, competitor and problem-aware terms, see the B2B PPC playbook.

FAQ