Branded paid search creates one of the most uncomfortable budget questions in marketing: are we buying new demand, defending against competitors or paying again for customers who were already looking for us?
A fresh Search Engine Land article published on August 17 gives the question useful numbers. In the case described, a company paused roughly $113,000 per month in paid search across four major markets. By week 13, organic and direct revenue had recaptured a reported 65% of the paid-attributed revenue. In one branded campaign analysis, only 10.9% of spend appeared incremental while 89.1% looked like brand defense.
Why attribution makes the question hard
The platform report will usually credit the ad that received the click. That does not answer whether the company needed to buy that click. A user who searches the brand name and clicks the ad may be incremental if a competitor would otherwise win, if the ad changes trust or if the organic result is weak. The same click may be defensive or cannibalized if organic already held the answer and the customer had high intent.
This is why old studies can seem contradictory. Google research found many ad clicks incremental on average, while eBay’s famous branded-search experiment suggested a strong brand could recover much of the demand organically. Both can be true because the answer changes by brand strength, query type, competitor pressure, SERP layout, organic rank and conversion behavior.
A practical overlap model
Start by joining Google Ads search terms with Search Console queries for the same period. Then classify each query group into three buckets. Defensive spend protects the brand from competitors or SERP risk. Dependent spend helps demand that weakens materially when ads disappear. Incremental spend reaches users organic could not have captured.
The mistake is treating all brand search as one bucket. A pure brand-name query, a brand plus category query, a store-near-me query and a competitor comparison query do not carry the same risk. The test design should respect that difference.
How to run the test
- Choose query groups where organic rank, competitor pressure and commercial value are known.
- Pause or reduce spend in a controlled market, geography, device segment or query cluster.
- Annotate the start date and keep the test long enough for conversion lag.
- Measure organic clicks, direct traffic, total revenue, margin and competitor movement, not only paid-attributed revenue.
- Decide whether the result justifies a full cut, a smaller defensive budget or a different bid strategy.
What not to cut blindly
Do not remove paid search where organic rankings are unstable, competitors are aggressive, the brand is new, retailers or marketplaces crowd the result, or paid traffic converts at a materially higher margin. Also avoid a holiday-season test unless the business is comfortable with the risk. Incrementality testing is a business experiment, not a symbolic gesture against platform costs.
The better outcome may be a priced defense: keep the campaigns that protect margin or high-value queries and reduce the campaigns that mostly collect credit for demand already won elsewhere.
The CMO takeaway
The right question is not “is branded search good or bad?” The right question is “what are we paying it to do?” A channel report answers who received the click. An incrementality test answers whether the business needed to buy it.
Use the test to price the defense, not to win an argument between SEO and PPC. If organic can carry a query without material profit loss, budget can move. If paid search protects profitable demand, keep it. Either way, the decision becomes a measured operating choice rather than a habit.
