By Ted Gibson — RippleEffect
There is no split that is right in general, and the published guidance proves it. For a mature brand, Ad Badger says branded spend should end up below 10% of total ad spend, while Olifant Digital puts a scaled account at 20% to 25%. Both are describing the same kind of business. On a £10,000 monthly ad budget that is £1,000 against £2,000 to £2,500; on a $10,000 budget it is $1,000 against $2,000 to $2,500. The two figures cannot both be right, so the question worth asking is narrower: what is your branded spend buying that you would not have got anyway?
The published splits contradict each other
Olifant Digital's table moves with the age of the account. At launch, non-branded takes 60% to 70%, product targeting 20% to 30% and branded 5% to 10%. Through the growth phase it becomes 50% non-branded, 30% product targeting and 15% to 20% branded. At scale, 40% to 45% non-branded, 35% product targeting and 20% to 25% branded. Ad Badger describes brands putting 30%, 40% and even 70% of budget into their own name, and treats getting that figure below 10% as the goal. Neither cites a study, and neither knows your category. It is the same problem as published TACOS benchmarks, where the ranges are wide enough that one number is simultaneously good and bad depending on whose page you landed on.
Branded clicks convert because of who is clicking
Tanveer Abbas at Ecom Brainly puts conversion on your own branded keywords at 25% to 40%, against 10% to 15% on non-branded terms. That gap is the trap rather than the achievement. Someone typing your brand name into Amazon has already decided, so the ad converts beautifully, reports a flattering ACOS and pulls your account average down with it. A good share of that spend is buying a sale organic search would have delivered for nothing.
Branded spend earns its place when it is doing a job you can name. A competitor bidding on your brand and holding the top of the page is a job. So is controlling which of your own products the brand searcher lands on, since a shopper searching your name should reach your hero line rather than the variant you are quietly discontinuing. Holding position in a category where conquesting is normal counts too. If you cannot name the job, the number should come down.
The test that settles it for your account
Run a holdout. Turn branded campaigns off for two to four weeks, keeping clear of promotional periods and anything seasonal, and watch total sales on your brand terms in Search Query Performance rather than ad sales. Ad sales will fall, which proves nothing. If total sales hold, you were paying for demand you already owned. If they drop, you have measured what defence actually costs you, which is a better basis for a budget than a percentage from someone else's table.
If switching off feels reckless, Ad Badger's staged version does the same work more slowly: halve branded spend in the first month, watch search query performance, then step it down again. Olifant Digital's guard rail is worth keeping either way. Above 90% impression share on your branded terms the defence is working and more budget buys nothing; below 80%, raise it until the term is covered, then leave it alone.
We run this as a deliberate split, defending branded search while buying non-branded growth and keeping the two in balance, and it sits among the levers behind our headline case study. If you want to know what your own branded spend is actually buying, separating those two is one of the first things a free account audit makes visible.