Across 170+ Amazon brands and $29M+ in managed ad spend, the single most common self-inflicted PPC misdiagnosis we see is the same: an account with a “healthy” 22% blended ACOS that’s actually bleeding on non-branded discovery — but the branded terms mixed into the same campaigns average the number out to look fine. The dashboard says healthy. The dashboard is wrong.
This post is the diagnostic that separates the two — why branded search hides weak non-branded performance, the 14-day Branded Spend Dial-Down Test that surfaces the real number, and the isolation architecture that stops the leak from coming back. Not “stop bidding on your brand.” A more specific move: manage branded and non-branded as two different problems, because they are.
What’s the difference between branded and non-branded search terms?
The distinction sounds obvious, but Amazon’s search term report doesn’t label them. Every branded term looks like every other search term in the CSV — just another row with clicks and orders. Unless you actively separate them, they blend into your campaign metrics and the strong branded numbers cover for the weak non-branded ones.
For the broader framework on how search terms feed your account, the complete Amazon search terms guide is the hub. This post is the branded-specific deep-dive that most PPC audits skip.
Why does branded PPC make everything look healthier?
Here’s what the pattern looks like in a mixed campaign:
| Term type | Clicks | Orders | Spend | Revenue | ACOS |
|---|---|---|---|---|---|
| Branded (“YourBrand collagen powder”) | 200 | 60 | $200 | $2,400 | 8% |
| Non-branded (“collagen powder for women”) | 800 | 40 | $1,200 | $1,600 | 75% |
| Campaign blended | 1,000 | 100 | $1,400 | $4,000 | 35% |
The 35% blended ACOS looks acceptable. The 75% non-branded ACOS is bleeding. Sellers who don’t split branded from non-branded never see the second row — they see the campaign summary and assume the account is fine.
What is the Branded Spend Dial-Down Test?
The test is designed to be reversible — 14 days is short enough that you can turn spend back up if revenue drops, and long enough that day-of-week and weekly noise average out. The 80% cut (rather than zero) preserves enough branded presence to defend against competitor conquest ads showing up on your brand terms.
Here’s the sequence:
- Isolate branded into its own campaign if it isn’t already (see below).
- Baseline — record total brand revenue for the prior 14 days.
- Cut branded spend to 20% of baseline for the next 14 days. Keep everything else — non-branded, Sponsored Brands, Sponsored Display, external traffic — exactly the same.
- Measure total brand revenue for the 14-day window.
- Compare. If revenue is within ±3% of the baseline, most of that branded spend was buying sales you already owned. If revenue drops 8%+, some meaningful portion was incremental.
Across the brands we’ve audited, the test recovers 8–15% of total ad spend on average. Some brands find 20%+ of their branded ad spend was pure overattribution.
How should branded and non-branded campaigns be structured?
The architecture matters because the two campaign types are solving different problems. Branded is a defense play — hold top-of-search share against competitor conquest ads and make sure shoppers who search for your brand find you first. Non-branded discovery is a growth play — find new shoppers who don’t know you yet and win them profitably.
The structure:
| Campaign | Purpose | Keywords | Match type | ACOS target |
|---|---|---|---|---|
| Brand Defense (SP) | Hold top-of-search on your brand | Brand name + variants + common misspellings | Exact | 10–20% (defensive) |
| Non-branded Discovery (SP) | Find new shoppers | Category + attribute + problem queries | Broad / Phrase / Exact | At or below SKU contribution margin |
| Competitor Conquest (SP) | Steal share from adjacent brands | Competitor brand terms | Exact (careful ACOS management) | Break-even to defensive |
| Sponsored Brands / Display | Awareness + retargeting | Broader themes | Product/audience targeting | Longer payback window |
The wasted PPC spend audit covers the isolation move as filter #4 — it’s often where the biggest single chunk of recoverable spend sits in a first audit.
What ACOS targets should apply to branded vs. non-branded?
The mistake most sellers make is applying one blended ACOS target across both. If you set a 25% target for the whole account, you’ll under-invest in branded defense (because branded routinely runs 8–15% ACOS) and over-invest in non-branded discovery (because forcing 25% ACOS on non-branded usually means bidding above what the incremental math supports).
For the deeper framework on ACOS-to-margin math, how to reduce Amazon ACOS without losing sales volume walks through the bid-to-margin logic and the branded piece together.
When is branded PPC actually worth the spend?
The decision checklist per brand term:
- Are competitors running ads on your brand term? → Yes: bid defensively. No: consider dial-down.
- What’s your organic top-of-search share on the brand term? → 80%+: consider dial-down. Under 60%: bid to defend.
- Are you in launch or category-climb? → Yes: bid to protect velocity. No: efficiency-first.
- Does the brand term have upsell/cross-sell potential (Sponsored Brands, video)? → Yes: consider Sponsored Brands with branded video, not just Sponsored Products. No: minimal SP defense.
The test isn’t “should I stop bidding on my brand?” — it’s “how much of what I’m spending is actually buying incremental sales?” Some branded spend is always worth it. Most brands overspend on it and don’t realize until they run the Dial-Down Test.
Where does this fit inside The Profit-Leak Method?
The full framework at /playbook/profit-leak-method/ sequences this correctly: search-term hygiene first (fastest, lowest-risk), branded isolation second (biggest single hidden waste), then the four remaining leaks in order. Reversing that sequence usually finds smaller wins because the branded number is easier to see once the search-term data is clean.
Frequently asked questions
What’s the difference between branded and non-branded Amazon search terms?
A branded search term contains your brand name (or a common misspelling). A non-branded search term is a category query without your brand — the shopper doesn’t know you yet. Branded converts at 25–40% because intent is high; non-branded converts at 5–15% depending on category and creative quality.
Should I bid on my own brand terms in Amazon PPC?
Some yes, in a dedicated brand-defense campaign — enough to hold top-of-search share against competitor conquest ads. But most established brands overpay for branded search because the sales would happen organically anyway. The Branded Spend Dial-Down Test reveals the actual incremental portion.
How much branded ad spend is realistic to save?
For established brands, the Branded Spend Dial-Down Test typically recovers 8–15% of total ad spend without losing revenue. Some brands find 20%+ of their branded spend was pure overattribution. The number depends on brand strength and how much competitor conquest activity is happening in your category.
What’s the Branded Spend Dial-Down Test?
A 14-day experiment: cut branded ad spend by 80% (not to zero — keep defense), hold non-branded and everything else steady, and measure total brand revenue against the prior 14 days. If revenue holds within noise, that branded spend was buying sales you already owned. If it drops, some portion was genuinely incremental.
Why does mixing branded and non-branded in one campaign cause problems?
The strong branded ACOS averages out weak non-branded terms in the same campaign. You look at a healthy 22% campaign ACOS and never notice the non-branded search terms inside are running at 60%+. Isolation separates the two signals so each can be managed to its own target.
How do I isolate branded queries into their own campaign?
Create a brand-defense Sponsored Products campaign with your brand terms as exact-match keywords. Set a defensive budget (usually 5–10% of total ad spend). Then add every branded term and common misspelling as negative exact in every non-branded discovery campaign so they can never leak back in.
The bottom line
Branded search terms aren’t the problem — mixing them into non-branded discovery campaigns is. The strong branded ACOS covers for bleeding non-branded terms and the account dashboard looks healthy while 8–15% of total spend leaks quietly. Isolate branded into a defense campaign, negate it out of discovery, and run the 14-day Dial-Down Test to see which portion of your branded spend was ever incremental in the first place. Most established brands find the answer smaller than they expected.
That’s the branded-overattribution piece of the ad-waste leak inside The Profit-Leak Method — usually the second-biggest chunk of recoverable margin after search-term hygiene, and always hiding in plain sight in the campaign summary.
Want us to run the Branded Spend Dial-Down Test on your account? Get a free Amazon PPC audit — we’ll separate branded from non-branded, quantify the overattribution, and design the 14-day test with your team.