New-To-Brand Rates By Ad Type
New-to-brand rates rise sharply as you move up the funnel. Sponsored Products typically bring 10–20% new customers; Sponsored Brands video runs far higher; DSP prospecting is the highest of all. In our own accounts, DSP prospecting audiences run about two-thirds new-to-brand.
Why does new-to-brand matter more than ROAS?
ROAS tells you what an ad returned. New-to-brand tells you whether it grew anything. A campaign at 8x ROAS composed entirely of existing customers is a discount programme wearing an advertising costume — profitable on paper, inert as growth.
This is the crux of the incrementality objection every brand raises, and it is the objection we hear most often in sales conversations. The honest version of the answer is not a bigger ROAS number, it is a new-to-brand share plus a measurable change in organic demand.
The clearest demonstration we have is a fitness-publishing account where organic order share moved from 13.5% before DSP to 27.9% after, and kept climbing for four weeks after all DSP spend stopped. Advertising-attributed cost of sale compressed from 48% to 37% and held. That is what buying new customers looks like downstream.
What are the benchmark rates by ad type?
The table below has two kinds of rows and we label them, because the difference matters. The DSP prospecting figure is ours, measured across accounts we run. The by-ad-type bands were synthesised by our team from published Amazon case studies and vendor benchmarks — they are a useful frame, not our measurement, and we cite them that way.
How should you read these numbers against your own account?
Compare shape before level. If your Sponsored Products new-to-brand rate is 30%, you are probably under-invested in branded defence or genuinely early in your category. If your DSP prospecting rate is 20%, your audiences are not prospecting — they are retargeting with a different label.
The measurement caveat worth stating: retargeting new-to-brand rates look low because that is the job. Retargeting is supposed to convert people who already met you. Judging a retargeting line on new-to-brand share is the same category error as judging a prospecting line on first-order ROAS.
- Prospecting lines: judge on new-to-brand share and downstream organic lift.
- Retargeting lines: judge on ROAS and frequency, not new-to-brand.
- Streaming TV: judge on video completion rate and branded-search lift, with roughly 1.0 ROAS as the bar.
- Everything: read on four-to-eight-week windows, because attributed sales keep moving after the fact.
What does it cost to buy a new customer this way?
Cheaper than the search auction, in our accounts, by an order of magnitude. Display and online-video CPCs on DSP have run between $0.21 and $0.44 in our managed programmes, against category search CPCs in the $5.00–$6.60 range.
One account moved from ROAS 10 in month one at $0.21 CPCs to ROAS 17.68 by month four on 12,000-plus units. Another reached 4x ROAS within three weeks on $6,000 of spend during a Prime week that pushed CPCs up across the whole platform.
That gap between upper-funnel and search-auction pricing is the entire structural argument for running both. See DSP prospecting for how the audiences are built.
| Ad type | New-to-brand share | Sourcing |
|---|---|---|
| Sponsored Products | 10–20% | Team synthesis from published Amazon case studies and vendor benchmarks |
| Sponsored Brands headline — branded terms | 5–15% | Team synthesis from published Amazon case studies and vendor benchmarks |
| Sponsored Brands headline — non-brand terms | 20–30% | Team synthesis from published Amazon case studies and vendor benchmarks |
| Sponsored Brands video | 40–60% | Team synthesis from published Amazon case studies and vendor benchmarks |
| DSP retargeting | 10–20% | Team synthesis from published Amazon case studies and vendor benchmarks |
| DSP prospecting — benchmark band | 50–70% | Team synthesis from published Amazon case studies and vendor benchmarks |
| DSP prospecting — our measured figure | ≈66% | First-party, across our managed DSP portfolio |
| DSP in one recovery-device account | 80% | First-party, single account, reached ~2.5 ROAS by week three |
Add new-to-brand share to every weekly report next to ROAS. If you cannot say what percentage of last month's ad-driven orders came from customers who had never bought from you, you cannot yet say whether your advertising is growing the brand.
Frequently asked
What counts as new-to-brand?
An order from a customer who has not bought that brand on Amazon within the preceding twelve months. It is a brand-level definition, not a product-level one, so a customer switching between your own products is not new-to-brand.
Why is your DSP prospecting figure lower than the 50–70% benchmark band's top end?
Because it is measured rather than modelled. Roughly two-thirds is what our accounts actually produce across a real mix of categories, spend levels and audience maturity. We publish the measured number first and the benchmark band second, clearly labelled.
Can I improve my new-to-brand rate without spending more?
Usually yes, by moving spend rather than adding it. The largest single lever we see is cutting branded spend that is buying orders you already own and redeploying it into non-branded and upper-funnel lines.
How often do you refresh this page?
Every January as part of a scheduled statistics refresh, and whenever a portfolio-level figure moves materially. The updated date at the top of the page is the real one.
Dr. PPC runs your Amazon ads daily — an AI agent doing the work, operators from a $500M+ Amazon team supervising. $300/mo + 3% of ad spend, published and capped, month-to-month.
Start free for 30 daysPart of
- Amazon Advertising Statistics (2026)Pillar guide
Read next
Related
- DSP Prospecting Runs About Two-Thirds New-To-BrandData · remktr.com