What New-to-Brand measures, what it misses, and how to read it as a growth signal

A customer who purchased your product on Amazon 365 days ago shows up in your Amazon reporting as new-to-brand. So does the shopper who purchased your product for the first time on Amazon yesterday, who spent five years buying from your Shopify store. New-to-brand is treated like an acquisition metric, but it can get caught counting people who already know you.

The metric is still important, but it’s worth analyzing what it truly captures, what it misses, and what any given NTB number means next to the campaign that produced it.

What NTB does measure

Amazon counts an order as new-to-brand when the shopper hasn't purchased from your brand on Amazon in the previous 365 days. It's a simple definition, but it answers a much narrower question than many marketers assume.

NTB doesn't ask whether someone is a new customer to your business. It asks whether Amazon has seen that shopper buy from your brand on its marketplace within the last year. That distinction matters because it shapes every NTB number you see.

The definition has three built-in boundaries. It's measured at the brand level, so customers buying a new product line still count as existing if they've purchased from your brand before. It's measured on Amazon, so shoppers who have spent years buying from your website, retail partners, or other marketplaces still appear as new when they make their first Amazon purchase. And it's measured over a 365-day lookback window, meaning a customer who returns after more than a year is treated as new again.

The four things NTB can't see

New-to-brand measures Amazon’s relationship with the shopper. Your brand's relationship with that same person is the part it can't see. Every blind spot comes from who Amazon considers an existing customer, or which touchpoint it credits for the sale.

• The lapsed buyer: A shopper purchased from your brand 366 days ago and places another order today, and Amazon calls them new-to-brand. Obviously, they aren't a new customer at all, just a repeat buyer who fell outside Amazon's lookback window, so part of what reads as acquisition is really re-activation. How much this measurement distorts your NTB number depends on your category. Brands with long repurchase cycles, durables and seasonal goods and high-consideration purchases, collect far more of these false positives than a consumable good that people reorder every few weeks.

• Shoppers buying from your DTC site: A shopper spends five years buying from your Shopify store, places a first Amazon order today, and lands in your reports as new-to-brand. Amazon only sees Amazon. Every existing customer you move onto the marketplace, from your own site, your email list, or a retail relationship, arrives looking like someone you've never met, and inflates the number on the way in.

• Existing customers buying a new product line: New-to-brand is measured at the brand level, so a customer who's bought your cleanser for a year and adds your new moisturizer today still counts as an existing customer. That moisturizer just pulled someone into a product line they'd never tried, and NTB records nothing for it.

• Shoppers who saw an ad but never clicked: A shopper sees a Sponsored Display impression, doesn't click, then searches your brand by name three days later and buys. NTB keys off the click, so that view-influenced purchase lands as organic with no ad credit, or drops out of the NTB picture entirely. The ad created a new customer. The reporting has no clean way to say so.

The first three gaps are about who Amazon counts as new. The fourth is about which touchpoint Amazon credits. Both are Amazon's rules rather than your customer history, and that's the whole point: NTB reports Amazon's relationship with the shopper, not your brand's. It doesn't make the metric useless. It makes it a signal you read in context rather than at face value.

What NTB tells you depends on what the campaign was trying to accomplish

A NTB number only means something once you know what the campaign was trying to accomplish. There's no universal healthy rate, because there's no universal campaign, and the same number can mean very different things depending on what you asked that campaign to do. Three things put it in context:

• What acquiring a new customer actually costs you: Divide ad spend by new-to-brand orders and you get what it actually costs to acquire a new customer, a different figure from a blended cost per order that mixes new and returning buyers together. Blended cost tells you how efficiently you're converting demand; new-customer acquisition cost tells you how much you're paying to create it. Perpetua's DSP engine exposes New-to-Brand ROAS and New-to-Brand cost per purchase directly, so you can optimize toward the cost of acquisition itself instead of rebuilding it from raw reports every month.

• Whether the NTB rate fits what the campaign was built to do: A branded-defense campaign exists to protect shoppers already searching your name, so a high new-to-brand rate there should make you suspicious rather than pleased. It's a signal worth investigating, and the likeliest explanation is weak organic presence on your own brand terms, with paid capturing shoppers your organic listings should already be winning. A campaign built to defend branded traffic and recapture it efficiently should skew heavily toward customers you already have. Category and conquesting campaigns carry the opposite job. They go after people shopping the category or a competitor, shoppers who mostly don't know you, so a high NTB rate is the campaign doing exactly what it was built for. Same number, two jobs, two entirely different ways to read the result.

• Whether that new customer ever came back: A single new-to-brand sale at a high acquisition cost isn't good or bad by itself. What settles it is what the cohort does over the following year. A group that comes back and repurchases across the next few quarters was an investment that paid off; one that buys once and disappears was an expensive transaction you mistook for growth. Perpetua's Consumer Long-Term Value report tracks each new-to-brand cohort month by month across a 12-month or 5-year window, showing how many of those buyers came back, how long they took to place a second order, and the average lifetime value of everyone acquired in that period. It also breaks that value down by the product a customer entered on, so you can see which first purchases pull shoppers deepest into your catalog.

The takeaway

New-to-brand is one of the few Amazon metrics that points toward incrementality, toward whether your advertising is bringing in people who weren't already heading your way. That's what makes it worth reading carefully. If you read it as a straight count of new customers, it overstates your reach and buries your best cross-sell, but if you read it as a proxy, next to the campaign that produced it and the cohort it created, it becomes one of the clearest signals you have for whether spend is buying growth or harvesting demand you already had.