Definition
New-customer share is the proportion of conversions, or of orders, that come from first-time customers. The affiliate reporting and incrementality guide lists it in its report table as the proportion of conversions from first-time customers, flagged in the order system, and the exclusion and testing guides read it as the new-customer share of credited orders: the orders a particular campaign, channel or partner is given credit for.
Conversions from first-time customers ÷ All conversions(%)
- Conversions from first-time customers are the credited conversions placed by someone with no earlier purchase or contract, by the business's own definition of an existing customer. That definition has to be written down first: how far back the records go, whether a lapsed customer counts as new again, and how a household or a company with several buyers is treated.
- All conversions are every conversion credited to the same campaign, channel or partner in the same period, counted the same way. Mixing a platform's conversion count in the denominator with the business's own flag in the numerator produces a figure that describes neither.
- The source of the flag is the business's order system or customer file, not the campaign's report. The guide to excluding existing customers explains why: the platform counts only the conversions it can see and labels them by its own lists, so it cannot say whether its exclusions are working.
Recording whether each conversion came from a new or returning customer, at the moment it happens and without double counting across channels, is the job of the guide to new and returning customers in marketing data. This page is the measure that recording makes possible.
How to interpret it: guardrail and partner quality signal
The share has no direction that is always better, which is why its direction is context. It does two jobs on the site.
A guardrail beside the main measure. Automated bidding and broad targeting find whoever converts most cheaply, and existing customers convert more easily than strangers. A ProspectingAdvertising aimed at people who have not yet visited, engaged with or bought from the business. campaign drifting towards them looks like a success: cost per conversion falls and return rises while the number of new customers stays flat or shrinks. The guide to Advantage+ audiences points to a breadth test that uses cost per purchase as the measure and new-customer share as the guardrail, and the guide to testing and scaling lists a falling share among the signals that should pause a budget increase. In paid search the sales objective page for search reads it with contribution: a rising return beside a falling share means the account has drifted towards brand searches and returning buyers.
A quality signal for partners. In affiliate programmes the share separates partners who bring demand from partners who sit at the end of a journey that was already under way. The affiliate reporting guide calls the proportion of a partner's conversions from first-time customers the best single quality signal: voucher, cashback and browser-extension partners tend to show low shares, while content, comparison and referral partners tend to show higher ones. A partner with a high share is probably bringing customers the business would not otherwise have had.
Read it by campaign, channel or partner, never only as a blended figure for the whole business, and always beside the count of new customers and the cost per new customer. A share can rise simply because the campaign's total conversions fell, which is not progress.
A cookware store finds who its cheapest purchases come from
Yzmelda Cookware runs a broad prospecting campaign beside a narrower interest-based one. In the first month of a quarter the broad campaign is credited with 410 purchases at £31 each; by the third month it has 520 at £24, and the team plans to move budget into it. Before doing so they match the credited orders against the customer file. In the first month 238 of the purchases (58%) came from first-time buyers; in the third, 177 (34%). Spend barely changed, so the cost per new customer rose from about £53 to about £71, while the narrower campaign held a share of 61%. The cheaper purchases were mostly repeat buyers, so the team adds a customer exclusion to the broad campaign and keeps the budget split until the share recovers.
What moves it
- Exclusions and their freshness. Customer lists exclude only the people they can recognise. A list that is incomplete, unmatched or weeks old lets recent buyers back into prospecting, and the share falls without any change to the campaign.
- Automated bidding and broad targeting. The more freedom the system has to find converters, the more it will find people who already buy unless an acquisition goal or exclusion tells it otherwise.
- Brand and returning-customer queries. In search, spend that shifts towards brand terms lowers the share, because those searchers mostly know the business already.
- Partner mix. In affiliate programmes, growth in end-of-journey partners lowers the programme's share even when every partner's own figure is unchanged.
- The definition of existing. Changing the lookback, or starting to count lapsed customers as new, moves the share without any change in behaviour. A definition change needs a note in the report.
What good looks like
The site cites no reliable benchmark for new-customer share, and a figure from another business would mislead more than it helped. What counts as healthy depends on what the spend is for. A campaign whose job is acquisition should hold a high share; a retention or remarketing campaign is meant to reach existing customers and its share should be low. The exclusion guide also sets out when existing customers should not be excluded at all, and a business whose margin comes from repeat orders may be right to spend on bringing people back.
The useful comparisons are internal: the same campaign or partner against its own earlier periods, campaigns with the same job against each other, and the share read together with the cost per new customer. The test that matters is economic rather than conventional: whether the new customers the spend finds cost less to acquire than they are worth to the business over the time they stay.
Related metrics
The cost of each new customer is customer acquisition costCost to acquire one new customer., and the share is what stops a cost per conversion being mistaken for it. Repeat Purchase RateShare of customers who placed more than one order in the period. measures the other side of the customer base, how many buyers come back, and the two are best read together. A low share is often the first sign of weak IncrementalityConversions caused by marketing., because orders from people already buying are the ones most likely to have happened anyway, but only a test settles that question. ProspectingAdvertising aimed at people who have not yet visited, engaged with or bought from the business. is the activity whose health the share is most often used to check.
Common mistakes
- Taking the share from the platform's report, which counts only what it can see and classifies by its own lists.
- Reading the share without the count, so a campaign that lost most of its conversions looks better because the few that remain are new.
- Judging retention campaigns by it, when reaching existing customers is their purpose.
- Changing the definition of an existing customer mid-series without noting it, so the trend reflects the rule rather than the customers.
- Treating a high share as proof of incrementality. New customers can still be people who would have bought anyway; the share is a signal, not a test.
