.svg){kind=small}
TL;DR
- New customer revenue is revenue from first-ever buyers in the period. Returning customer revenue is revenue from anyone with a prior order. Together they always sum to 100%.
- Tag orders at the moment of sale using the customer’s lifetime order count — one prior order or more equals returning. Do the split at order level, not customer level.
- Healthy mix for a brand past 18 months: 35–55% returning. Under 25% is a leaky bucket. Over 70% is acquisition stagnation.
- Watch the shape, not the snapshot. A rising returning share with flat new revenue often means the ad account is starving, not that retention is improving.
- Fairview tags every Shopify and Stripe order as new or returning, tracks the mix by cohort and channel, and surfaces the drift before the board deck catches it.
New customer revenue and returning customer revenue are the two halves of every order report — one tells you the business is acquiring, the other tells you it is retaining. Track them separately, read the ratio between them, and you have the single fastest diagnostic for whether a commerce brand is compounding or quietly breaking.
Most Shopify dashboards show total revenue in large type and hide the split two clicks deep. That framing makes a brand look healthy for far longer than it actually is. A brand posting $1.2M a month with 82% returning revenue is not growing — it is harvesting a base it has already paid to acquire, and the new-cohort engine has stalled.
This guide covers what new vs returning customer revenue actually measures, how to calculate each without the usual errors, how Shopify classifies the two, what a healthy mix looks like by brand stage, and the cohort view that stops the ratio from lying to you. Pair it with ecommerce retention metrics, increasing LTV without ads, and D2C unit economics.
What is new vs returning customer revenue?
Definition
New customer revenue is revenue from customers placing their first-ever order with the brand inside the reporting period. Returning customer revenue is revenue from customers who had placed at least one prior order at any point in the brand’s history. Together they always equal 100% of order revenue.
The line between the two is drawn at the customer’s lifetime order count the moment the transaction clears. Zero prior orders, this order is new. One or more prior orders, this order is returning. The reporting window does not change who a customer is — it only changes which orders land inside the period.
Keep two confusions out of the definition. First, returning does not mean recent. A customer who last bought 18 months ago and came back today still counts as returning. Second, new does not mean new in the window. A customer who has been buying monthly for three years is still a returning customer on every order after their first — the window never resets them.
How to calculate new vs returning customer revenue
Three steps, in order, every time. Doing them out of order is what produces the wrong answer.
- Step 1 — tag each order at the moment of sale. Use the customer’s lifetime order count at checkout. Zero prior → new. One or more prior → returning. The tag belongs to the order, not the customer.
- Step 2 — sum revenue in each bucket for the period. Include discounts and refunds but not shipping and tax, unless your financial reporting also includes them (be consistent).
- Step 3 — compute the mix. New share = new revenue ÷ total revenue. Returning share = returning revenue ÷ total revenue. Report both in every weekly review.
Key insight
The split is an order-level calculation, not a customer-level one. If you sum revenue per customer and then label the customer, you lose the signal the moment a customer straddles the definition — their first order is new, their second is returning, and those two facts live inside the same reporting period.
The common shortcut — take all customers with two or more lifetime orders, sum their period revenue, call it returning — overcounts the returning bucket by 8 to 15% in most brands, because it attributes the customer’s very first order to the returning side. Over a year, that alone is enough to hide a new-acquisition problem for two full quarters.
New vs returning customer revenue in Shopify
Shopify classifies customers, not orders. The moment a customer’s order count in your store goes from zero to one, the customer is marked as returning on every subsequent order. The first order itself stays on the new side. This matches the order-level method above if you pull the data through the Orders API rather than the customer object, because the Orders API exposes customer.orders_count at the moment the order was placed.
Three Shopify-specific quirks to know before you trust the dashboard:
- Multi-storefront brands. A customer who bought on the US storefront and then the UK storefront shows as new on each one. If you operate multi-region, you need to reconcile by email or by customer ID outside Shopify — the native reports will not.
- Subscription app orders. Recharge, Bold, and Skio treat recurring charges as orders in Shopify. Depending on setup, a month-2 subscription charge can appear as a first-time order from Shopify’s perspective if the subscription is new to that email — check the app’s sync settings.
- Guest checkout. Two guest orders with the same email but different capitalisations or name differences will sometimes be treated as two customers, inflating new revenue. Shopify’s customer merger addresses this but only after the fact.
Shopify’s built-in Sales by New vs Returning Customers report is the right primary source. If the numbers look off, reconcile against the Orders API directly — that is what Shopify’s own analytics is reading anyway, without the subscription and multi-store complications. For context on the wider ecommerce reporting stack, see Shopify revenue dashboard limits.
What a healthy new vs returning revenue mix looks like
The right mix depends on the brand’s age, category, and whether the product is a one-and-done purchase or a replenishment category. The ranges below are directional, drawn from Shopify Plus benchmark studies and composite D2C operator reports.
| Brand stage | Replenishment category | Considered / durable | Read the ratio |
|---|---|---|---|
| 0–6 months | 85–95% new | 90–100% new | Too early to benchmark |
| 6–12 months | 65–80% new | 80–95% new | First repeat cohort appearing |
| 12–24 months | 45–60% new | 70–85% new | Retention mechanics showing |
| 2–4 years | 35–50% new | 60–75% new | Mix should be stable |
| 4+ years | 30–45% new | 55–70% new | Watch for new-share drift |
Two failure modes to watch for at every stage:
- Returning share under 25% past year one. This is the leaky-bucket pattern. Customers are buying once and leaving. Fixing it is usually cheaper than finding more customers.
- Returning share over 70% for a brand under three years old. Acquisition has stalled. The business is harvesting its base. Six months later revenue starts to fall and it will not be obvious why.
Quote-ready
A rising returning-revenue share is not always a good sign. If it rises because new revenue is shrinking, the brand is sliding down its own acquisition curve — the ratio just hides the slide for a quarter or two.
Track the split by cohort, not just by period
A monthly mix number is a lagging indicator. The leading indicator is how each acquisition cohort is contributing over time. Lay out the acquisition month on the vertical axis and the calendar month on the horizontal, and put the customer’s second-order revenue inside the cell. Now you are looking at a cohort retention curve expressed in dollars.
Three reads become obvious in that view:
- Cohort decay. If the January 2026 cohort shows $180 per customer of returning revenue by month six, and the July 2026 cohort shows $120, something in product, onboarding, or acquisition quality changed.
- Channel contamination. Break the cohort by acquisition channel. Meta prospecting cohorts often decay 30–50% faster than organic or email-acquired cohorts. The blended returning number hides this entirely.
- Pull-forward from promos. A BFCM cohort with huge month-one revenue and near-zero returning revenue for the next six months is not a cohort — it is a one-time discount event. Pull it out when computing the healthy mix.
The pillar on ecommerce retention metrics covers the full cohort layout with worked examples. The short version for a weekly review: track returning revenue per acquisition cohort at months 1, 3, 6, and 12, and compare across cohorts rather than across periods.
Five mistakes when tracking new vs returning revenue
Every broken new-vs-returning report I have seen traces to one of these five.
- Labelling at the customer level. Tagging a customer as returning and summing their period revenue pushes their first order onto the returning side. Label orders, not customers.
- Ignoring cohort drift. The blended mix can stay stable while every new cohort is decaying. Look at acquisition cohort revenue, not the aggregate.
- Reading the ratio without the absolute numbers. A 50/50 mix on $1.2M is very different from a 50/50 mix on $600K. Never report mix without the absolute revenue alongside.
- Letting subscription charges land as new customers. Recharge and other subscription apps can flag recurring charges as new depending on integration. Reconcile against the Orders API, not the app.
- Treating discount-driven revenue as healthy new revenue. A $10-off welcome code cohort often has 60% lower month-6 returning revenue than a full-price cohort. New revenue from discounts is a lower-quality version of the metric — track it separately.
How Fairview tracks new vs returning customer revenue automatically
Fairview connects to Shopify, Stripe, Google Ads, Meta Ads, and the rest of the commerce stack via native OAuth. Every order is tagged as new or returning at the moment it is pulled in — using the customer’s lifetime order count at transaction time, reconciled across storefronts and subscription apps so the Shopify edge cases do not leak in.
The Margin Intelligence layer takes the split further: it splits not just revenue but contribution margin, CAC payback, and channel ROAS by new vs returning. So when the Meta prospecting channel starts acquiring lower-quality cohorts, Fairview shows a returning-revenue decay curve on that specific channel two to three months before the blended number moves.
When the mix drifts beyond its rolling baseline, Fairview writes a named next-best action: “New customer revenue dropped 18% week-over-week while returning revenue held flat. Paid CAC up 22% on Meta; new-to-brand orders down on organic. Investigate acquisition quality before scaling budget.”
See pricing and tiers for the plan that fits your stack.
Order-level
Tag every order at source, not at the customer
Per channel
New/returning split by acquisition source
Daily
Mix, cohort, and drift refreshed every morning
Key takeaways
- New vs returning customer revenue is an order-level split, not a customer-level split. Tag at the moment of sale.
- Healthy returning share rises with brand age: under 25% past year one is a leak, over 70% under year three is acquisition stall.
- Watch cohort decay, not only the blended mix. The aggregate can stay stable while every new cohort is getting worse.
- Shopify’s native report is correct for single-storefront brands. Multi-store and subscription brands need to reconcile against the Orders API.
- Always pair the mix percentage with absolute revenue. A 50/50 mix at $1.2M and a 50/50 mix at $600K tell you very different stories.
See your new vs returning split by channel and cohort, every morning.
Connect Shopify and Stripe. Fairview tags every order at source and shows the mix, the cohort decay, and the drift before the board deck sees it. 14-day trial, no card required.
Frequently asked questions
New customer revenue is revenue from customers placing their first-ever order with your brand in the reporting period. Returning customer revenue is revenue from customers who have placed at least one prior order at any time. Together they make up 100% of order revenue, and the ratio between them is one of the fastest diagnostics of brand health.
Tag each order as new or returning using the customer’s lifetime order count at the moment of purchase — one prior order or more equals returning. Sum revenue in each bucket for the period, then divide each by total revenue to get the mix percentage. Use order-level, not customer-level, revenue so refunds and partial orders flow through correctly.
For an established D2C brand past 18 months, a 35–55% returning customer revenue share is the healthy range. Under 25% signals a leaky bucket — every acquisition dollar has to keep working twice as hard. Over 70% signals acquisition stagnation. Early-stage brands under 12 months will naturally run 70%+ new revenue and should not be benchmarked against mature brands.
Shopify flags a customer as returning the moment their total order count at the store exceeds one. The classification sticks to the customer, not the order — so a customer’s second order is returning, as is their tenth. Shopify counts across all sales channels on that storefront but does not span multiple storefronts or subscription app customers automatically.
Almost, but not identical. New customer revenue is defined at the order level in a reporting period — a customer placing their first order inside the window. New-to-brand revenue is the lifetime-unique definition — the customer had never bought from the brand before. The two match when the reporting window is the customer’s first ever window, and diverge when you re-slice history.
Both, but in sequence. Early brands under 12 months should grow new customer revenue first because there is no meaningful base to re-activate. Brands past 18 months with under 30% returning revenue should stop adding ad spend and fix repeat economics first — otherwise every new cohort becomes the last cohort. Growing only one half of the split for two consecutive quarters is a signal the business is out of balance.