GA4 ecommerce benchmarks
What 52 webshops saw in GA4 in October 2026: conversion rate, where sessions come from, and how many purchases GA4 can't tie to a real channel.
Every business counts once (one shop per GA4 account), so one big shop or a brand with ten country sites never sets the number. We show the median and the middle half: half of all shops sit inside that range, a quarter below it and a quarter above.
- 1.6%ecommerce conversion rate (middle half 0.9% – 3.1%)
- 24.4%of purchases credited to Direct (middle half 18% – 31.8%)
- 4.5%of purchases credited to Unassigned (middle half 1.2% – 8.1%)
Ecommerce conversion rate
| Median | Middle half | Shops | |
|---|---|---|---|
| All shops | 1.6% | 0.9% – 3.1% | 52 |
| Netherlands | 1.6% | 1.1% – 3.2% | 45 |
| General retail | 1.4% | 0.9% – 2% | 13 |
Purchases divided by sessions in GA4, per shop. A country is the shop's main market (most purchases); an industry is the category the shop picked in its GA4 settings. A group only shows with at least 10 shops.
Conversion rate by device
- 1.7%Mobile · 0.9% – 3.1%
- 2%Desktop · 0.8% – 3.4%
- 0.8%Tablet · 0.5% – 1.5%
Where sessions come from
| GA4 channel | All shops | Netherlands |
|---|---|---|
| Organic Search | 15.1% · 7.6% – 26.1% | 15.1% · 6.7% – 26.7% |
| Paid Search | 7% · 2.4% – 14.8% | 6.2% · 2% – 14.5% |
| Direct | 17.2% · 12.2% – 24.7% | 16.9% · 12.2% – 22.8% |
| Cross-network | 4% · 0.4% – 22.1% | 5.4% · 0.5% – 22.1% |
| Paid Social | 8.5% · 0.7% – 14.8% | 9.2% · 0.7% – 18.8% |
| Organic Social | 1.4% · 0.7% – 2.8% | 1.3% · 0.5% – 2.8% |
| 1.5% · 0.2% – 4.8% | 1.6% · 0.2% – 5.4% | |
| Referral | 1.1% · 0.6% – 2.5% | 1.2% · 0.6% – 2.6% |
| Paid Shopping | 4.1% · 0% – 11.3% | 3.8% · 0% – 11.3% |
| Unassigned | 2% · 0.6% – 4.2% | 1.9% · 0.6% – 4.1% |
Share of a shop's sessions per GA4 default channel group: median, then the middle half. Medians per channel don't add up to 100%. Channels under 1% for most shops are left out.
Direct and Unassigned: worth checking
At the median shop GA4 credits 24.4% of purchases to Direct, and at a quarter of shops it's over 31.8%. Unassigned is 4.5% at the median and over 8.1% at a quarter of shops. Some Direct is real: people type your address or use a bookmark. Unassigned has no real-world source. These numbers can't tell you which is which for your shop, but when yours sits in the top quarter, these are the setup problems we find most often:
- Click IDs and UTM tags lost on the way, for example by a redirect, a consent banner that reloads the page, or a browser that strips them.
- Returning visitors who look new because Safari and other browsers shorten cookie lifetimes, so the original source is gone.
- Purchases sent server-side or through the Measurement Protocol without the session they belong to. GA4 files those under Unassigned.
- A payment page or second domain that starts a new session, so the order lands on Direct or on the payment provider.
Frequently asked questions
What is a good ecommerce conversion rate in GA4?
Across 52 webshops in October 2026 the median was 1.6%, and half of the shops sat between 0.9% and 3.1%. Below 0.9% is in the bottom quarter, above 3.1% in the top quarter. Compare against shops in your own industry and market before you draw conclusions.
Does the conversion rate differ between mobile and desktop?
In October 2026 the median shop converted 1.7% of mobile sessions and 2% of desktop sessions. Read that with care: people who browse on their phone and buy later on a laptop show up in GA4 as a mobile visit without a purchase and a desktop visit with one.
How many purchases does GA4 credit to Direct?
At the median shop GA4 credits 24.4% of purchases to Direct, and a quarter of shops are above 31.8%. If your shop is in that top quarter, check for lost click IDs, short cookie lifetimes and a payment page that breaks the session before you credit it to loyal customers.
What does Unassigned mean in GA4?
GA4 received the event but couldn't tie it to a channel, usually because it arrived without a usable session: a server-side or Measurement Protocol purchase, or a session_id that went missing. The median shop has 4.5% of purchases in Unassigned, a quarter of shops more than 8.1%. Unassigned is never a real traffic source, so any share of it is worth tracing.
More Direct or Unassigned than this?
We trace where your sources get lost, from the first click to the purchase in GA4, and fix it.
Method · October 2026
- Source: the GA4 properties New North Digital can read, web data only, one shop per GA4 account (copies, roll-ups and a brand's other country shops in the same account removed), shops with at least 100 purchases that month. This month: 52 shops, mostly Dutch webshops of every size. Not a random sample of all webshops.
- Conversion rate is GA4's purchases divided by sessions. Visitors who refuse cookies are only partly in GA4 (modelled or not at all), so the rate is what GA4 shows, not what the shop sold.
- Every figure is a median over shops plus the middle half (25th to 75th percentile), each taken as the average of the three shops nearest that point, so no figure is one shop's own number. A country (the shop's main market, with its figures over all its visitors) or industry only appears with at least 10 shops. No shop is named.
- Updated monthly with the previous month's data.