← Back to blog
Blog

The traffic source saved on every WooCommerce order — how to read it and where it stops

WooCommerce records where each customer came from on the order itself. Line those order sources up by channel and the order records tell you which channels are producing sales. But only the last channel is kept, and visits that did not end in an order leave nothing behind. Based on the official documentation, this post covers how to read it and what to decide with it when it disagrees with GA4.

5 min read
  • Reading the numbers
  • GA4

In an earlier post (Before you chase more traffic), we put sales and conversion rate side by side for each traffic source and picked one channel to grow. That post used GA4. This one does the same job with WooCommerce order data.

WooCommerce stores have something most carts do not give you: every order carries a record of where that customer came from. GA4 derives traffic sources from visit data. WooCommerce attaches the source to the order record itself. The one rule this post argues for: decide order counts and sales by channel from the source saved on the order. Conversion rate is the only thing you read on the visit side.

Everything below about WooCommerce comes from the official Order Attribution Tracking documentation, read in November 2026.

1. The source saved on the order — where did this order come from?

When an order is placed, Order Attribution Tracking stores:

  • Referring source — the URL of the site that sent the visitor to your store
  • UTM parameters — the tags you added to ad or newsletter links (source, medium, campaign, content and term)
  • Device type — desktop, tablet or mobile
  • Session page views — how many pages the visitor viewed in that visit before ordering

You can see it in two places. The order edit screen gets an Order attribution box showing the origin, the device and the page views. The Orders list gets a new column showing each order’s origin. The feature is switched on or off under WooCommerce, Settings, Advanced, Features.

A purchase in GA4 is a signal sent from the browser each time an order completes. What WooCommerce saves is a record attached to the order itself, so the count matches your real orders. That difference is what you lean on later, when the two disagree.

2. Line up orders and sales by channel — which channel is producing sales?

The approach is the same as in the earlier post. Take roughly the last four weeks of orders and list each source with its order count and total sales. A single week is too noisy for channels with only a handful of orders.

Then decide one thing: which channel is carrying your sales? Search, a newsletter campaign tagged with UTM, a particular referring site — the order records answer that directly. Because there are no missed purchases mixed in, you can treat the order side as the source of truth for how many orders each channel produced and how much they were worth.

It helps to keep device type and page views next to the same table. If the orders from one channel lean heavily toward mobile, for example, you know to check that channel’s landing page on a phone first.

3. Only the last channel is kept — where does the customer who found you on social and bought through search end up?

Three habits of the data are worth knowing, all stated in the documentation.

First, it uses last-click attribution. The documentation says all of the credit goes to the last interaction before the customer lands on your site. UTM-tagged and organic search visits override the earlier source. A direct, typed-in visit never overrides a source that was already identified.

Second, it does not follow visitors across sessions. The cookies it uses expire with the session, and a session lasts 30 minutes by default. So if someone discovers your store on social media and comes back the next day by searching your store name, the order is credited to search. Nothing is recorded against social.

Third, orders placed while the feature was switched off carry no source. If you turned it off at some point, check that the period you are analysing does not overlap with that gap.

None of this is an error in the data; it is the counting rule. So read the table in section 2 as a table of the channel that closed the sale. It cannot tell you which channel first introduced the customer to your store.

4. When it disagrees with GA4 — which one decides what?

Put WooCommerce order sources next to purchases by source in GA4 and they will not match. They count different layers. The order source belongs to the order layer; GA4 purchases belong to the visit layer.

Here is where to draw the line.

What you want to decideWhich numbers to use
How many orders, and how much in sales, each channel producedWooCommerce order sources
What share of visitors bought (conversion rate)Inside GA4 only, dividing purchases by sessions for the same channel

Avoid dividing WooCommerce order counts by GA4 sessions to get a conversion rate. The numerator and denominator are counted by different systems, so when that rate moves you cannot tell whether customers changed or the counting drifted.

5. What is missing — where are the people who came and did not buy?

Order sources are strong on counts and sales, but one decision is still out of reach. Order records contain nobody who did not buy.

Recall the four types from the earlier post. A channel that brings many visitors who rarely buy and a channel that brings few visitors who buy often look the same from the order side: both are just channels with few orders. And only the second one is worth growing.

So to decide where to add ad spend or posts, you need the order sources and the visit counts for each channel, matched using the same channel definitions. What is missing is not effort; it is that matching data.

If you want to understand the mismatch between WooCommerce sources and GA4 as a difference in layers, see Why your order count and conversion count never match. For another platform, see the post on Color Me Shop.

Related articles