Which Traffic Sources Send Buyers, Not Browsers

The traffic source sending you the most sessions is often not the one making you the most money. A channel can convert well and still be a poor source of revenue if the orders it drives come back at twice the rate of everything else. Ranking sources by sessions or last-click revenue alone hides that, because both metrics stop counting the moment the sale completes.

This guide walks through building a session-quality table, source by source, using conversion rate, average order value, and return rate together. The goal is a single view that tells you which channels send buyers worth keeping, and which ones send tourists who click, sometimes buy, and often send it back.

Why Volume and Last-Click Revenue Both Lie

Most traffic reports rank sources by sessions or by revenue attributed at the click. Both are incomplete on their own. A source with huge session volume and a mediocre conversion rate can still generate the most raw orders, which looks like a win in a dashboard sorted by revenue. But if a meaningful share of those orders come back, the revenue number booked at checkout was never real profit to begin with.

Return rate is the missing variable most traffic reports never join in, because it lives in Shopify's returns data, not in the ad platform or GA4. A source that looks like your best performer on conversion rate can be your worst on a return-adjusted basis, and you won't see it until you pull the three metrics into one table.

Building the Session-Quality Table, Step by Step

You need three numbers per source: conversion rate, average order value, and return rate. Conversion rate and sessions come from GA4 or your ad platform, AOV and return rate come from Shopify. None of this requires a data team, just three exports and one join key: the source/medium value.

Step 1: Pull Sessions and Conversion Rate by Source

In GA4, go to Reports > Acquisition > Traffic acquisition. Set the dimension to Session source / medium and the date range to a full month, longer if a source has low volume. Export sessions and the "Session conversion rate" column (add it via the metric selector if it isn't shown by default).

If you run paid ads, cross-check paid sources against your ad platform's own conversion tracking too, since GA4 and ad-platform pixels sometimes attribute differently. For this exercise, GA4's session source/medium is the more consistent join key across sources, so use it as the base list.

Step 2: Pull Orders and AOV by Source

In Shopify, if you have UTM parameters or referral data flowing into orders, go to Analytics > Reports and look for a report broken out by referrer or marketing channel, or export orders with the Referring Site and UTM Source/Medium fields included via Orders > Export. Group orders by source/medium and calculate average order value as total revenue divided by order count for each group.

If your UTM tagging is inconsistent, this step will surface it fast: any source with a suspiciously large "direct" or "none" bucket usually means links went out untagged. Fix tagging going forward before trusting the historical split too heavily.

Step 3: Pull Return Rate by Source

This is the step most traffic reports skip entirely. In Shopify, go to Orders, filter by Return status, and export returned or refunded orders for the same period, again with source/medium attached. Divide returned orders by total orders for each source to get a return rate per channel.

If your Shopify data doesn't tag returns with the original order's traffic source directly, join the returns export back to your Step 2 order export using the order number as the key. This is the one step worth doing carefully, since it's the number that changes the ranking most.

Step 4: Join the Three Into One Table

Put source/medium as rows, then sessions, conversion rate, AOV, and return rate as columns. This is the table that changes the conversation, because it's the first time volume, quality, and durability of the sale sit next to each other.

Here's a worked example for a mid-size D2C apparel brand, one month of data:

Source	Sessions	CVR	AOV	Return rate	Verdict
Meta Prospecting	42,000	1.1%	$58	22%	Worst by profit
Google Search (Brand)	6,200	6.8%	$71	8%	Best per session
Google Shopping	18,500	2.4%	$64	14%	Solid, watch return rate
Email/Klaviyo	9,800	4.2%	$82	9%	Strong, undervalued
Influencer/Affiliate	15,300	1.8%	$49	19%	High volume, low quality
Organic Search	11,000	2.6%	$67	10%	Steady baseline
Direct	7,400	3.9%	$75	7%	High intent, low risk

Meta Prospecting has more sessions than every other source combined, and it looks fine on conversion rate alone at 1.1%, roughly in line with cold top-of-funnel traffic norms. But it pairs a below-average AOV with a return rate nearly three times the site average. Every other source in this table converts fewer sessions into orders that are worth less time and money to acquire and process than what Google Search (Brand) and Email deliver on far less volume.

Reading the Table: Where the Real Losses Hide

The pattern that matters most is a source with both high volume and high return rate, because that combination compounds. Meta Prospecting isn't just sending lower-quality orders, it's sending a lot of them, which means the fulfillment cost, restocking cost, and refund processing cost scale with the channel's spend, not against it.

A low-volume, high-conversion, low-return source like Google Search (Brand) or Email is usually underfunded relative to how well it performs, precisely because it doesn't show up as impressive in a sessions-sorted report. These are the channels worth defending, or even shifting budget toward, before chasing more volume elsewhere.

A high-volume, low-conversion, high-return source like Meta Prospecting or an under-vetted affiliate deal isn't automatically wrong to run. Cold prospecting has to convert worse than branded search by definition, since it's reaching people earlier in their decision. But it needs a return-adjusted view before you decide how much budget it deserves relative to everything else, and before you accept its raw revenue number at face value in a monthly report.

What to Do With This Once You Have It

Rebuild this table monthly, not quarterly. Return rates and AOV drift with seasonality and promotions, and a source that looked fine in a slow month can look very different once holiday-driven impulse purchases and their returns run through the funnel. If a source's return rate climbs two or three points above the rest of the table for two months running, that's worth a real conversation before the next budget cycle, not after it.

Use the table to reallocate, not just to report. Shift incremental budget toward sources sitting in the high-CVR, low-return quadrant even if their raw session volume looks small next to a prospecting channel. And before cutting a high-volume source outright, check whether its AOV benchmark against D2C ecommerce benchmarks for your category, since a low AOV can be normal for a channel that's meant to acquire, not to maximize order size.

Frequently Asked Questions

Why isn't sessions or last-click revenue enough to judge a traffic source?

Both stop counting at the point of sale. Sessions ignore whether a visit ever converts, and last-click revenue books an order as a win the moment checkout completes, even if it's returned later. A source can have strong session volume and solid revenue on paper while quietly running a high return rate that erases the margin, which neither metric on its own will ever show.

How do I calculate return rate by traffic source?

Export returned or refunded orders from Shopify for a period, tagged with the original order's source/medium, then divide returned orders by total orders for each source. If returns aren't tagged with source directly, join the returns export back to your original orders export using the order number as the key, then group by source/medium on the joined data.

What counts as a good return rate for a traffic source?

It depends heavily on category and price point, so compare sources against each other within your own store rather than an external benchmark. As a rough signal, a return rate for one channel running several points above your site-wide average, alongside high session volume, is worth investigating before that channel gets more budget. Apparel and footwear naturally run higher than most other categories.

Should I cut a high-volume, high-return traffic source entirely?

Not automatically. Cold, top-of-funnel sources are expected to convert and retain worse than branded search or email, since they reach people earlier in the decision. The better move is usually reallocating incremental budget toward better-performing sources first, then testing creative, targeting, or landing page changes on the high-return source before cutting it outright.

How often should I rebuild this traffic source quality table?

Monthly, at minimum. Return rates and AOV shift with seasonality, promotions, and product mix, so a channel that looks healthy in one month can look different the next. A quarterly cadence is too slow to catch a return-rate spike before it has already eaten several months of ad spend on a source that wasn't actually returning it.

Conclusion

Sessions and last-click revenue tell you who showed up and who checked out. They don't tell you who kept what they bought. A source with strong conversion rate can still be your least profitable channel once return rate is in the table, and a quiet performer like branded search or email can be your best one at a fraction of the volume.

Build the table once, keep it current monthly, and let it drive budget decisions instead of a sessions-sorted dashboard. Pair it with the AOV calculator to model what a return-rate improvement is worth per source, and check the average order value glossary entry for how AOV should be read alongside conversion rate. If your funnel itself needs a closer look before the traffic question, start with funnel analysis in GA4 for Shopify stores. Both feed into the broader store funnel playbook.