← Back to blog
Blog

Before you chase more traffic — the channels that bring visitors and the channels that bring buyers are not the same

Before adding more ads or social posts to grow revenue, line up conversion rate and revenue by traffic source. A falling overall conversion rate does not always mean your store got worse. Once you know which entrance brings buyers, you can pick one channel to grow and one landing page to fix. Here is where it lives in GA4, and what GA4 alone cannot settle.

6 min read
  • Reading the numbers
  • GA4

This picks up from the first of the five numbers to check each week: splitting sessions by traffic source. Last time, the split was for finding why sessions moved. This post goes one step further and lines up, for each source, whether the people who came actually bought.

If you’re thinking “more revenue means more traffic” and are about to add ads or social posts, look at this view once before you do. The entrance that brings the most people and the entrance that brings buyers are usually not the same one.

A lower overall conversion rate doesn’t mean the store got worse

Take a made-up store. Average order value is ¥5,000 both last month and this month.

Sessions last monthOrders last monthSessions this monthOrders this month
Search1,000301,00030
Social1,00053,00015
Total2,000354,00045

This month you pushed hard on social and sessions doubled. Orders went from 35 to 45, revenue from ¥175,000 to ¥225,000, and the overall conversion rate fell from 1.75% to 1.125%.

Looking only at the total, it reads as “more people came, but the site got worse at selling.” Split by source, though, search is still at 3% and social is still at 0.5%. Neither entrance changed at all. What changed is the share of visitors coming through the entrance that converts less.

It’s the same shape as average order value moving with the mix of new and returning customers, covered under ③ in the five numbers post. A total moves whenever the mix of the groups inside it moves.

Line up three columns per source — sessions, conversion rate, revenue

Where it lives in GA4: Reports → Acquisition → Traffic acquisition. Set the rows to Session default channel group and look at three columns:

  • Sessions — how many people came
  • Session key event rate — what share of them bought (pick purchase as the key event; see ② in the five numbers post)
  • Total revenue — how much they spent

Use roughly the last four weeks. With one week, conversion rates for the smaller entrances swing too much to read.

For a rough outside reference, just one: in Contentsquare’s published figures (the version seen as of September 2026, covering Q4 2025 and more than 6,500 sites), paid search converted at 2.8%, AI referrals at 1.3%, and organic social at 0.7%. Read it only as “a gap of several times between entrances is normal.” It’s not something to hold your own store’s numbers up against.

Read the rows as four types, and pick one entrance to grow

Sort the rows by how many people come and how often they buy.

Buy oftenBuy rarely
Many comeToday’s revenue base. Protect itCandidate for fixing the landing page
Few comeCandidate to growLeave it for now

There is one decision here.

If you’re going to grow traffic, grow the entrance where few people come but they buy often. In the example above, bringing in more people from search gets you closer to revenue than pushing social further, for the same effort. For your store it might be newsletter readers, a particular referring site, or people searching your brand name. This one view tells you which.

For the entrance where many come but few buy, look at the landing page before sending more people there. Check whether social visitors land on a product page or the home page, and whether that page’s add-to-cart rate (④ in the five numbers post) is lower than the rest.

What GA4’s revenue by source can’t settle on its own

That gets you a candidate to grow. But one decision remains: can you cut back an entrance where people rarely buy? The GA4 screen alone can’t answer that.

The reason is that each row in Traffic acquisition is the entrance of the session in which the purchase happened. If someone discovers your store on social and buys a few days later after searching your store name, that revenue lands in the search row, not the social row. On top of that, Safari deletes cookies set by JavaScript after 7 days, so a return visit after a gap isn’t linked back to “the person who came from social earlier.” In Japan, Safari holds 48.7% of mobile browsing (as of August 2026), so this isn’t an edge case.

GA4’s sessions and revenue are also approximations, with a reporting delay. Orders whose tracking was blocked by an ad blocker, or orders taken by phone, don’t show up in any source’s row.

So GA4’s revenue by source is a trend. To decide “social isn’t driving revenue, so stop it,” you need material that ties each real order in your cart to the entrance that person came through. What’s missing isn’t effort. It’s that material.

Change one thing, and look at the same row four weeks later

Once you’ve picked one entrance to grow, write down the date and, four weeks later, look at the same row in Traffic acquisition. Check two things: did sessions from that entrance go up, and did its conversion rate hold. If more people came but the conversion rate dropped sharply, the way you grew it (ad targeting or what you posted) is bringing in a different crowd.

In NextRise Analytics, you can line up sessions, conversion rate, average order value and revenue by traffic source, with a comparison to the previous period. Start by opening the Acquisition tab of the sample store dashboard and sorting its entrances into this post’s four types.

If revenue by source doesn’t match what your cart’s admin screen says, the reason is in Why your order count and conversion count never match.

Related articles