Shopify conversion rate: exclude bot traffic

Most founders I talk to think their conversion rate is 2 to 3%. Often it is not, because a part of the sessions it is divided by were never people.

Bots visit your store, load pages and leave. They will never buy. But every bot session counts as a visit, so it pushes your conversion rate down and makes the store look worse than it is.

Why bots end up in your numbers

Conversion rate is orders divided by sessions. Orders come from real customers. Sessions come from anything that loads your pages: shoppers, crawlers, price scrapers, uptime checks and scripts nobody can name.

Shopify now labels every session as human or bot. It looks at each event in a session against a set of criteria, then classifies the whole session based on all of its events. Shopify says this is designed to be conservative, because it is better to miss some bots than to label real customers as bots. (Shopify Help Center)

So the filter will not catch every bot. But what it does catch, you can trust.

What changed in September

When the filter first arrived in October 2025, you applied it yourself: a “Human or bot session” filter and dimension for sessions reports. (Shopify changelog)

Then Shopify changed the default. With the session measurement update that rolled out between 21 and 23 September 2026, identified bot sessions are filtered out of session related reports by default. The same update changed how sessions themselves are counted, so session numbers can drop compared to before. Shopify says orders and sales are not affected. (Shopify Help Center)

That means a lot of stores saw their conversion rate move up at the end of September without changing anything on the site. If that is you, it is not a bug. That is the real number, and it should be your new baseline.

Two limits worth knowing:

• Bot classification only applies to data from 7 October 2025 onwards. Older sessions cannot be classified, so a comparison with earlier periods still has bots in it.
• It only applies to session based metrics.

How to check your own store

1. In your Shopify admin, go to Analytics, then Reports.
2. Open any report built on sessions.
3. Find the “Human or bot session” filter in the configuration panel and check what it is set to. It should exclude bots.
4. Now add “Human or bot session” as a dimension instead. You get human and bot sessions side by side, so you can see how much of your traffic was never a person.
5. Do the same with a report that breaks sessions down by traffic source. A channel with plenty of sessions and almost no human ones is not a channel. It is noise.

How to read your numbers now

What I usually see once bots are out: they make up a meaningful chunk of sessions, the real conversion rate jumps noticeably, and the channels that actually bring buyers become obvious. Ad spend decisions suddenly make more sense.

A few things I keep in mind when I read the data:

• Draw a line at late September. Do not compare October with August as if nothing happened. The way sessions are counted changed in between.
• Compare like with like. Shopify’s own reports filter bots now. Your other tools may count visits their own way, so the numbers will not match, and that is expected.
• Look at sessions per channel, not just the total. A source that only ever sends bots should not get credit in your reports, and it should not get budget.

A higher conversion rate after filtering does not mean the store got better. It means you finally see how it performs for people. The leaks are still there, they are just easier to find. On mobile, one of the most common ones is a sticky add to cart hidden behind an app widget.

What I do

On every audit I start with clean numbers. I check the bot filter, split sessions by human and bot per channel, and note the date of the September change so nobody compares two different ways of counting. Only then do I look at where people drop off.

If your conversion rate jumped in October and you are not sure what to trust, send me your store and I will tell you what your real numbers say.