What Is Visitor Segmentation? Splitting Your Traffic

What Visitors Get Grouped By

A segment is a group of visitors who share something. The useful ones come from things you can see during a visit:

Behavior: products viewed, time on a page, cart adds, a dropped checkout.
Visit history: a first visit, or the third one this week.
Traffic source: an email, a Google search, a social ad.
Device: phone, tablet or desktop.
Location: country or region, which can change shipping costs.

Segmentation only draws the lines. Deciding what each group sees comes next. When that's based on what people do, it's called behavioral targeting.

Why One Store-Wide Average Misleads

Your conversion rate is one number for everyone, so it mixes people ready to pay with people just looking. In our visitor data from 958 Shopify stores, the average conversion rate was 2.2%. So about 98 of every 100 visits end without an order.

Those 98 aren't alike. A shopper who left a full cart is close to buying. Someone who bounced off your home page isn't, and a message that helps one can annoy the other. For discounts, the split that matters most is between dedicated buyers and walk-away customers.

Is Your Segment Big Enough to Learn From?

Example: A leather wallet store gets 6,000 sessions a month. The owner builds a really narrow segment: returning visitors on a phone who came from Instagram and viewed at least three wallets.

That matches 60 sessions a month.
Say this group buys at 5%. That's 60 × 5% = 3 orders a month.
If a new message gets it to 7%, that's 60 × 7% = 4.2, so about 4 orders.
One extra order a month can easily be luck. You'd need many months of data before trusting the change.

Drop two filters and keep just "returning visitors on a phone." That's 1,200 sessions a month. If it buys at 4%, that's 1,200 × 4% = 48 orders, and a real change has a far better chance of showing up. Telling a real difference from luck is what statistical significance checks.

When Segments Help and When They Hurt

Segments pay off when groups really behave differently and you've got something different to do for each. If every segment sees the same banner, nothing changed.

They hurt when:

You slice too thin. Ten tiny segments give ten sets of numbers too small to read.
You can't act on the segment. You rarely know an anonymous visitor's age or income, so segments built on them stay on paper.
One weak signal decides everything. A phone visitor who scrolls fast may be in a hurry, not uninterested.

Where Shopify Groups Your Visitors

Shopify's analytics can split sessions by device type, location and traffic source. That's a look back: you see how each group did last month.

Shopify also has customer segments. They filter your customer list, for example by orders placed or amount spent, which is customer segmentation. A first-time visitor who browses and leaves without sharing an email isn't on that list. Visitor segmentation counts them too.

Grouping visitors while they're still in your store and changing what they see usually takes an app. Our smart targeting guide for visitor segmentation covers picking signals and turning groups into rules.

How Growth Suite Segments Visitors

Growth Suite is our app, and splitting visitors by intent is its core. It tracks each visit live and predicts how likely that visitor is to buy. Dedicated buyers, who look ready to buy, don't get an offer. Walk-away customers may see a time-limited offer once they show engagement, like time on page or cart activity.

For those visitors, more interest means a smaller discount for a shorter time, within limits you set. You can also target cart abandoners, return visitors and device types, and connect email capture to Klaviyo or Mailchimp. Its A/B testing module compares offer settings by conversion rate, average order value or total revenue.

A Segment Is Not a Permanent Label

People often treat a segment as a tag stuck to a person, like "bargain hunter." A visitor segment describes a visit, though. A shopper lands as a browser. Ten minutes later she adds a wallet to her cart and joins your "items in cart" group. Next week she's a returning visitor on a laptop. Good segmentation reacts to where someone is now, not where they got filed last month.