Zum Hauptinhalt springen

Cohorts and Retention

Getting a first order is nice. Getting the second order is where a store starts to grow. These two pages answer the questions every store owner asks sooner or later:

  • Retention: Do my first-time buyers come back? How soon? Is my customer base growing or leaking?
  • Cohorts: Who exactly are those people, and how do I reach them?

Both live in PathGuide → Customer Insights, next to the Customer Directory. Open Customer Insights and pick Retention or Cohorts in the row of sub-pages at the top.

Now open for every store

Retention and Cohorts finished their test run in November 2026 and are now switched on for every PathGuide store. No setup needed. If you open one and see "not set up for this store yet", tell us and we'll take a look.

Before you start: where the numbers come from​

Both pages read your store's order history, not your visits. That has three effects:

  1. They have their own time ranges. The Date range select at the top of PathGuide does not change them. Each page has its own selectors, explained below.
  2. They see the orders PathGuide has collected. When you install PathGuide, it quietly imports your order history in the background. Shopify gives apps the last 60 days of orders by default, and every new order is added as it happens. So a young install has a short memory, and a long-time customer can look "new" at first. The tables fill in over time, like a photo album.
  3. Test, cancelled and refunded orders can be counted differently than in Shopify's own reports. Small gaps between PathGuide and Shopify are normal.

Retention​

The Retention page has the subtitle "How many first-time buyers order again, and whether your customer base is growing." It has two parts.

Part 1: Repeat-purchase retention​

Three big numbers sit at the top:

NumberWhat it means
Order again within 30 daysOf customers whose first order is at least 30 days old, the share that ordered a second time within 30 days
Order again within 90 daysThe same, for 90 days (only customers whose first order is at least 90 days old)
Median days to second orderFor customers who did come back, the typical number of days between order one and order two

Below them is the repeat-purchase table. The First orders from select picks Last 6 months (the default) or Last 12 months.

An example repeat-purchase table. Rows are the month of the first order, columns are the months after it, and each cell is the share of that month's new customers who ordered again. One highlighted cell reads 14 percent: of August's 400 new customers, 14 percent ordered again one month later.

How to read it:

  • Each row is a group of customers who placed their first order in that month. The New customers column says how many.
  • Each column is a later month: Same month, Month 1, Month 2, and so on.
  • Each cell is the share of that row's new customers who placed another order in that later month. Darker cells mean more customers came back.
  • A star (*) marks the newest cell in each row. That month is still running, so the number will still go up. Don't panic about a low starred cell on the 3rd of the month.
  • Months are calendar months in UTC.

Click a cell to see how many customers it holds and the revenue they brought, then save them as a cohort with one click. They show up on the Cohorts page as a fixed list.

Part 2: Customer lifecycle​

A bar chart that sorts your customers into four groups for each period:

GroupMeaningWhere it shows
NewFirst order in this periodAbove the line
ReturningOrdered in the period before tooAbove the line
Won backOrdered again after a gapAbove the line
Went quietOrdered in the period before, not in this oneBelow the line

Pick a Period length of 30 or 90 days; the chart shows the last six periods. When the bars above the line grow faster than the bars below it, your customer base is growing. When "Went quiet" keeps getting taller, it's time for a win-back plan (see the playbook below).

Click a bar segment to save exactly those customers as a cohort. "Everyone who went quiet last month" is now a list you can act on.

Cohorts​

A cohort is a group of your customers, built from their order history. The page subtitle tells you how many customers PathGuide knows about: "Groups of your N customers, built from their order history."

Start from a template​

Four ready-made cohorts, one click each:

TemplateThe question it answersRules
Lapsed repeat buyersWho bought more than once and then stopped?2+ orders, last order 60+ days ago
First-time buyers, last 30 daysWho just bought for the first time?1 order, first order within the last 30 days
VIP (5+ orders)Who are my most loyal customers?5+ orders
One-time buyers, 90+ days agoWho bought once and never came back?1 order, last order 90+ days ago

A template only fills in the builder. You can change any rule before you save.

Build a cohort​

Add conditions with the + buttons. Every condition has to be true for a customer to match (it's an "and", not an "or").

ConditionChoices
Number of ordersis at least, is more than, is at most, is less than, is exactly, is between
Lifetime valuethe same six choices, in your store's currency
Last orderwas at least … days ago, was within the last … days, was between … and … days ago
First orderwas within the last … days, was on or after, was on or before, was on, was between
Customer tag: has any ofup to 10 of your Shopify customer tags
Customer tag: has none ofup to 10 tags to leave out

So "tagged wholesale, lifetime value at least 500, no order for 90 days" is one cohort. Fancy.

Click Show matching customers to see how many customers match, their lifetime value and their orders. Then name the cohort and pick how to save it:

  • Save as live cohort always shows who matches today. A customer who lapses tomorrow joins on their own.
  • Save as fixed list freezes exactly the customers who match right now. Handy for "everyone who got the spring offer".
Customer tags need a quick import first

The first time you open Cohorts, PathGuide imports your customers' Shopify tags in the background. A progress line shows how far it is. The tag conditions appear once the import has finished; until then you can't save a cohort with a tag rule, so you never get a half-filled list. If the import fails, the page says why and offers Try again.

Upload a list​

Already have a list from somewhere else? Upload a CSV or text file with one Shopify customer id per line (the id first on each line), up to 5 MB. Shopify's own customer export works: PathGuide reads its "Customer ID" column. It becomes a fixed list.

Spreadsheets love to "help"

If a spreadsheet app opened your file, it may have turned long ids into numbers like 7.74E+12. PathGuide can't read those and tells you how many it found. Export the file again without opening it, or format the id column as text before saving.

What you can do with a saved cohort​

Every saved cohort has a row of buttons:

ButtonWhat it does
OpenShows the cohort's size, lifetime value and orders right on the page
Tag in ShopifyAdds the tag pathguide-cohort:<name> to up to 500 customers per click. Click Tag next customers to continue
Create Shopify segmentSaves the cohort as a real customer segment in Shopify (Customers → Segments), named "PathGuide: <cohort name>"
Sync to KlaviyoAdds up to 500 customers per click to the cohort's own Klaviyo list. Click Sync next customers to continue
Download CSVDownloads the cohort's customers as a spreadsheet
DeleteRemoves the cohort (asks you to confirm first)

Shopify segments, the short version. A live cohort becomes a rule in Shopify's own segment language, so Shopify keeps it current all by itself. A fixed list becomes "customers with this cohort's tag", so use Tag in Shopify first. After creating, PathGuide shows the rule and a link that opens the segment in Shopify.

Klaviyo, the short version. The first sync creates a list called "PathGuide: <cohort name>"; later syncs add to it. It never lands in your newsletter list by accident. The list is a copy from the moment you click, so sync again to add customers who joined the cohort since.

You are in charge of marketing consent

Sync to Klaviyo stays locked until you tick "I confirm that the customers I sync to Klaviyo have agreed to receive marketing from my store." PathGuide does not check consent for you, and it never sends an email itself. But any campaign or flow you attach to that list will reach these customers, so only sync people who said yes.

Playbook: four cohorts worth building this week​

  1. Win back the lapsed regulars. Use the Lapsed repeat buyers template, save it as live, sync it to Klaviyo, and attach a friendly "we miss you" flow. These people already like you.
  2. Nail the second order. Look up Median days to second order on Retention. Build First-time buyers, last 30 days and time a follow-up email a few days before that median.
  3. Thank your VIPs. Save VIP (5+ orders) as a Shopify segment and give them early access or a small surprise. Loyalty is cheaper to keep than to buy.
  4. Learn from the quiet ones. On the lifecycle chart, click last period's Went quiet bar and save it. Read a few of their orders: is it one product, one season, one shipping problem?
Technical details
  • Addresses: /app/audience-insights/retention and /app/audience-insights/cohorts. The older /app/retention and /app/cohorts redirect there.
  • Limits: up to 100 saved cohorts per store; a fixed list holds up to 10,000 customers (when more match, the 10,000 with the highest lifetime value are kept); names up to 80 characters; tag and Klaviyo actions handle 500 customers per click; uploads up to 5 MB.
  • Shopify segments use Shopify's segment query language (number_of_orders, amount_spent, last_order_date, first_order_date, customer_tags CONTAINS). Shopify allows at most 10 conditions per segment; the builder warns before you save a cohort that would need more. Shopify works out the numbers from its own all-time order data, so a segment's size can differ from the cohort's, especially when PathGuide holds less than about 180 days of your orders (the page warns you). Needs the write_customers permission.
  • Shopify tags end with part of the cohort's id (for example pathguide-cohort:vip-5-orders-1a2b3c), so two cohorts with look-alike names never share a tag.
  • Klaviyo: PathGuide writes the profile properties path_explorer_cohort and path_explorer_label and adds the profile to the cohort's list. Customers without an email address are counted and skipped.
  • Customer tags are kept as a copy (customer id and tags only) and updated by Shopify's customer and order webhooks, so cohort pages stay fast even when one tag covers most of your customers.
  • Order history: the background import uses a Shopify bulk export. Without Shopify's read_all_orders permission, Shopify limits it to the last 60 days; every order after install arrives by webhook.

Next steps​