Cohorts

Measure how well you retain visitors over time.

Overview

A cohort is a group of visitors who share the same acquisition date — the day (or week, month, or year) they first visited your website. The Cohorts report compares these groups side by side over time, showing you how many visitors come back and how their engagement evolves. It’s the clearest way to answer: are we keeping the people we acquire?

There’s nothing to set up — cohorts are built automatically from your existing tracking data. Find the report under Visitors → Cohorts in the left reporting menu.

Reading the Cohorts Report

The report is a retention matrix:

• Each row is a cohort — visitors whose first visit happened in that row’s period (e.g., the week of March 2)
• Each column is a period after the cohort first appeared — Day 0 is the acquisition day itself, Day 1 the day after, and so on (or Week 0/1/2…, Month 0/1/2… depending on the selected period)
• Each cell shows the selected metric for that cohort in that period — by default, Returning Visitors %: the percentage of the cohort that visited again during that period

Deeper color shading highlights higher values, so retention patterns (and drop-offs) stand out at a glance. Newer cohorts have fewer elapsed periods, which gives the matrix its staircase shape. The accompanying graph plots the same data as one line per cohort.

Choosing Period and Metric

• Period — Cohorts follow the period selector: choose day, week, month, or year at the top of the page to group cohorts accordingly
• Metric — Use the dropdown at the top right of the report to switch from Returning Visitors % to other metrics, including visits, conversion rates, and revenue (overall or per goal). Your metric choice is remembered for next time.
• How many cohorts — Adjust the limit selector to show more or fewer cohorts (ten are shown by default)

Note that user-based metrics are only available for day and week periods.

Drilling into a Cohort

Every insight in the matrix can be investigated:

1. Click the icon on a cohort’s row label to open the segmented visits log for exactly those visitors
2. From there, open the segment editor pre-filled with the cohort’s definition (based on the visitor’s first-visit date)
3. Save that segment and apply it to any report — see what a specific cohort searched for, which pages they read, or what they converted

You can also apply your existing segments on top of the Cohorts report — for example, cohorts of mobile visitors only, or cohorts of visitors from a specific campaign.

Use Cases

Measure the Impact of Changes

Released a redesigned patient portal or a frequently requested feature? Compare cohorts acquired before and after the release date — if the change worked, later cohorts should retain better.

Diagnose Engagement Drops

If a specific cohort retains far worse than its neighbors, look at what happened on its acquisition date: a campaign that attracted the wrong audience, content that didn’t match expectations, or a site issue that day.

Compare Acquisition Channels

Apply a segment (e.g., visitors from paid search vs organic) to see which channels bring visitors who actually come back — not just visitors who bounce once.

Healthcare Retention Patterns

• Do patients acquired during an awareness campaign return to book appointments?
• Does a new-patient portal onboarding change how often patients come back?
• Which service-line content builds a returning audience vs one-off traffic?

Best Practices

• Match the period to your visit cycle — weekly or monthly cohorts suit most healthcare sites; daily cohorts suit high-frequency products
• Compare like with like — seasonal effects hit all cohorts; judge a cohort against its neighbors, not against a cohort from another season
• Give cohorts time — recent cohorts have few elapsed periods; don’t judge retention from Day 1 alone
• Pair with campaigns — check cohorts after every major acquisition push to measure quality, not just volume

Limitations

• Cohorts are defined by acquisition date; for other groupings (e.g., “everyone who saw page X”), build a segment and apply it to the report
• Recognizing a returning visitor relies on the standard visitor identification your tracking uses — privacy features that limit cross-visit recognition also limit cohort accuracy
• Reports exist from when the feature was activated onward; if you need cohorts computed for older historical data, contact support to re-process past periods

Next Steps

• Visitor Reports — Understand who your visitors are
• Funnels — Analyze the paths that drive conversions
• Custom Reports — Build your own retention views