Cohort Retention

What is this page?

The Cohort Retention page groups customers by the shop-calendar month of their first qualifying purchase and shows what percentage return in each later completed month. Qualifying purchases have status AWAITING_SHIPMENT, IN_TRANSIT, DELIVERED or COMPLETED; creator samples and return-order records are excluded. The heatmap helps you compare customers acquired in different months at the same stage after their first purchase.

When an exact view is not ready yet, the page says it is preparing the Cohort Retention view and will update automatically. You can leave and return later. If a complete earlier result is available, it stays visible while an updated view is prepared. Dashboardly does not mix partial results into that completed heatmap.

Key Metrics on This Page

Retention Rate

What it shows: The percentage of distinct customers in a cohort who made a qualifying purchase in a specific later completed calendar month.
Formula: (Customers who purchased in period N / Total customers in cohort) x 100
Important: Higher retention rates mean more customers are coming back. Even small improvements in retention can significantly increase total revenue over time.

Cohort Size

What it shows: The number of distinct customers whose first qualifying purchase falls in each cohort month. The first purchase is determined from available eligible history, not just the selected cohort window.
Important: Larger cohorts give more statistically meaningful retention rates. A very small cohort (e.g., 5 customers) can have volatile retention percentages.

Summary Cards

What it shows: Total Cohorts counts the displayed acquisition months. Avg Month 1 Retention, Avg Month 3 Retention and Avg Month 6 Retention summarize the corresponding completed months.
How averages work: Each summary is the arithmetic mean of eligible cohort percentages for that month, not a combined customer-weighted rate.
Important: Cohorts that have not completed the relevant month are excluded from its average. If no cohort is old enough, the summary is N/A, not zero. This page does not show an overall Repeat Purchase Rate card.

Features

Retention Heatmap

The main visualization is a color-coded table (heatmap):

Rows: Each row represents customers whose first qualifying purchase was in the same shop-calendar month.
Columns: M0 is the acquisition month; M1, M2, M3 and later columns are the subsequent calendar months, not rolling periods after each customer's purchase date.
Cell color: The blue/teal scale indicates lower to higher retention. Its appearance adjusts to the selected theme; read the percentage rather than treating a color-band label as an industry benchmark.
Cell value: The retention percentage for a completed month. N/A means that month has not finished yet, not that no customer returned.

Read the heatmap left to right. M0 is always 100% because every customer in the cohort made a qualifying purchase in its acquisition month. Later completed columns show the percentage who purchased in that specific month; a customer can return again after skipping a month.

Cohort Period

Use Cohort Period to choose how far back to analyze monthly cohorts:

3 Months -- analyze the recent three-month cohort window.
6 Months -- analyze the recent six-month cohort window; this is the default.
12 Months -- analyze the recent twelve-month cohort window.

Under Track Retention For, choose 3, 6, 9 or 12 months after first purchase; the default is 6. These controls change the analysis window and number of later months, not the monthly grouping. Weekly and quarterly grouping are not offered on this page.

Retention Trends

Compare the same M column across cohort rows to see whether retention at that age improved. Read across one row to see purchases in its later months. The current page uses the heatmap and summary cards, not a separate trend-line chart.

Cached Data

Use cached data is checked by default. Uncheck it to request a fresh calculation, which can take longer. A fresh calculation still cannot make an unfinished calendar month complete; those cells remain N/A.

Common Questions

How do I read the retention heatmap?

Each row is a group of customers who first bought in the same period. Read across the row to see what percentage came back. For example, if the "January 2026" row shows 15% in the "Month 2" column, it means 15% of customers who first bought in January 2026 made another purchase two months later.

What is a good retention rate?

There is no universal target established by this page. Compare your own completed cohorts at the same age, considering cohort size, product repurchase cycle and seasonality. Month 2 does not have to be lower than Month 1: customers may skip a month and return later. A small cohort can change sharply when only a few customers return.

Why is the first column always 100%?

M0 represents the cohort's acquisition month. Every included customer made their first qualifying purchase in that month, so its value is 100% by definition.

Why do newer cohorts have fewer columns?

Later cells are N/A until the entire corresponding shop-calendar month has ended. For a cohort acquired last month, M1 is still the current unfinished month and remains N/A. A completed month with no returning customers shows 0%; an unfinished month is not converted to zero or included in summary averages.

How is this different from the Purchase Journey page?

Purchase Journey shows which products customers buy in sequence. Cohort Retention shows whether they buy again in each later completed month, regardless of which products they choose. Journey answers "what do they buy next?"; retention answers "when do they return?"

How can I improve my retention rates?

Several strategies can help:

Follow-up communication -- use TikTok Shop Seller Center Buyer Messages or another consented channel in accordance with its rules. Do not send messages directly to the anonymized buyer email supplied by TikTok.
Product quality -- customers who receive quality products are more likely to return
Loyalty incentives -- offer returning customers exclusive discounts or early access to new products
Product variety -- consider relevant complementary products that give customers a reason to return; evaluate the result against your own cohorts.

What does it mean if newer cohorts have better retention?

This can be encouraging, but the heatmap alone does not prove which change caused it. Compare completed cohorts at the same month offset and check sample size, product mix and seasonality before attributing the difference to recent improvements.

First purchase product

This control is available for selected shops. If your Cohort Retention page shows First purchase product, choose Products or Product Variants (SKU) and search for a product name, variant name or seller SKU. The heatmap shows customers acquired through a matching item.

Each customer keeps the month of their first qualifying purchase. Later purchases of any product count toward retention. N/A means a later month has not finished yet and is excluded from the summary averages. Clear the search to restore the complete heatmap.