How to Find Which Customers Are About to Leave Your Store (Without Hiring an Analyst)

In 2022, I was managing eCommerce for a footwear brand. Sales looked fine on the surface — new orders coming in, revenue stable, nothing alarming in the dashboard.

Then I exported three years of customer data and ran some basic numbers.

Sixty-one percent of customers who had placed a second order never came back for a third. They were gone — silently, without a complaint, without an unsubscribe. Just... gone.

The data had been sitting there the whole time. Nobody was reading it.

That moment changed how I think about eCommerce. It is not enough to know how many people bought today. You need to know who is quietly disappearing from your customer base — before they are gone for good.

This guide will show you exactly how to find those customers. Whether you have 200 orders or 20,000, the principle is the same.

What Is an At-Risk Customer?

An at-risk customer is someone who has bought from your store before — but based on their behaviour, is very unlikely to buy again without some kind of intervention from you.

They have not formally "left." They have not complained. They probably still open your emails occasionally. But the window for winning them back is closing, and most store owners never act because they never identify who these customers actually are.

Here is a simple definition that works for most eCommerce stores:

A customer is at-risk when their last purchase was 91 to 365 days ago, or when they have placed two or more orders but their most recent one was more than 90 days ago.

This is different from a lost customer — someone whose last purchase was over a year ago, or a one-time buyer who has not returned in more than six months. At-risk customers are still reachable. Lost customers require a completely different strategy.

The distinction matters because your approach — the message, the offer, the urgency — should be different for each group.

Why Most Store Owners Never Catch This

The honest answer is that identifying at-risk customers manually is genuinely time-consuming. It requires you to:

Export your full order history Calculate the number of days since each customer's last purchase Segment customers by purchase frequency Cross-reference those two dimensions Do it again next month, because the list changes constantly

If you are running a store alone or with a small team, this kind of analysis falls to the bottom of the list every single week. There is always something more urgent — a campaign to launch, stock to reorder, a customer complaint to handle.

The result is that most stores operate blind. They spend money acquiring new customers while existing ones quietly disappear.

3 Signals That a Customer Is About to Churn

Before we get into the how, it helps to understand what you are actually looking for. At-risk customers tend to show the same patterns across very different types of stores.

Signal 1: The Gap Is Growing

A customer who used to buy every 45 days has not purchased in 90 days. The gap between their orders is widening. This is the earliest and most actionable signal — the customer is still within reach, but the window is closing.

For example: a beauty brand customer who was ordering skincare products monthly has now gone quiet for three months. She has not unsubscribed. She is probably buying the same products somewhere else.

Signal 2: Decreasing Order Value

A customer who spent £80, then £60, then £35 across three orders is showing declining engagement. Each purchase is a smaller commitment. This is often a sign that they are testing alternatives — your store becomes one of several rather than the default.

Signal 3: The Second Order Never Came

This is the most common pattern in eCommerce and the one most store owners underestimate. A customer placed one order six months ago and never returned. Statistically, a customer who has not made a second purchase within 90 days of their first is unlikely to do so without a specific trigger.

In most stores I have analysed, between 45 and 65 percent of customers fall into this category. They came, they bought once, and they never came back.

How to Find At-Risk Customers Manually

If you want to do this yourself, here is the process. I will be honest: it takes two to three hours the first time, and you will need to repeat it every month.

Step 1: Export Your Order History

Every major eCommerce platform allows you to export a CSV of your order history. You need at minimum three columns:

Customer email address Order date Order value

In Shopify: go to Orders → Export → All orders → CSV for Excel. In WooCommerce: use the built-in export under WooCommerce → Orders → Export. In BaseLinker: go to Orders → Reports → Export to CSV.

Step 2: Open in Excel or Google Sheets

Once you have your CSV, open it and create a new column called Days Since Last Order. Use this formula:

=TODAY() - MAX(IF(A:A=email, B:B))

This calculates how many days have passed since that customer's most recent order.

Step 3: Count Orders Per Customer

Add another column: Total Orders. Use COUNTIF to count how many times each email address appears in your data:

=COUNTIF(A:A, email) Step 4: Apply Your At-Risk Filter

Now filter your data to show customers where:

Days Since Last Order is between 91 and 365 OR Total Orders is 2 or more AND Days Since Last Order is over 90

This is your at-risk segment.

Step 5: Sort by Value

Sort this list by total customer value (sum of all their orders). Start with your highest-value at-risk customers — these are the ones worth reaching out to personally.

The Faster Way: 60 Seconds With RetentionShift

I built RetentionShift because I got tired of doing the above process manually every month.

The tool does everything described above — and more — in about 60 seconds. You upload the same CSV you would export for the manual method, and instead of spending three hours in Excel, you get:

Your overall retention score (0-100) Three customer segments: Loyal, At Risk, and Lost The number of customers in each segment The revenue at risk from your at-risk and lost segments Specific win-back recommendations for each segment, including email templates with subject lines ready to send

The analysis uses the same logic I described above — recency, frequency, and value — but applied automatically to your entire customer base, with no formulas and no pivot tables.

You can try it for free. The free plan includes one full analysis per month for up to 100 customers.

What to Do With Your At-Risk Customers

Finding them is only the first step. Here is what to do once you have the list.

Action 1: Send a Personalised Win-Back Email

Do not send a generic newsletter. At-risk customers need a message that acknowledges the time that has passed and gives them a specific reason to come back.

A simple structure that works:

Subject line: We noticed you haven't been back in a while, [First Name]

Email body:

Hi [First Name],

It has been a few months since your last order, and we wanted to check in.

We have added new products since you last visited, and we wanted to make sure you did not miss them.

As a thank you for being a customer, here is 15% off your next order: [DISCOUNT CODE]

This offer is valid for the next 7 days.

[Shop Now]

Keep it short. Keep it personal. One clear call to action.

Action 2: Make the Offer Time-Limited

Discounts without urgency do not work. Add a deadline — seven days is enough to create genuine urgency without feeling manipulative.

"Valid for the next 7 days" consistently outperforms "use whenever you like."

Action 3: Segment by Value Before You Send

Do not give your highest-value at-risk customers the same offer as your lowest-value ones. A customer who spent £800 with you deserves a personal email from you directly — not an automated sequence.

Segment your at-risk list into three tiers:

High value (top 20% by spend): Personal outreach or premium offer — free shipping plus 20% off Mid value: Standard win-back email sequence with a 15% discount code Lower value: Newsletter re-engagement, no discount needed How Often Should You Check?

Once a month. Customers move between segments — someone who was at-risk last month might be lost this month if you did not act.

Building a monthly habit around this data is one of the highest-ROI activities a store owner can do. You are not acquiring new customers — you are recovering revenue from people who already trusted you enough to buy once.

Summary

At-risk customers are the most overlooked growth lever in eCommerce. They have already bought from you. They know your brand. They just need a reason to come back — and in most cases, a timely, personalised message is enough.

Here is what to do this week:

Export your order history CSV from your eCommerce platform Either run the manual Excel analysis above, or upload to RetentionShift for a 60-second report Identify your top 20 at-risk customers by spend Send them a personal win-back email this week

The customers are there. The data is there. You just need to read it.

Want to find your at-risk customers automatically? Upload your CSV at RetentionShift and get a full retention report in 60 seconds — free.