Most small businesses find out a customer has churned after they are already gone. The subscription lapses. The repeat purchase never comes. The email bounces. By then, you are reacting, not preventing.
The fix is not a better goodbye email. It is a retention dashboard that flags at-risk customers before they cancel, so you can act while it still matters.
Here is how to build one using tools you probably already have.
Why You Need a Retention Dashboard
Subscription businesses lose roughly 5 to 7 percent of customers per month on average. That is not dramatic. It is a slow bleed. The problem is that churn hides in plain sight. Unless you are tracking the right signals, you do not notice until the revenue drops.
A retention dashboard pulls those signals into one place so you can see who is drifting, who is engaged, and who is about to leave. It turns guesswork into a system.
The 5 Metrics to Track
Before you build anything, you need to decide what to measure. These five metrics are the ones that actually predict churn for small businesses.
1. Purchase Frequency Decline
Track the average days between purchases for each customer. When that number starts increasing, the customer is pulling away. Someone who bought every 30 days and suddenly stretches to 45 is signalling disengagement before they tell you anything.
Set a threshold: if purchase interval increases by more than 30 percent compared to their personal average, flag them.
2. Email Engagement Drop
If a customer who used to open your emails suddenly stops, that is a churn signal. Track open rates and click rates per customer over the last 90 days. A decline of more than 50 percent from their personal baseline is a warning.
This is not about your overall open rate. It is about individual behaviour change.
3. Support Ticket Spike
Customers who submit more than two support tickets in a single month are either having a real problem or losing patience. Either way, they are closer to the exit than the entrance.
Track ticket volume per customer per month. Flag anyone with more than two tickets in a 30-day window.
4. Session Duration Drop
If you have analytics on your site, track how long each logged-in customer spends browsing. A customer who used to browse for 3 minutes and now bounces after 20 seconds is not exploring. They are checking a price, comparing you to a competitor, or confirming a detail before they stop shopping entirely.
5. Subscription Downgrade
If you offer tiers, a downgrade is the clearest signal. They are not gone yet, but they are trying to spend less. This is your window to reach out, ask why, and fix the issue before they cancel entirely.
How to Build the Dashboard
You do not need a custom app. You can build this in a spreadsheet connected to your store data, or in a tool like Google Looker Studio that pulls from Shopify, your email platform, and your helpdesk.
Step 1: Export Your Customer Data
Start with a CSV export from your store platform. You need:
- Customer name and email
- Total orders and order dates
- Average order value
- Current subscription tier (if applicable)
- Last purchase date
For Shopify stores, you can export this from Customers in the admin panel. For other platforms, check your orders or customers export.
Step 2: Add Engagement Data
Next, layer in email engagement. Most email platforms (Klaviyo, Mailchimp, Omnisend) let you export per-contact engagement data. Match it to your customer list by email address.
Add these columns:
- Emails sent (last 90 days)
- Emails opened (last 90 days)
- Emails clicked (last 90 days)
- Open rate trend (up, flat, down)
Step 3: Add Support Data
If you use a helpdesk (Gorgias, Zendesk, Freshdesk), export ticket counts per customer. Match by email. Add:
- Tickets submitted (last 30 days)
- Tickets submitted (last 90 days)
- Average resolution time
Step 4: Calculate Risk Scores
Create a simple risk score using conditional logic. A basic version:
- +2 points if purchase interval increased more than 30%
- +2 points if email engagement dropped more than 50%
- +1 point if customer submitted more than 2 support tickets in 30 days
- +2 points if subscription was downgraded this month
- +1 point if last purchase was more than 60 days ago
Anyone scoring 3 or above is at risk. Anyone scoring 5 or above is likely to cancel within 30 days.
Step 5: Visualise and Set Alerts
Plug your data into Google Looker Studio or a similar free tool. Create a table view sorted by risk score, highest first. Add conditional formatting: red for 5+, orange for 3-4, green for 0-2.
Set up a weekly email or Slack notification that lists the top 10 at-risk customers. This turns passive data into an action trigger.
What to Do With At-Risk Customers
Once your dashboard flags someone, do not send a generic “we miss you” email. That is what everyone does, and it does not work.
Instead:
- If purchase frequency dropped: Send a personalised offer based on what they used to buy. “Your last order of [product] was 6 weeks ago. Here is 15 percent off your next one.”
- If email engagement dropped: Change your cadence. Send fewer emails, but make them more relevant. Ask what they want to hear about.
- If support tickets spiked: Reach out proactively. “I saw you had a few issues last month. I want to make sure everything is sorted. Here is what we fixed.”
- If they downgraded: Ask why. A short survey with 3 options gives you data and shows you care.
The point is not the dashboard itself. It is what the dashboard lets you do: intervene early, personally, and with something relevant.
Maintenance
Update your data export weekly. Check the dashboard every Monday. Set a calendar reminder if you need to.
Every month, review your risk thresholds. If too many false positives show up, raise the bar. If you are missing real churners, lower it. The dashboard gets better the longer you use it.
The Bottom Line
Churn is not a surprise. It is a pattern that shows up in your data weeks before the customer cancels. A retention dashboard makes that pattern visible so you can act on it.
Start with a spreadsheet. Track purchase frequency, email engagement, support tickets, session duration, and downgrades. Calculate a simple risk score. Sort by risk. Reach out to the people at the top of the list.
You do not need a fancy tool. You need a system that flags the right people at the right time, and a process for doing something about it.
Want more like this? Join MOKU Club for free. Weekly resources, early access to new guides, and occasional templates you can actually use. Join below.
