What are the churn warning signs?

Churn does not come out of nowhere. There are five warning signs that show up in your data weeks or months before a customer leaves. Here's what to look for and how to find them.

7 min read

The short answer

The five most reliable churn warning signs are: declining order frequency, shrinking order size, increasing gaps between purchases, late or missing payments, and reduced engagement with your team. The first three are visible in your accounting data. The last two require CRM or communication data.


Why understanding churn signals lets you prevent churn, not just measure it

Most businesses measure churn after it happens. They count lost customers, calculate a rate, and move on. That is useful for tracking trends but does nothing to save the customers who are leaving right now.

If you know what churn looks like before it happens, you can intervene. A customer whose order frequency dropped by 30% over the last two months might just need a phone call. A customer who has not responded to your last three emails might need a different approach. The warning signs give you a window of time to act.

The economics are straightforward: preventing one customer from churning saves you the full cost of replacing them, which is typically 5-7 times the cost of retaining them.


The five churn warning signs in order of reliability

1. Declining order frequency (most reliable). A customer who used to order every 3 weeks is now ordering every 5 weeks. Their cadence is stretching.
2. Shrinking order size. Average order value has dropped 20%+ compared to their historical baseline. They are buying less even when they do buy.
3. Growing gap since last purchase. The time since their last order exceeds 1.5x their typical purchase interval. They are overdue.
4. Late or missing payments. A customer who always paid within 15 days is now at 45 days. Payment behavior often deteriorates before purchasing stops.
5. Reduced engagement. Fewer emails opened, fewer support tickets, fewer meetings scheduled. Silence is often the final warning sign.

Signals 1-3 come from your accounting data. Signal 4 comes from your accounts receivable. Signal 5 requires CRM or email data. The more signals you can monitor, the earlier you catch the problem.


How to check for churn warning signs in QuickBooks Online

  1. 1
    Go to Reports → Sales by Customer Detail

    Set to the last 6 months. Export to CSV. This gives you every transaction with dates and amounts by customer.

  2. 2
    Build a customer-level summary in a spreadsheet

    For each customer, calculate: number of transactions per month, average transaction size, and days since their most recent transaction.

  3. 3
    Compare recent behavior to baseline

    For each customer, compare their last 2 months of behavior to their 6-month baseline. Flag anyone whose frequency dropped by 25%+ or whose average order size dropped by 20%+.

  4. 4
    Check receivables for payment delays

    Go to Reports → Accounts Receivable Aging Summary. Any customer who is newly appearing in the 31-60 or 61-90 day columns (and was not there before) might be disengaging.

  5. 5
    Cross-reference with your CRM (if available)

    Check HubSpot, Salesforce, or your CRM for engagement data. In HubSpot, go to Contacts → filter by last activity date. Customers with no recent activity who are also showing declining purchase patterns are highest risk.

Total time: 45-60 minutes. Heavy spreadsheet work, multiple data sources, and manual comparison against baselines. This is realistically a quarterly exercise, not a monthly one.


How to check for churn warning signs in Xero

  1. 1
    Export invoices from Business → Invoices

    Set to the last 6 months. Export to CSV.

  2. 2
    Build frequency and size analysis per customer

    Same spreadsheet work as QuickBooks: transaction count per month, average invoice amount, and days since last invoice.

  3. 3
    Check Aged Receivables for payment warning signs

    Go to Accounting → Reports → Aged Receivables. Cross-reference newly overdue customers with the declining frequency list.

Total time: 45-60 minutes. Same analytical burden as QuickBooks, with the added step of building customer summaries from raw invoice data.


Why proactive churn detection is nearly impossible manually

  • 45-60 minutes minimum, and that is per analysis. The spreadsheet work alone takes nearly an hour. And it needs to happen every month to be useful.
  • You need to compare current behavior to each customer's own baseline. Industry averages do not work here. A customer who always orders every 6 weeks is not at risk because they have not ordered in 5 weeks. But a weekly buyer who has not ordered in 3 weeks is. The baseline has to be per-customer.
  • Data lives in multiple systems. Transaction data is in your accounting tool. Engagement data is in your CRM. Payment behavior is in your AR reports. Pulling all three together manually for every customer is not practical.

Or get churn warning signs detected automatically

Bottomline monitors all five warning signs across your accounting data, CRM, and payment records. Each month, it compares every customer's current behavior to their individual baseline and flags anyone showing churn risk signals.

Customers showing churn risk signals
Metro Supply Co.3 signalsHigh$4,100/mo
Greenfield Builders2 signalsMedium$2,800/mo
Sunrise Bakery2 signalsMedium$1,600/mo
$8,500/mo in revenue from customers showing active churn risk. Metro Supply Co. has declining frequency, shrinking orders, and a late payment.
From a real Bottomline report. At-risk customers are flagged before they churn, giving you time to intervene.

This is the difference between a churn report and a churn prevention system. The report tells you who left. Bottomline tells you who is about to leave, with enough time to do something about it.

Get your answer. Every month, automatically.

Connect your accounts in 5 minutes. Your first report arrives within 24 hours.

Works with QuickBooks, Stripe, HubSpot, Google Ads, and more
© 2026 Bottomline