Most hotels do not announce that they are leaving. The general manager stops opening the dashboard, the front desk drifts back to a spreadsheet, and three months later the renewal email gets a polite "we have decided to go another way". The decision was visible in your product data long before it reached your inbox.

This list sets out ten usage signals that tend to show up before a hotel churns, what to measure for each, and the first response that fits. It is written for founders, customer success leads and revenue leaders at hotel software companies. If you already run a health score, use it as a check on what your score is missing; if you do not, start with our guide to building a customer health score for hotel properties.

Why usage beats sentiment as a warning

Surveys and QBR moods lag. Usage moves first. Gainsight describes a customer health score as consolidating product usage, support history, NPS and engagement into one score, and its worked example weights usage at 40%, support trends at 25%, sentiment at 20% and executive engagement at 15%. Gainsight's DEAR framework (Deployment, Engagement, Adoption, ROI) makes the same point: deployment and adoption gaps are early risk signals you can act on before issues escalate.

The stakes are higher than they look. The SaaS Capital 2025 survey of private SaaS companies puts median net revenue retention at 102% for companies with annual contract values of $25,000 to $50,000, with the lowest quartile at 97%. For a hotel tech vendor whose average property pays far less than that, a few silent churns a quarter can wipe out the expansion revenue. Our post on net revenue retention for hotel software works through the maths.

Ten usage signals that predict a hotel will churn

1. Logins fall below the property's own baseline

Absolute login counts mislead in hotels. A 12-room guesthouse and a 300-room city hotel have nothing in common on that measure. Compare each property with its own trailing average instead. A sustained drop of a third or more over four weeks, outside a known seasonal closure, deserves a call.

2. Active users shrink to one person

When a property goes from five active users to one, your product now depends on a single employee. Hotels lose staff fast: the US Bureau of Labor Statistics JOLTS data shows a quits rate of 3.5% for accommodation and food services in August 2026 (preliminary), against 2.1% for total private industry. That is a monthly figure. When the one remaining user leaves, the knowledge leaves with them. See how front desk turnover breaks software adoption.

3. The core workflow stops running through your product

Every hotel product has one or two jobs that define its value: rate pushes for a channel manager, check-ins for a PMS, price acceptance for an RMS, replies sent for guest messaging. The Pendo Feature Adoption Report, based on usage across 615 subscriptions, found an average of 12% of features generate 80% of daily usage volume, while 80% of features are rarely or never used. Track the 12%, not the rest. If price recommendations are viewed but no longer accepted, the hotel has already stopped trusting the product.

4. Exports to spreadsheets climb

A rise in CSV and report exports often means staff are rebuilding the workflow outside your product. Sometimes it is harmless (an owner wants a monthly file). Often it is a workaround for a missing view or a slow screen. Ask what they do with the export.

5. Integrations break and stay broken

A disconnected PMS, a failing payment gateway or a channel mapping error that sits unresolved for days tells you two things: the hotel is suffering, and nobody on their side is chasing you to fix it. Reliability matters to hoteliers. Hotel Tech Report reports that 48% of hotels would switch PMS vendors over reliability issues and 42% would leave due to cybersecurity concerns.

6. Feature breadth narrows

A property that used reports, automations and the mobile app in month three but only the calendar in month nine has quietly downgraded itself. Breadth narrowing usually precedes a request to move to a cheaper plan or a competitor that "does the basics".

7. The decision maker stops logging in

Front-line users keep the product running; the general manager, revenue manager or owner renews it. When the senior login goes dark for six weeks, the person who signs the renewal no longer sees value first-hand. This is the usage version of losing your sponsor, and it often follows a management change. Our post on what happens when a hotel changes ownership covers what to do when that happens.

8. Support goes quiet after being noisy

Falling ticket volume looks like good news. It is not, if usage is falling at the same time. A property that raised ten tickets a month and now raises none has often given up asking. Look at tickets alongside usage, never alone. Tickets asking how to export all data or close an account are a late, explicit signal.

9. Seasonal settings are not updated

Hotels that are engaged update rates, restrictions, room types, packages and message templates as seasons change. A property still running last summer's settings in October is either closed or no longer managing through you.

10. Seats removed and billing friction

User removals, a failed card that nobody fixes, a request for an invoice copy "for the owner" and a question about notice periods are commercial signals that usually appear last. By this point you are negotiating, not preventing.

The signals at a glance

How to use the list without drowning in alerts

• Pick three signals first. Logins against baseline, active users and the core workflow cover most early risk. Add the rest once those alerts are trusted.

• Adjust for seasonality. Flag known closures and low seasons so a Swiss ski hotel in May does not trigger a churn alert. Our post on reviewing pipeline through hotel seasonality shows the same logic on the sales side.

• Combine signals. One signal is noise; two or three on the same property in the same month is a pattern.

• Give every alert an owner and a play. An alert without a named response becomes a dashboard nobody opens.

• Check your score against real churn. Every quarter, look at who actually left and whether the signals fired. Reweight the ones that did.

Usage signals only help if they reach someone with time to act, which is why the renewal is usually decided months before the renewal date. Retention is the expand and measure stage of the system laid out in our hotel tech GTM strategy guide, and many churn problems start much earlier, in onboarding; our post on instrumenting activation covers the events you need to capture from day one.

If churn is eating into growth and you are not sure whether the leak starts in onboarding, adoption or renewal, the free GTM leak audit will show where properties start to drift.

Frequently Asked Questions