When a hotel cancels, someone fills in a reason. Price. Moved to a competitor. Budget cuts. New management. Hotel sold. Those reasons go into a report, the report goes to the leadership meeting, and the company decides it has a pricing problem, a competitive problem or bad luck.

In a lot of hotel tech companies I have looked at, a meaningful share of those cancellations were decided months earlier, in the first weeks after the contract was signed. The churn was real. The reason on the form was not.

The attribution error

Churn has a delay built into it. A property that never really got going does not cancel the next day. It carries on paying, using the product lightly or not at all, until something forces a decision: a renewal date, a budget review, a new GM, a cheaper offer.

At that point, the property gives the reason that is easiest to say. Price is the easiest of all, because nobody can argue with it. But a hotel getting obvious value from a tool rarely leaves over a modest price gap. Price becomes the reason when value was never established.

I wrote about the month three churn pattern before. This is the same problem seen from the other end. The cancellation shows up at month twelve, but the seed was planted at month one.

What onboarding failure looks like in a hotel

Onboarding failure in hospitality is rarely dramatic. It looks like a setup that is ninety percent done, with the last integration never connected because nobody at the property had the credentials. It looks like a training session delivered to the day shift while the night team never saw the product. It looks like a GM who signed, handed the project to a busy department head, and moved on.

Properties without IT staff are especially exposed, a point I covered in designing onboarding for properties with no IT staff. When the vendor assumes technical capacity the hotel does not have, the setup stalls quietly, and nobody raises a flag because nobody owns it.

Why vendors miss it

There are three reasons. First, onboarding and retention are often owned by different people, so nobody sees the whole journey. Second, go-live is treated as the finish line, when it is really the starting line for value. Third, there is rarely a clear definition of what an activated property looks like, so there is nothing to measure against.

Without that definition, a property that logged in twice and one that runs its whole morning on the product look identical in the account list. Both are green until one of them cancels.

How to test whether this is your problem

Pull your churned accounts from the last year and your retained accounts from the same signing period. For each, look at a few onboarding markers: how long it took to go live, whether key integrations were completed, how many staff were active in the first month, and whether the property used the core workflow that delivers value.

If churned accounts cluster on the wrong side of those markers, you have your answer. The cancellation reasons may say price or competitor, but the pattern says the property never activated. Instrumenting activation properly is what turns this from a one-off analysis into an early warning.

What to change once you see it

If the analysis points to onboarding, resist the urge to add more steps. Most hotel onboarding problems come from too much being asked of a property that has too little time. Start by defining the one or two milestones that separate properties that stay from properties that leave, such as a completed integration or the core workflow running for a full week.

Then make those milestones someone's job on both sides. On your side, a named person responsible for getting each new property there by a set day. On the property's side, a named person who owns the setup, not just the GM who signed. Check progress at fixed points, for example day seven and day fourteen, and escalate when a property stalls instead of waiting for them to ask for help.

Finally, change how you record churn. Keep the reason the hotel gives, but add the onboarding markers alongside it. Within a couple of quarters, your churn report will tell you something you can act on.

Consider a guest experience platform

Picture a hypothetical guest experience platform that sees churn rising and hears price as the most common reason. The leadership team debates a discount programme for renewals.

Before doing that, customer success compares churned and retained properties. Most churned properties never connected their PMS, so staff were copying guest data by hand, and usage dropped after the first month. Retained properties connected in the first two weeks.

A discount would have kept some of those properties for another year while they continued not using the product. Fixing the integration step during onboarding, with a named owner at the property and a check at day fourteen, addresses the cause.

Reading your churn differently

None of this means price and competitors never matter. They do. It means cancellation reasons are the start of an investigation, not the end of one. Before changing pricing or positioning in response to churn, check whether the properties leaving ever reached the point where the product was doing its job.

If you are not sure whether your biggest leak sits in activation or somewhere else, the 12-point leak audit takes about twenty minutes and gives you a first ranking of where revenue is escaping.

Frequently Asked Questions