Retention 6 min read · · Last updated:
By Mark Ashworth · Founder, ChurnTools

How to Write a Cancellation Save Offer That Works

A good save offer matches the cancellation reason. A bad one is the same 20% discount for everyone. Here is how to write save offers that actually save customers, by reason, with examples you can steal.

📊

Want a personalized score for your situation?

Take the free 60-second Churn Health Check

Score me →

Most cancellation save offers are the same 20% discount, no matter who's leaving or why. That offer saves 5-10% of cancellation attempts. A reason-matched save flow saves 15-25%. Same traffic, 2-3x the impact.

Here's how to write save offers that actually work, by reason, with templates you can steal.

The principle: match offer to reason

The single biggest unlock is asking why before offering. Make cancellation reason a required field (5-6 dropdown options, not free text), then route to a different save based on the answer. The same customer leaving for "too expensive" needs a completely different offer than the customer leaving for "missing a feature."

The seven most common cancellation reasons (in rough order of frequency):

  1. Too expensive
  2. Not using it enough
  3. Missing a feature
  4. Switching to a competitor
  5. Don't need it right now
  6. Bad support experience
  7. The champion left the company

Each one needs a different play. Here are the templates that work.

"Too expensive"

What works: Tiered response based on tenure and usage.

Template for high-tenure, high-usage customers:

"You've been with us 14 months and ran 247 campaigns this year. We don't want to lose you over price. Here's 40% off for the next 6 months while you decide if it's worth full price again. No commitment, no auto-renew at the higher rate. Stay if it works."

Template for short-tenure or low-usage customers:

"Our Pro plan might be more than you need right now. Our Starter plan is $X less per month and includes everything you've actively used so far. Want to downgrade instead of cancel?"

What to avoid: Offering 40% off to everyone. You'll teach customers that cancelling = discount, and you'll burn margin on customers who would have downgraded for free.

"Not using it enough"

What works: Pause subscription + value reminder.

Template:

"In 8 months with us, you've built 127 saved searches, 14 reports, and 3 custom dashboards. If you cancel, all of that disappears in 30 days. Want to pause for 1-3 months instead? Your data stays intact, your team's workflows don't break, and you pick up exactly where you left off."

What to avoid: Offering a discount. They're not leaving because of price, they're leaving because they aren't getting value. A cheaper version of "not using it" is still "not using it."

"Missing a feature"

What works: Check if the feature exists (it often does), show the roadmap if planned, ask for specifics.

Template:

"You mentioned missing [feature]. Two questions before you go: (1) have you tried [actual existing feature that's similar]? Quick 90-second video here. (2) If it's not on our roadmap, would you stay 60 days while our team reviews building it? We respond to every feature request from cancelling customers within 5 business days."

What to avoid: Promising to build features you won't build. That comes back as a trust issue when you reach out 6 months later.

"Switching to a competitor"

What works: Personalized comparison + acknowledge what they may be missing.

Template:

"Sounds like you're evaluating [competitor]. Quick read before you switch: you currently use [features A, B, C]. Here's how those work in [competitor]: [specific differences]. If it's still the right move, we'll help with the migration. If not, here's a [free month / consultation / Pro plan upgrade] while you decide."

What to avoid: Trash-talking the competitor. It signals desperation and lowers the trust they had in you. The comparison should be factual.

"Don't need it right now"

What works: Pause with automatic reactivation. The single highest-converting save type.

Template:

"Sounds temporary. Want to pause for 1-3 months instead? Your account stays, your data stays, and we'll send you a reminder when the pause ends so you can decide whether to reactivate or cancel for real."

What works about this: 30-40% of "permanent" cancels are actually temporary. A pause option recovers them automatically.

"Bad support experience"

What works: Acknowledge specifically what went wrong + concrete recovery.

Template:

"You opened a ticket on [date] about [topic] and it didn't get resolved. That's on us. Before you cancel, can we connect you with our team lead for a 15-minute call to fix it? Plus a service credit on next month's invoice. If you still want to cancel after that, no pushback."

What to avoid: Generic "Sorry you had a bad experience." Specifics matter. Pull the actual ticket history into the save flow.

"Champion left the company"

What works: Re-onboarding for the new owner.

Template:

"Looks like [original champion] moved on. We see this often. Want a 30-minute concierge re-onboarding for whoever's taking over? We'll cover their use cases, custom setup, and answer questions specific to their goals. Free, no commitment."

What about every reason

If you can only ship one save flow, use this: pause subscription as the default option for every reason, with reason-specific add-ons. Pause beats cancellation 40-50% of the time across all reasons because most cancellations are situational, not permanent.

Don't make cancellation harder

Adding friction to cancellation hurts more than it helps. Customers who feel trapped churn harder, write bad reviews, and tell others. The save flow should feel helpful, not adversarial. Make canceling easy, but make staying obviously the better option through good offers.

For the full save flow architecture, see AI cancellation save flows and the save flow MVP experiment.

Score your cancellation flow

Take the 60-second Churn Health Check. It scores your save flow maturity (including whether you match offers to reasons) and tells you what to fix first.

Free interactive tool

Score your retention setup in 60 seconds

8 questions. Get your tier (Critical to Best-in-Class), your weakest spots, and 3 specific things to fix next.

Take the Health Check

Frequently asked questions

Answers to the questions I get most often about this topic.

What is a good cancellation save offer?

A good save offer matches the cancellation reason. For price concerns: tiered discount or downgrade option. For low usage: pause subscription or guided walkthrough. For missing features: roadmap preview or beta access. For switching to competitor: personalized comparison and migration help. Generic 20% off for everyone saves 5-10%; reason-matched offers save 15-25%.

Should you offer a discount to canceling customers?

Only as one option among several, and only for price-sensitive cancellations. Offering everyone a discount teaches customers that canceling is the path to a deal, trains future churn, and wastes margin on customers who would have stayed for a non-monetary save. Match the offer to the reason.

How much discount should a save offer include?

For long-tenure, high-LTV customers: 25-40% off for 3-6 months. For short-tenure or low-LTV customers: start with downgrade or pause options before any discount. Always cap the discount window so it expires; permanent discounts create margin problems and signal that your normal pricing is wrong.

What words should you use in a save offer?

Lead with acknowledgment ("Sounds like price is the issue"), not denial ("Are you sure?"). Use specific data from their account ("In 8 months you ran 47 campaigns") to remind them what they've built. Avoid guilt ("You'll lose access to everything"). Frame the save as helpful, not adversarial.
MA

Written by Mark Ashworth

Founder of ChurnTools. I spend my time studying how SaaS companies lose customers and building tools to help them stop. Previously worked in SaaS growth and retention across multiple B2B products.

Ready to run your first retention experiment?

Browse 30+ proven playbooks for reducing churn across every stage of the customer lifecycle.

Browse Experiments →