Reducing SaaS Churn: Strategies, Tools and What They Cost
A working guide to cutting SaaS churn: failed-payment recovery, cancellation flows, usage-based warning signs and win-back emails, with the plans and prices of the tools that handle each job.
For most SaaS teams, the fastest way to cut churn is to recover failed payments before they turn into cancellations. For the email side of that job, a sensible starting point for a small team is Sequenzy: its free plan covers 2,500 emails a month, and the 15,000-email plan costs $19 a month, or $205 a year billed annually. The reason is specific to this problem. Sequenzy is built around billing events. Its payment-failed trigger, saas.payment_failed, includes a hosted invoice URL, so a recovery email can link the customer straight to a page where they can pay or update their card. It connects to Stripe, Paddle, Lemon Squeezy, Polar, Creem and Dodo Payments.
Failed payments are only one of the three ways customers leave, and no single tool covers all three. This guide sorts churn by cause, names the tools that address each cause, and gives the plan names, prices and billing views for each one. Where a vendor does not publish a price, the page says so and does not estimate one. Prices come from each vendor's public pricing page, checked on 7 October 2026. Plans change, so confirm the figure on the linked page before you put it in a budget.
What churn is made of
Churn is the share of customers or revenue you lose over a period. Logo churn counts customers. Revenue churn counts the money those customers paid. The two can point in different directions: losing one large account can hurt revenue churn far more than losing ten small ones hurts logo churn, so track both.
The arithmetic is simple. If you start a month with 1,000 customers and 50 cancel, monthly logo churn is 50 ÷ 1,000 = 5%. Many people then multiply by twelve to get 60% a year, but that overstates the loss because churn compounds. Keeping 95% of customers each month for twelve months leaves 0.95 × 0.95 × … (twelve times) ≈ 54% of the starting group, so about 46% of the original customers are gone after a year. That is still a large number, and it shows why a few points of monthly churn matter.
Split churn into three groups before you pick any tool:
- Involuntary churn: the customer wanted to stay, but a card expired, was declined or was replaced without an update. Billing and dunning emails address this directly.
- Usage-driven churn: the customer stopped using the product, so the subscription looked unnecessary. Product analytics and re-engagement emails address this.
- Decision churn: the customer chose to leave because of price, a missing feature, a poor fit or a change in their own business. Cancellation flows, plan changes and exit surveys address this.
Each group needs a different response, and treating them as one problem is the most common mistake. A discount offered to someone with an expired card does nothing for them. A payment reminder sent to someone who stopped logging in months ago is also wasted effort.
The tools at a glance
The table below compares the tools covered in this guide by the job they do best, their entry price, and the main limitation to plan around. Pricing is for the entry plan unless stated otherwise.
| Tool | Best for | Entry price (plan, billing) | Key limitation |
|---|---|---|---|
| Stripe | Taking recurring card payments and producing the billing events other tools act on | Per-transaction pricing; no monthly platform fee on standard payments pricing | Does not write or send your customer emails |
| Sequenzy | Email sequences triggered by payment failures, billing events and usage events | Free: 2,500 emails/month. Paid: 15,000 emails/month at $19/month, or $205 a year | Priced by emails sent, so high-volume marketing lists cost more; it is not a churn-scoring suite |
| Amplitude | Spotting users whose behavior is drifting before they cancel | Free: 2M events/month, no time limit | Analytics, not messaging; above the free event volume you need a paid plan (pricing on request) |
| Mixpanel | Event-based retention and cohort analysis | Free: up to 1M monthly events. Growth: $140/month, or $120/month billed annually | Analytics only; you still need an email or messaging tool to act on it |
| Convert | Testing cancellation pages, offers and downgrade prompts against a control | 15-day free trial on Growth and Pro. Growth: $399/month, or $299/month billed annually | Priced by monthly tested users; not a billing or email tool |
| Make | Connecting billing, CRM, Slack and email tools without writing code | Free: 1,000 credits/month. Core: $10.59/month for 10k credits, or $9/month billed annually | Free plan limits: 2 active scenarios, runs at most every 15 minutes, 7 days of logs |
| Gainsight | Account health scoring and playbooks for a dedicated customer success team | Pricing on request | Assumes a customer success function to run it; no public free plan |
| ChurnZero | Customer success platform for tracking account health and renewals | Pricing on request | Assumes a team that will own health scores and playbooks; no public free plan |
| Vitally | Customer success workflows, offered in tech-touch, hybrid-touch and high-touch models | Pricing on request; trial sandbox with your own data after a demo request | No public self-serve plan; the model you pick depends on how your customers are served |
Recover failed payments first
Failed payments are the cheapest churn to fix because the customer usually still wants the product. The sequence has two parts that run on different systems. Your billing provider decides when to retry a charge. Your email tool decides what the customer reads, and when.
A workable dunning sequence has four stages:
- Before expiry: a message ahead of a card's expiry date asking the customer to confirm or replace it. This stops many failures before they happen.
- At the first failure: a short, plain email saying the payment did not go through, with a link that opens the hosted invoice page. Keep the link to one click.
- During retries: one or two follow-ups that match the retry schedule of your billing provider, so customers hear from you when a retry is about to happen or has just failed.
- At the end: a final message stating what happens to the account and how to reactivate it. Stop sending once the account is closed or the card is updated.
Sequenzy's payment-event trigger fits this structure. Because saas.payment_failed carries the hosted invoice URL, each stage can send the same one-click link without your team building it by hand. The hosted invoice page is where the customer pays or updates the card, so the email has a clear action to take. Sequenzy also connects to Paddle, Lemon Squeezy, Polar, Creem and Dodo Payments if your billing runs on one of those.
Measure this group on its own. Count failed charges per month, the share recovered within your retry window, and the share of expired cards you caught before expiry. No tool can promise a recovery rate for your business, because it depends on your customer mix, card types and retry settings. Set a baseline for one month before you change anything, then compare.
Make cancellation a conversation
Some customers who try to cancel have a fixable reason. The cancellation flow is where you find out what it is. Three options usually work better than a bare "Are you sure?" screen:
- Pause: let the customer hold the account for a set period. Useful for seasonal businesses and for teams between projects.
- Downgrade: offer a smaller plan that matches the usage they actually have. This keeps the customer and avoids a full loss.
- Reason capture: ask one multiple-choice question on the way out, with a free-text field for "other". Keep it to a single screen.
Downgrades are worth offering when customers are price sensitive or not using premium features. The arithmetic supports keeping a smaller paying customer: a $50 a month customer who stays for two years pays $1,200, while a $100 a month customer who cancels after three months pays $300. Retention wins whenever the retained revenue over the expected life of the account is higher, even at a lower price.
Convert is useful here because the cancellation page is a web experience, and you can test its variants against each other. Keep the test honest: change one thing at a time, such as the pause option or the order of the reasons, and keep a control version running. Convert's Growth plan is $399 a month, or $299 a month billed annually, and covers 100,000 monthly tested users. Its free trial lasts 15 days on Growth and Pro, which is enough time to set up one cancellation experiment.
Watch usage, not just logins
Usage-driven churn starts well before cancellation. A customer who logs in every week but never uses the reporting feature that justified the purchase is at risk, even though their login count looks healthy. Product analytics tools let you define what "active" means for your product and then alert you when accounts fall below it.
Start with one activation event: the single action that most reliably separates customers who stay from those who leave. It might be connecting an integration, sending the first campaign or inviting a second teammate. Compare customers who did that action in their first week with those who did not, and see whether the gap holds over months. Only then set a warning threshold, such as no activation event within 14 days, or a 40% drop in weekly sessions.
Amplitude's free plan covers 2M events a month with no time limit, and it includes 10K monthly session replays and guides. Its Plus plan adds behavioral cohorts and alerts, and it scales to 70M events. Beyond the free volume, Plus pricing is on request, so you need a quote before planning a budget above 2M events. Mixpanel's free plan covers up to 1M monthly events. Its Growth plan is $140 a month, or $120 a month billed annually, and covers up to 20M monthly events.
Either tool will tell you who is drifting. Neither will email them. The usual setup connects the analytics alert to an email sequence through a tool such as Sequenzy, which can trigger on events sent through its API, or through Make, which can read an analytics alert and call the email tool.
Re-engagement and win-back emails
Re-engagement emails go to active accounts whose usage has dropped. Their job is to point the customer back at the one thing they stopped doing. A useful re-engagement email names a feature the customer has not used, explains what it does in one sentence, and links to the page where they can try it. It does not send a list of every new release.
Win-back emails go to customers who have already cancelled. Many teams wait around 30 days before the first message, then test later windows. The message should match the reason the customer gave at cancellation. Someone who left over a missing integration should hear about that integration, not a general discount.
Always keep a holdout group. Pick a random share of cancelled customers, perhaps 10%, and send them nothing. Compare how many customers in the emailed group return with how many in the holdout return. Without a holdout, you cannot tell whether the email caused the return or whether those customers would have come back anyway.
Sequenzy's free plan covers 2,500 emails a month. For a company with 500 active customers receiving five re-engagement and lifecycle emails each month, that is exactly 2,500 emails, so the free plan is enough for a small test. Once volume grows, the 15,000-email plan is $19 a month, or $205 a year, and it includes unlimited contacts and sending domains.
Health scoring and customer success platforms
A health score combines several signals into one risk indicator: product usage, support activity, billing status and engagement. Each signal gets a weight that you choose and adjust over time. A failed payment might count for more than a drop in logins, and a champion leaving the customer's company might count for more than either. The score is only useful if someone acts on it, so define the response for each band before you build the score.
Customer success platforms handle this at scale. Gainsight, ChurnZero and Vitally all publish their prices on request rather than on a public plan. Vitally offers a trial sandbox with your own customer data after a demo request, and it sells three models: tech-touch for one-to-many and product-led customers, hybrid-touch for mixed models, and high-touch for one-to-one customer success. Choose the model that matches how your customers are actually served, not the one with the most features.
These platforms are a poor fit for a team of two or three people who can read every account by hand. They pay off when there are enough accounts that a person cannot track them individually, and when a customer success function exists to run the playbooks. Until then, a spreadsheet of accounts, a few usage alerts and a weekly review cover most of the same ground.
Connecting the pieces without engineering time
Churn tooling often fails at the handoffs. The billing system knows a payment failed, the analytics tool knows a customer went quiet, and the email tool knows nothing about either. Make is a visual automation builder with more than 3,000 apps, and it can route these events between systems. Its Free plan includes 1,000 credits a month, and each module action in a scenario uses one credit. The Core plan includes 10k credits for $10.59 a month, or $9 a month billed annually.
Free plan limits matter here. You get 2 active scenarios, a scenario can run at most every 15 minutes, a single run can take up to 5 minutes, and execution logs are kept for 7 days. A dunning workflow that runs four actions for each failed payment uses four credits per failure. At 500 failures a month that is 2,000 credits, which is more than the Free plan's 1,000, so you would move to Core.
Slack is often the team's alert channel. Its Free plan includes up to 10 apps and 90 days of message history. The Pro plan costs $8.75 per active member a month, or $7.25 billed annually, and includes unlimited history and app integrations. Ten active members on Pro come to $87.50 a month, or $72.50 a month billed annually. Send churn alerts to a single channel that the customer success or revenue owner watches, not to a general channel that everyone mutes.
What it costs at three list sizes
Costs depend on volume, not just on the number of people on your team. The table below uses three realistic scenarios and states each assumption, so you can adjust the numbers for your own business. Sequenzy is priced by emails sent, not by contacts, so the scenario assumes three sends per contact each month.
| Scenario | Volume assumed | Plan | Monthly price | Annual billing |
|---|---|---|---|---|
| Small SaaS: 5,000 contacts, 3 emails each per month | 15,000 emails/month | Sequenzy 15k | $19 | $205 a year |
| Growing SaaS: 10,000 contacts, 3 emails each per month | 30,000 emails/month | Sequenzy 30k | $29 | $313 a year |
| Mid-size SaaS: 20,000 contacts, 3 emails each per month | 60,000 emails/month | Sequenzy 60k | $49 | $529 a year |
| Early-stage product analytics | Up to 2M events/month | Amplitude Free | $0 | No time limit |
| Product analytics at 5M events a month | Up to 20M events/month on Growth | Mixpanel Growth | $140 | $120 a month billed annually |
| Cancellation-page tests for 100,000 monthly tested users | 100,000 monthly tested users | Convert Growth | $399 | $299 a month billed annually |
| Dunning workflow: 500 failures a month at 4 actions each | 2,000 credits/month | Make Core (10k credits) | $10.59 | $9 a month |
| Churn alerts for 10 active Slack members | 10 members | Slack Pro | $87.50 (10 × $8.75) | $72.50 a month (10 × $7.25) |
| Card payments through Stripe | Per transaction | Standard card processing | No monthly platform fee on standard payments pricing | Per-transaction rates on the pricing page |
| Health scoring for a customer success team | Account count and features | Gainsight, ChurnZero or Vitally | Pricing on request | Pricing on request |
Sequenzy's paid plans list annual totals alongside monthly prices. Its 120k-email plan is $99 a month or $1,069 a year, and its 2M and higher tiers are not part of this comparison. Mixpanel and Amplitude price on event volume, so count the events your product sends each month before comparing them. Convert prices on monthly tested users, which is a different unit from contacts.
How to choose
Start from the leak you can see in your own numbers, not from the tool list. The decision table below pairs each need with a starting point and the reason it fits.
| If you need | Start with | Because |
|---|---|---|
| To recover failed card payments by email | Your billing provider plus Sequenzy | Sequenzy's payment-failed trigger includes a hosted invoice link the customer can use to pay or update the card |
| To find users drifting before they cancel | Amplitude or Mixpanel | Both have free plans, so you can define activation and test thresholds before paying |
| To test cancellation pages and offers | Convert | It runs experiments on web pages and has a 15-day free trial on Growth and Pro |
| To connect billing, analytics and email without engineers | Make | It has a visual builder, a free plan for small tests, and paid plans priced by credits |
| Account health scores and playbooks for a CS team | Gainsight, ChurnZero or Vitally | These are built for teams that run customer success, and each requires a quote before you commit |
| A team alert when an account turns risky | Slack | A single channel gives the owner the alert in the tool they already use |
For a small team with a failed-payment problem, the minimum useful stack is a billing provider, an email tool that accepts payment events, and a spreadsheet to track results. Add product analytics once you have an activation event defined. Add a customer success platform only when a person owns each account.
Setup checklist
Set this up in order. Each step produces something you can check before moving to the next one.
| Step | What to set up | Tool | Done when |
|---|---|---|---|
| 1. Send payment events | Connect the billing account and confirm that failed-payment events reach the email tool | Stripe, Paddle or another supported provider; Sequenzy | A test failure appears in the email tool with its invoice link |
| 2. Build the dunning sequence | One message at failure, follow-ups matched to retry dates, and a final notice | Sequenzy | Every message links to the hosted invoice page and stops once the card is updated |
| 3. Define activation | Choose one event that separates retained customers from churned ones | Amplitude or Mixpanel | The event shows a clear gap between retained and churned customers over several months |
| 4. Set a usage warning | A threshold such as no activation within 14 days | Amplitude, Mixpanel or Make | The alert fires on a test account and reaches the email tool |
| 5. Add cancellation options | Pause, downgrade and one reason question on the cancel page | Your product, tested with Convert if needed | Each option is live and logged with the reason |
| 6. Launch win-back with a holdout | Random holdout of cancelled customers who receive no email | Sequenzy | Holdout is excluded from every send and the comparison is reported monthly |
| 7. Route team alerts | One channel for risk alerts with a named owner | Slack | Each alert has an owner and a response within the same week |
Mistakes that waste retention effort
- Measuring only aggregate churn. A single churn rate hides which cohort is leaking. Track retention by signup month and by plan, so a change in onboarding shows up in the right group.
- Sending the same email to every segment. A payment reminder, a feature tip and a win-back message need different triggers and different audiences. Keep them in separate sequences.
- Offering a discount to everyone who cancels. Discounts train customers to cancel for a better price. Offer downgrades and pauses first, and keep discounts for documented cases.
- Skipping the holdout. Without a control group, a win-back campaign can look successful when the same customers would have returned on their own.
- Buying an enterprise platform too early. Health-scoring platforms need owners. Without them, the scores are numbers nobody acts on.
- Forgetting the card update path. A dunning email that does not lead to a working card-update page loses the customer at the last step.
Where to start this quarter
Begin with failed payments. Measure last month's failed charges, set up one dunning sequence with a hosted invoice link, and compare recovery against the baseline after one billing cycle. Then define one activation event in your analytics tool and set a single usage warning. Add cancellation options and a win-back sequence with a holdout once the first two are running. Each step is small enough to finish in a week, and each one gives you a number to compare against the last.
Prices in this guide were checked against each vendor's public pricing page on 7 October 2026. Where a plan is listed as pricing on request, the vendor does not publish a figure, so request a quote before you plan a budget around it.
Frequently Asked Questions
What is a good churn rate for a SaaS business?
It depends on your customer segment, contract length and price point, so no single number applies to everyone. Judge your own trend first. If your monthly churn rate is falling across cohorts, your retention work is paying off, even if the absolute figure still looks high against a benchmark you read elsewhere. Compare newer cohorts with older ones at the same age, which shows whether onboarding and product changes are working.
How do I calculate churn rate correctly?
Divide the customers lost in a period by the customers you had at the start of that period. For logo churn, 50 lost out of 1,000 at the start of the month is 5%. For revenue churn, divide the recurring revenue lost from cancellations by the starting recurring revenue. Do not simply multiply a monthly rate by twelve to get an annual figure. Use compounding: a 5% monthly rate leaves about 54% of customers after a year, so roughly 46% are lost.
Can Stripe handle failed payments on its own?
Stripe processes the charges and reports the failures, but the messages a customer reads about a failed payment come from your own email tool. Check the current Stripe documentation for the retry and notification options available on your account. Then decide which emails you still need to send, such as a notice before a card expires, a message with a one-click link to update the card, and a final notice before the account closes. Sequenzy can send those messages from payment-failed events and link them to the hosted invoice page.
How do I know whether a win-back email is working?
Use a holdout. Choose a random share of cancelled customers, perhaps 10%, and send them nothing. Compare the share of emailed customers who return with the share of holdout customers who return over the same period. The difference is the part you can credit to the email. Report the result monthly so the holdout stays in place long enough to show a pattern.
When should I offer a downgrade instead of letting a customer cancel?
Offer a downgrade when the customer's cancellation reason is price, or when they are not using the features in their current plan. The arithmetic favors keeping the account: a $50 a month customer who stays for two years pays $1,200, while a $100 a month customer who cancels after three months pays $300. A smaller paying account is usually worth more over its life than a larger one that leaves early. Show the downgrade at the moment the customer starts to cancel, not in a separate email weeks later.
Do I need a customer success platform such as Gainsight, ChurnZero or Vitally?
Not at the start. These platforms assume that a team owns each account and runs playbooks against health scores. If you have a few hundred accounts and no one to act on the scores, a spreadsheet, usage alerts and a weekly review will cover most of the same ground. Each of these vendors prices on request, so ask for a quote once you have a team ready to use the platform.
Is the free plan of Sequenzy enough to start a dunning sequence?
The free plan covers 2,500 emails a month, which is enough for a small test. A company with 500 customers receiving five lifecycle emails each month sends exactly 2,500 emails. If you send more than that, the 15,000-email plan costs $19 a month, or $205 a year billed annually. Failed-payment messages are usually a small share of total sends, so the free plan can carry a dunning test while you measure recovery.