Should a SaaS offer a win-back discount to churned customers?

Should you offer a win-back discount in SaaS? Yes, but only to customers who left over price, only for a fixed number of months, and only when the extra returners it buys outweigh the discount you also hand to people who would have come back at full price. On 500 churned accounts at $150 MRR, a 30%-off-for-3-months coupon needs 3 extra returners to break even; a permanent 30% discount needs 13.

Rules of thumb (each is arithmetic from the worked table below or a cited source):

• Break-even extra returners = full-price returners x discount cost per full-price returner / (MRR x months an extra returner stays – discount cost per extra returner).
• A 30%-off-for-3-months coupon on $150 MRR costs $135 per returner; with 20 full-price returners on a 500-account list it pays for itself with 3 extra returners.
• Make the same 30% discount permanent and the break-even rises to 13 extra returners, because every full-price returner now costs $450 instead of $135.
• Proving a win-back lift from 4% to 7% takes about 900 churned accounts in each arm of a test (95% confidence, 80% power), so most SaaS lists can only hold out a control group and watch the direction.
• Expect discount-bought customers to be worth less: Paddle reported customers who came in through discounts had lifetime value “upwards of 32.41% less” than its core group (Paddle; acquisition data, not win-back).

Should you offer a win-back discount in SaaS? The break-even calculation

A SaaS win-back discount breaks even when the recovered MRR from extra returners, times the months they stay, exceeds the discount paid to every returner, including those who would have come back at full price. That last group is the cost most teams leave out, and it decides the answer.

No credible public benchmark for SaaS win-back rates with and without a discount was found, so every input below is a labelled assumption.

• 500 accounts churned voluntarily in the last 90 days, each at $150 MRR (assumption).
• Without a discount, 4% come back at full price: 20 accounts, each staying an assumed 10 months, worth $1,500 each.
• Extra returners that only the discount brings back stay an assumed 7 months, 30% fewer, in line with the direction of the Paddle finding above.
• Each design is tested at the same assumed lift, from 4% to 7% (15 extra accounts), so only the cost of the discount changes.

How each row is built, using the 30%-off-for-3-months coupon: discount cost per returner = $150 x 30% x 3 = $135. Twenty full-price returners now receive it too, so cannibalisation costs 20 x $135 = $2,700. An extra returner brings $150 x 7 months = $1,050, less the $135 coupon, so $915. Break-even = $2,700 / $915 = 2.95, rounded up to 3 accounts. At 15 extra accounts the campaign nets 15 x $915 – $2,700 = $11,025.

The forever row charges $45 a month for as long as each customer stays, so its net value per extra account falls to $735. Deeper discounts may lift the win-back rate more; the table holds the lift constant to show what each design must buy.

FAQ

How big should a SaaS win-back discount be?

Small and short is often enough. In the worked model, 50% off one month or 30% off three months breaks even with 2 to 3 extra returners, while 50% off for 12 months needs 29. Pick the smallest discount a price leaver would notice, then test it against a holdout.

Should the win-back discount match the save offer shown at cancellation?

No. If the cancellation coupon and the win-back coupon are the same, cancelling becomes the route to a lower price. Keep them different in size or form, for example a pause or smaller plan at cancellation and a time-limited coupon after churn.

Do discounted win-back customers churn again faster?

There is no public win-back study on this, but Paddle reported that customers acquired through discounts had lifetime value “upwards of 32.41% less” than its core group (Paddle, SaaS discounting strategy). That is acquisition data, so treat it as a warning rather than a win-back rate, and measure 90-day re-churn yourself.

Can Stripe limit a win-back coupon to a few months?

Yes. Stripe coupons can apply once, repeat for a set number of months, or apply forever, according to Stripe’s coupon documentation. Use “once” or “repeating” for win-back offers.

Is a win-back discount unfair to loyal customers?

It can be if the offer beats what loyal customers pay for long. A discount that ends after a few invoices and reverts to the standard price keeps returning customers on the same terms as everyone else once it expires.

How cannibalised full-price returners change the win-back maths

Cannibalised full-price returners are the churned customers who would have reactivated anyway, and a blanket win-back discount pays all of them for nothing. In the worked example they cost $2,700 on a 3-month coupon and $9,000 on a permanent one.

The ratio that matters is full-price returners per extra returner. A list where many customers come back on their own has a large cannibalisation bill; a list where almost nobody returns unprompted has a small one, and even a modest coupon can pay. Before you set any offer, count how many accounts reactivated last year with no offer at all: that is your full-price baseline.

Two places cut cannibalisation without dropping the discount. First, send the offer only to accounts coded as price leavers, not the whole churned list. Second, lead with a no-discount message and keep the coupon for the second touch, so customers who were coming back anyway return before they ever see it. The reason-matched offers for every other segment are set out on how to win back churned SaaS customers; this page covers only whether the discount itself pays.

Which churn reasons deserve a win-back discount?

Only the price churn reason deserves a win-back discount; customers who left over a missing feature, no usage or a company change need a different offer, because a lower price does not fix why they left.

Accounts with no recorded reason get a one-question email or short call before any offer. Customers who cut their plan rather than cancelled are a separate list with their own economics, covered in how to win back SaaS customers who downgraded. Failed payments are not win-back cases at all: they belong in the dunning queue described in how long a SaaS dunning period should be.

How to structure a win-back discount so it stays temporary

A win-back discount stays temporary when it is set up in billing as a coupon with a fixed number of invoices, never as a new list price or a coupon with no end date.

Stripe’s documentation describes three coupon durations: “once”, the default, applies only to the first invoice; “repeating” with a set number of months; and “forever”, which applies to all invoices indefinitely. Its own example is a 50% coupon with a 4-month duration, which on a monthly subscription “applies to the first 4 invoices” (Stripe, coupons). The first two rows of the worked table map to “once” and “repeating”; the third is the “forever” setting, and it is the one that more than quadruples the break-even.

Keep the cancellation save offer and the post-churn win-back offer apart. Stripe’s customer portal can show a retention coupon at the moment a customer cancels (Stripe, cancellation page). If that coupon and the later win-back coupon are the same, customers learn that cancelling is the way to get a lower price. Rework’s guide calls this a “churn-discount loop” and advises reserving discounts “for customers who left explicitly due to pricing concerns” (Rework).

State the full price after the discount ends in the offer itself, so the return to full price does not trigger a second cancellation.

How do you test whether a win-back discount is working?

A win-back discount is tested by holding out a random group of price leavers who get the same message with no coupon, then comparing reactivation and 90-day re-churn between the two groups.

The honest limit is sample size. Detecting a move from a 4% to a 7% win-back rate at 95% confidence and 80% power needs about 900 accounts in each group, by the standard two-proportion formula. A list of 500 churned accounts cannot prove that lift. What a smaller list can do is show the direction and catch the expensive failure: if the holdout group reactivates at nearly the same rate as the discounted group, you are paying the cannibalisation bill and little else.

Track reactivated accounts, reactivated MRR and the share still paying 90 days after the discount ends. Report reactivated MRR on its own line, separate from retention, as the churned-customer win-back guide explains.

What a SaaS win-back campaign costs to run yourself

A SaaS win-back campaign run in-house costs mostly time: at an assumed 15 minutes per account to code the reason, write and follow up, 500 churned accounts take about 125 hours.

The coupon is the cheap part; the expensive part is reaching price leavers on a channel they answer before the offer expires. Email alone may mainly reach the people who were coming back anyway. A call or text to the price-leaver segment, timed after the discount decision rather than on a fixed drip, gives the coupon a chance to reach people who would not come back on their own; the mechanics are covered in recovering cancelled subscriptions with outbound agents.

Do it yourself when the list is small, the price segment is under a few dozen accounts, and someone on the team already talks to customers. Hand the outreach over when the list runs to hundreds of accounts and nobody owns it. LeadsNow, with 50,769+ AI-booked sales appointments since 2017 and 1M+ leads generated, runs AI appointment setting that calls, texts and messages leads within permitted calling hours; it is billed as a revenue share, a fee per appointment, or a mix of both, and can be cancelled any time with 14 days notice. For the wider SaaS pipeline picture, see SaaS lead generation and trial-to-paid conversion with outbound conversations.

Sources

1. Paddle: SaaS discounting strategy (published 2022; lifetime value of discount-acquired customers)
2. Stripe: Coupons and promotion codes (coupon durations once, repeating, forever)
3. Stripe: Add a cancellation page (retention coupon in the customer portal)
4. Rework: Win-back campaigns (churn-discount loop)