Co-authored with Katelyn Belyus, Global Director, Strategy

Preventing churn has become a core focus for subscription marketers — for good reason. But what happens after a subscriber cancels still matters: former subscribers are far more likely to return than cold prospects, especially in the days and weeks immediately after they leave. The challenge is that the odds of winning them back decline rapidly over time, making the post-churn window a valuable — and often underutilized — opportunity for retention marketers. 

To understand what makes a successful winback campaign, Piano's Strategic Services team ran an A/B/C test through Piano Test Lab – an initiative that helps publishers test and benchmark strategies through controlled experiments. The question we wanted to answer: does a well-timed winback modal recover lost subscribers? And if so, does a discount vs no discount make the difference on reader engagement post conversion? 

Winback modals bring churned subscribers back (especially with a discount)

We tested three experiences with previously subscribed users who'd churned: 

• A — No winback message (control): no winback message 

• B — Full-price modal: standard pricing offer 

• C — Discount modal: 20% off return offer 

The modal fired on the user's second pageview. Here are the results: 

The conclusion – winback campaigns work and have higher conversion rates compared with the benchmark median conversion rate across all experiences, which was 0.19% during the test period. There are a few other important takeaways: 

1. The discount drove a clear conversion lift 

A 20% discount noticeably changed user behavior. 

• 54% higher click-through rate 

• 88% higher conversion rate 

• 78% higher total revenue compared to full-price branch 

The takeaway: lead with a discount to drive conversions. Even though each subscriber paid less, the volume of conversions was high enough to drive more total revenue than the full-price offer. 

2. Full-price returners showed deeper engagement after conversion 

Converted users in both branches read more pages after resubscribing than they did before. Users who took the discount read 77% more pages after converting, and those who paid full price read 126% more pages – that's 63% higher for those who paid full price than those who returned at a discount. It stands to reason: readers who pay full value tend to use more of what they bought. But even any offer moves the needle on content consumption post-conversion. 

3. Returns increased regardless of conversions  

The modal made an impact well beyond the click. Comparing sessions to the control group before and after the modal fired: 

• No-discount branch: sessions up 34% 

• Discounted branch: sessions up 27% 

Exposure alone – regardless of whether the user converted – raised return-visit behavior. The modal works as a reminder that your publication is still there, still relevant, and still worth a look. 

What to test next 

If you're running your own winback experiments, here's where to focus: 

• Find the discount sweet spot. 20% worked, but it may not be optimal. Test 10%, 15%, 25%, and 30% to see where the curve bends. Be mindful not to discount more than your pricing in new business efforts. 

• Measure LTV, not just conversion. The 78% revenue lift on the discount group is short-term. Track how many subscribers are still active at 30, 60, and 90 days — and ideally 6 months — for both branches. Even though discounted returners didn't cancel auto-renewal at a notably higher rate in this test – that 78% revenue lift could reverse if they churn more over the following months. That makes long-term retention tracking the most critical follow-up. 

• Optimize trigger timing. Try scroll depth, different pageview counts, return visits after a delay, or even a paywall hit. Different triggers surface different intent levels, and you may find a higher-converting moment than the second pageview. 

• Cap frequency and test cadence. If a user dismisses the modal, repeating it risks annoyance and lower future engagement. Cap exposure (for example, two impressions in 30 days) and test cadence variations. 

• Iterate on creative and copy. This test compared presence vs. absence and discount vs. no discount — not creative variations. Try emotional framing (“come back to your unfinished story”) against transactional (“20% off”), and experiment with testimonials, urgency, and CTA wording. Small wins in creative will compound at scale. 

The bottom line 

Winback campaigns work to drive revenue, and they work hardest when paired with a discount. The results validate the broader pattern we keep seeing in Test Lab experiments: well-timed, well-targeted interventions can re-engage readers who would otherwise quietly slip away. 

This experiment was run through Piano Test Lab – an ongoing initiative from Piano's Strategic Services team. Piano Test Lab gives you a structured way to run experiments using guidance from Piano's product and strategy experts and benchmark yourself against thousands of industry peers. 

Ready to test this for your audience? Reach out to your Piano account representative or click below.