Resources

Practical guides on keeping the revenue you've already earned. Written from research and real industry data — not marketing fluff.

Why 9% of your MRR disappears to failed payments

The silent leak in every subscription business: what causes it, what it costs, and what actually fixes it.

Read the guide →

Dunning emails that actually recover revenue

The 3-email sequence that wins back expired cards — with the psychology behind each message and copy you can steal.

Read the guide →

How to reduce failed payments in Stripe

The full playbook in order: Smart Retries, card updater, dunning emails, in-app prompts — and when you need a dedicated tool.

Read the guide →

7 dunning email examples that actually recover revenue

Steal these templates: the day 1/4/7 core sequence plus pre-expiry, handoff, founder-note, and win-back variants.

Read the guide →

Involuntary churn: the silent revenue killer

What it is, how to measure it, industry benchmarks, and the recovery playbook — the cheapest growth you'll ever buy.

Read the guide →

A cheaper alternative to Churn Buster for indie SaaS

An honest comparison: $29/mo vs $269/mo. What you give up, what you keep, and which one fits your stage.

Read the guide →

Why 9% of your MRR disappears to failed payments

8 min read · Based on Baremetrics 2026 benchmarks and Recurly industry data

Here's a number that should bother every subscription founder: according to Baremetrics' 2026 benchmarks across 119 subscription companies, roughly 9% of MRR is lost to failed payments. Not churn. Not competitors. Just cards that didn't go through.

Recurly puts the industry-wide damage at $129 billion lost in 2025 alone. And the cruelest part? Most of these customers didn't want to leave. They liked your product. Their card just expired, or their bank flagged the charge, and nobody told them in a way that got them to act.

Why payments fail

Failed payments fall into two buckets. Soft declines are temporary: insufficient funds, a bank's fraud filter being cautious, a card reported lost and reissued. These often succeed on retry a few days later. Hard declines are permanent for that card: it's expired, cancelled, or invalid. No amount of retrying will fix those — only the customer updating their card will.

The split matters because each needs a different response. Soft declines want smart retries. Hard declines want the customer, reached quickly, with a frictionless way to update their details.

Why founders ignore it

Three reasons. First, it's invisible — Stripe shows failed payments in a dashboard tab most founders rarely open, so the leak never becomes a line item anyone owns. Second, it feels small per incident: one $29 failure doesn't trigger alarm bells, even though hundreds of them compound into real money. Third, fixing it sounds boring next to building features. It's plumbing, not product.

But do the math for a $20k MRR business: 9% failing means $1,800/month evaporating. Over a year, that's $21,600 — more than a full-time hire in many markets — walking out the door silently.

What actually fixes it

There are four layers, in order of effort:

  1. Smart retries. Stripe's built-in Smart Retries already attempts failed charges again at optimized times. Turn it on if you haven't — it's free and catches the easy soft declines.
  2. Card account updater. Stripe can automatically pull updated card details from card networks when cards expire or are reissued. This alone recovers a meaningful chunk with zero customer contact.
  3. Dunning emails. A short sequence — friendly reminder, firmer nudge, final notice — directing customers to a card-update page. This is where most of the recoverable revenue lives, because hard declines need a human to act.
  4. In-app prompts. If you have a logged-in product, banner the account owner when their payment fails. Email open rates hover around 20-30%; in-app visibility is near 100% for active users.

The honest bottom line

You will never recover 100% of failed payments. Some customers were already on their way out; the failed card just did the cancelling for them. A realistic recovery rate for a good dunning setup is 30-40% of failed revenue — which, on that $20k MRR example, is $6,500-$8,600 back per year for an afternoon of setup.

That's the whole pitch for caring about this unglamorous problem: it's the highest-ROI afternoon you'll spend this quarter.

How much are you losing?

Plug in your MRR and see your estimated failed-payment loss — and what 40% recovery would win back.

Try the ROI calculator

Dunning emails that actually recover revenue

6 min read · Copy you can steal, and the psychology behind each message

"Dunning" is the industry's ugly word for a simple thing: emailing customers whose payment failed and asking them to update their card. Done badly, it's nagging. Done well, it reads like good customer service — and it recovers 30-40% of failed revenue. Here's the 3-email sequence that works, and why each message is shaped the way it is.

The golden rules first

  • Assume goodwill. The customer didn't stiff you; their card expired. Every email should read like you're doing them a favor, because you are — you're saving their account.
  • One link, above the fold. Every email has exactly one job: get the click to the card-update page. Don't bury it under feature announcements.
  • Short beats thorough. Nobody reads a 400-word billing email. Say what happened, say how to fix it, get out.
  • Space them out. Day 1, day 4, day 7 is the standard cadence. Daily emails train people to ignore you; monthly emails let accounts die quietly.

Email 1 (Day 1): The friendly heads-up

Tone: casual, helpful, zero pressure. Most customers who will ever update their card do it after this email — it catches everyone who just needs a nudge.

Subject: Quick heads-up: your payment didn't go through

Hi {{first_name}}, just a quick heads-up — we tried to charge {{amount}} and the payment didn't go through. Usually just an expired card, nothing to worry about. Update it here (30 seconds): {{update_card_url}}

Why it works: it normalizes the situation ("usually just an expired card"), minimizes effort ("30 seconds"), and asks for exactly one thing.

Email 2 (Day 4): The firmer nudge

Tone: direct, slightly urgent, still warm. This one converts the procrastinators — people who meant to do it and forgot.

Subject: Your subscription is at risk — update your card

Hi {{first_name}}, we still haven't been able to process your {{amount}} payment, and your subscription will be paused soon. Update your card here: {{update_card_url}}. Reply to this email if you're having trouble — a human reads every reply.

Why it works: it names a consequence (paused subscription) without threats, and the "human reads every reply" line catches edge cases — wrong email on file, company card issues, people who actually want to cancel.

Email 3 (Day 7): The final notice

Tone: plain, factual, respectful. This is the last touch before cancellation. Keep it short and give the graceful exit.

Subject: Final notice: keep your account active

Hi {{first_name}}, final notice — we've tried to collect {{amount}} three times. Your subscription cancels tomorrow unless your card is updated: {{update_card_url}}. If you meant to cancel, no hard feelings.

Why it works: the deadline is real and specific ("tomorrow"), and offering the graceful exit paradoxically saves accounts — some people update the card precisely because you didn't guilt them.

What the data says

Across dunning tools, email #1 typically drives 50-60% of all recoveries, #2 drives 25-30%, and #3 picks up the remaining 10-15%. That distribution is why you don't skip straight to the final notice — you'd leave most of the money on the table and annoy everyone in the process.

Personalize the variables that matter — first name, amount, a direct card-update link — and keep the "from" name a human or your company, never "noreply." Replies to dunning emails are disproportionately from customers with real billing problems your automation can't solve.

Don't want to hand-roll this?

Recoverly runs this exact 3-email sequence automatically, with your branding and your voice. Set it up in an afternoon.

Calculate what you'd recover