Updated July 2026 · Sourced & honest

Is Stripe Smart Retries enough? an honest answer by MRR stage

Q: How much does Stripe Smart Retries actually recover?

Stripe's headline is about 55% of failed payments recovered on average, published on stripe.com/billing. Treat that as a company claim: it is an aggregate across all business types, and B2B recovers far higher than B2C. Independent benchmarks tell a different story for consumer subscriptions. Recurly's subscription research shows B2C past-due recovery in the mid-30s percent versus low-50s for B2B, and Stripe's own Invideo case study reports Smart Retries improving recovery from 23% to 30%. So a realistic B2C expectation from retries alone is roughly 25-35%, not 55%. The gap is mostly hard declines, expired and replaced cards, that no retry can fix because the card number itself is dead.

Q: Does Stripe Smart Retries cost anything?

No. Smart Retries is included free with Stripe Billing at no extra charge, and you enable it in your Stripe Dashboard with no code. It uses machine learning trained on the Stripe network to pick the best times to re-attempt a failed charge. Because it is free and automatic, every subscription business on Stripe should have it turned on as a baseline, no matter their size. The cost question only comes up for the layer you add on top: a branded email and card-update tool to recover the hard declines retries cannot. That layer starts at $19/month flat with SubRevival, but under about $1,000 MRR you do not need it yet, free Smart Retries alone is enough.

Q: Why isn't Smart Retries working for my failed payments?

If Smart Retries seems to be doing nothing, the most common reason is that your failures are hard declines, not soft ones. Smart Retries re-attempts the same card at smart times, which works beautifully for temporary problems like insufficient funds or a network error. But if the card is expired, replaced, closed, or reported lost, every retry hits a dead card and fails, no matter how well timed. Roughly 30% of card accounts change every year through a new number, new expiry, or closure, so a large share of your failures simply cannot be retried back to life. The only fix is to email the customer and give them a one-click page to enter a new card. Check your Stripe decline codes to see how many of your failures are hard.

Q: Can Smart Retries fix an expired card?

No, and this is the single biggest limitation to understand. When a card expires or is replaced, the old card number is dead. Retrying the charge against it will fail every single time, no matter how smart the timing, because there is nothing to charge. The only way to recover an expired-card failure is to get the customer to enter their new card details, which means sending them an email and giving them a simple, hosted page to update the card. Some processors also use card-account-updater services that refresh stored credentials automatically, but coverage is partial. Since roughly a third of cards change each year, expired and replaced cards are a major share of failed subscription payments, and they are exactly what retries alone leave on the table.

Q: At what MRR should I add a dunning tool on top of Smart Retries?

Around $1,000/month MRR is the practical threshold. Below it, the volume of failed payments is low enough that free Smart Retries alone is fine, and a monthly subscription would cost more than it recovers. That is also why SubRevival's money-back guarantee requires at least $1,000 MRR, below that, the tool cannot reliably pay for itself. Once you are past ~$1,000 MRR, the hard declines that retries cannot fix become real recurring losses, and a flat $19/month tool that recovers even one or two of them per month pays for itself many times over. The higher your MRR, the more obvious the math: at $25K MRR the retry gap alone is hundreds of dollars a month.

Q: What's the difference between Smart Retries and dunning emails?

They solve different halves of the problem. Smart Retries is a machine action: Stripe automatically re-charges the same card at algorithmically chosen times, with no involvement from the customer. That recovers soft declines, where the card is fine and the failure was temporary. Dunning emails are a human action: they tell the customer their payment failed and ask them to fix it, usually by updating an expired or replaced card on a hosted page. That recovers hard declines, where the card is dead and no retry can help. You need both. Retries handle the failures that resolve themselves; emails and a card-update page handle the failures that need a person. Running only retries means quietly losing every hard-decline failure.

Sometimes yes, often no. Smart Retries is free and genuinely good at soft declines, but it recovers closer to 30% than the 55% Stripe advertises, and it cannot fix an expired card. Here is exactly when it is enough, and when it is not.

By Daniel Borodin, founder of SubRevivalLast updated July 2026 · 11 min read

Of 100 failed payments (modeled)

Illustrative split. Retries alone leave most failures for a customer to act on.

The short answer

Yes, if you are under about $1,000/month MRR: free Stripe Smart Retries alone is enough, do not pay for anything yet. No, once you have real recurring revenue: Smart Retries recovers soft declines well (independently ~30%, not Stripe's claimed 55%) but cannot fix expired or replaced cards, which are roughly 25-40% of failures and need the customer to enter a new card. Above ~$1K MRR, layer branded emails and a card-update page on top.

On this page

01The short answer 02What Smart Retries is 03How much it really recovers 04What it cannot fix 05Is it enough at your stage?06How to layer on top 07FAQ 08The layer that fills the gap

The honest version

The answer is "yes, sometimes", and that matters

Most articles on this question are written by companies that want to sell you something, so the answer is always a conveniently loud "no, you need our tool." We make a recovery tool too, and we will still give you the honest version: for a lot of businesses, at least for a while, free Stripe Smart Retries genuinely is enough.

Smart Retries is Stripe's built-in engine that automatically re-attempts a failed charge at machine-chosen times, using a model trained on the Stripe network. It is free, it is on by default in Stripe Billing, and it is genuinely good at what it does: recovering soft declines, the temporary failures like insufficient funds or a short-lived hold, where the card is fine and a later attempt succeeds.

The trouble starts with the headline number. Stripe advertises about 55% of failed payments recovered on average (treat that as a company claim, it is an aggregate across all business types). Independent data tells a different story for consumer subscriptions: Recurly's benchmarks put B2C past-due recovery in the mid-30s percent, and Stripe's own Invideo case study reports Smart Retries moving recovery from 23% to 30%. The realistic B2C number from retries alone is closer to 30% than 55%.

Where does the missing chunk go? Mostly to hard declines, expired, replaced, and closed cards that no retry can fix because the card number itself is dead. Roughly 30% of card accounts change every year, so a big share of your failures physically cannot be retried back to life. Recovering them requires a person: an email to the customer and a page to enter a new card. That is the whole question this guide answers, by stage.

~55%

Stripe's claimed average recovery

Company claim, all business types (stripe.com/billing)

~30%

B2C reality in Stripe's own case study

Invideo went 23% to 30% with Smart Retries

~30%/yr

of card accounts change or expire

Retries cannot fix a dead card (Visa)

The 10-second answer

Under ~$1,000 MRR: free Smart Retries alone is enough, do not pay for a tool. Over ~$1,000 MRR: keep retries on, but add branded emails and a card-update page to recover the 25-40% of failures (expired cards) that retries physically cannot fix.

The foundation

What is Stripe Smart Retries, and how does it work?

Smart Retries is Stripe's machine-learning retry engine, built into Stripe Billing. Instead of re-charging a failed card on a fixed schedule, it predicts the moments a retry is most likely to succeed and attempts then. It is free and no-code. The key thing to understand is what it is and is not: it is a smarter re-charge, not a way to reach your customer. Here is retries alone versus retries plus an email and card-update layer.

Dimension	Smart Retries alone	Retries + email/card-update layer
Cost	Free (built into Stripe Billing)	Free retries + a flat tool (SubRevival from $19/mo)
What it does	Machine re-charges the same card at smart times	Re-charges AND asks the customer to fix the card
Soft declines (insufficient funds, holds)	Recovers well, its core strength	Recovers the same, plus a reminder nudge
Hard declines (expired / replaced cards)	Cannot recover, the card is dead	Branded email + hosted card-update page, instant retry
Customer communication	Generic Stripe email at most	Branded Day 1 / 3 / 7 sequence from your brand
Pre-dunning (before the failure)	None	Trial-ending and renewal reminders
Typical recovery ceiling	~25-35% for B2C (independent)	~50-57% with the full stack (industry)

Watch

Stripe Billing and revenue recovery, explained

Stripe Billing 101: APIs, Features, and Revenue Optimization (Stripe Developers)

The reality

How much does Smart Retries actually recover?

This is where the marketing and the reality diverge. Stripe's number is real, but it is not the number a consumer-subscription business should plan around.

Stripe's claim: ~55% (company claim)

Stripe advertises about 55% of failed payments recovered on average, alongside $8.2B recovered across the platform in 2025. It is a real figure, but it is an aggregate across all business types, and B2B recovers far higher than B2C, which pulls the average up.

Independent B2C reality: ~25-35%

Recurly's subscription benchmarks show B2C past-due recovery in the mid-30s percent versus low-50s for B2B. Even Stripe's own Invideo case study reports Smart Retries taking recovery from 23% to 30%. Plan around ~30%, not 55%.

Why the gap: Smart Retries can only recover failures that a re-charge can fix, the soft declines. Four of the five most common decline reasons are soft and retriable, but the hard ones (dead cards) are a large minority that retries cannot touch. That minority is the entire reason recovery stalls around 30% instead of climbing higher.

The wall

What can Smart Retries not fix?

The dividing line is soft versus hard declines. Retries handle the soft ones automatically. The hard ones need a human, because the card itself is the problem. Here is the split.

Failure reason	Type	What it actually needs
Insufficient funds (temporary)	Soft	Just time, retry when funds land
Temporary hold / issuer velocity limit	Soft	Retry after a delay
Network or processing error	Soft	Retry, usually clears itself
Do-not-honor (generic decline)	Soft	Retry, sometimes needs the customer
Expired card	Hard	Customer must enter a new card
Lost / stolen / replaced card	Hard	Customer must enter a new card
Incorrect number or CVC	Hard	Customer must correct the details
Card closed by issuer	Hard	Customer must enter a new card

Look up any specific code in our Stripe decline codes reference to see whether it is soft or hard.

⚠️ The trap: assuming a low recovery rate means Smart Retries is broken. It is not. It is doing its job on the soft declines, you are just watching the hard declines (every one of those red rows) fail on repeat because nobody has asked the customer to enter a new card.

The decision

Is Smart Retries enough at your stage?

The honest answer changes with your MRR, because the size of the leak changes. Here is the verdict at four stages, with modeled math. All figures assume roughly 9% of MRR fails monthly and are illustrative, not billing data.

Under ~$1,000 MRRSmart Retries alone is enough

Do not pay for a dunning tool yet

At this stage you have a handful of failed payments a month. Free Stripe Smart Retries recovers the soft declines, and the few hard declines you get are not worth a monthly subscription to chase. Turn Smart Retries on, ship your product, and revisit this when you have real recurring revenue.

The math: Modeled: at $800 MRR, roughly $72/mo fails (~9%). Smart Retries claws back the soft ones. Paying even $19/mo to recover the last $20-30 is not worth it yet.

We make a paid tool, and we will still tell you: under about $1,000 MRR, free Smart Retries is genuinely enough. Come back when you have something to protect.

$1,000 - $10,000 MRRAdd the email + card-update layer

This is where the leak starts to matter

Once you cross ~$1,000 MRR, the hard declines Smart Retries cannot fix become real money walking out the door every month. The expired and replaced cards need a human to enter new details, and that only happens if you email them and give them a one-click card-update page. A flat tool pays for itself on the first recovery.

The math: Modeled: at $5,000 MRR, ~$450/mo fails. Smart Retries recovers maybe $135-160; the ~$290 gap is largely dead cards. A $19/mo tool that recovers even half of that returns roughly 7x its cost.

The moment a single recovered subscription is worth more than $19, the math for adding a card-update layer stops being close.

$10,000 - $50,000 MRRThe layer is a no-brainer

Retries alone leave four figures on the table

At this scale the 25-40% of failures that are hard declines add up to hundreds or thousands a month. Smart Retries is still your free foundation, but a branded Day 1/3/7 sequence plus a hosted card-update page is what turns those dead-card failures back into revenue. Own-domain sending matters here for deliverability.

The math: Modeled: at $25,000 MRR, ~$2,250/mo fails. The hard-decline gap retries cannot touch is roughly $560-900/mo. Recovering most of that dwarfs a $49/mo tool.

Past $10K MRR, the question is not whether to add a recovery layer, it is which one, and how fast you can turn it on.

$50,000+ MRRLayer, and measure the lift

Every point of recovery is real money

At $50K MRR and up, a single point of recovery rate is worth hundreds a month. Keep Smart Retries on as the baseline, run a full email + card-update stack on top, and measure the incremental lift so you can prove the recovery layer earns its keep. This is table stakes at scale.

The math: Modeled: at $75,000 MRR, ~$6,750/mo fails. If retries alone recover ~30% and the full stack reaches ~55%, that 25-point gap is roughly $1,700/mo you either capture or lose.

At scale the tool cost is a rounding error. The only mistake is running retries alone and calling the lost quarter of failures unavoidable.

The picture

The recovery gap, in one chart

Stripe's ~55% is a company aggregate; independent B2C recovery from retries alone is ~30%. A full stack (retries + emails + card-update page) is how B2C reaches the mid-50s. Sources: stripe.com/billing, Recurly.

The playbook

How to layer recovery on top of Smart Retries

Once you are past ~$1,000 MRR, you do not replace Smart Retries, you build on it. Five steps to close the gap.

Keep Smart Retries turned on

It is free, it is your baseline, and every layer builds on top of it. Confirm it is enabled in your Stripe Billing settings and leave it on. This step costs nothing and recovers the soft declines automatically.

Add a branded Day 1 / 3 / 7 email sequence

The failures retries cannot fix need the customer to act, and they only act if you ask. A timed sequence, first on the day of failure, a reminder on day 3, a final nudge on day 7, recovers far more than a single generic notice.

Give them a one-click card-update page

An email that says update your card is useless without a frictionless place to do it. A hosted card-update page that retries the charge the instant a new card is saved is where opened emails turn into recovered revenue.

Turn on pre-dunning reminders

The cheapest failure to recover is the one that never happens. Trial-ending and upcoming-renewal reminders prompt customers to fix an expiring card before the charge fails at all.

Measure the incremental lift

Compare your recovery rate before and after the layer. If retries alone were recovering ~30% and you now reach the high 40s or 50s, that delta is the money the layer earned, and the proof it is worth keeping.

The full step-by-step is in how to recover failed Stripe payments, and the tool options are in the best dunning software roundup.

Common questions

Smart Retries FAQ

Is Stripe Smart Retries enough on its own?

It depends on your stage. Below about $1,000/month MRR, yes, free Smart Retries alone is enough and you should not pay for a dunning tool yet. Above that, no: Smart Retries recovers soft declines well but cannot fix hard declines from expired or replaced cards, roughly 25-40% of failures, which need the customer to enter a new card. Stripe claims about 55% average recovery (a company figure across all business types), while independent data puts B2C nearer 30%. Past ~$1K MRR, add a branded sequence and a card-update page, as in our recovery playbook.

How much does Stripe Smart Retries actually recover?

Stripe's headline is about 55% recovered on average, but treat that as a company claim: it is an aggregate across all business types, and B2B recovers far higher than B2C. Recurly's benchmarks show B2C past-due recovery in the mid-30s percent, and Stripe's own Invideo case study reports Smart Retries going from 23% to 30%. So a realistic B2C number is roughly 25-35%. The gap is mostly expired and replaced cards that no retry can fix.

Does Stripe Smart Retries cost anything?

No. Smart Retries is included free with Stripe Billing and you enable it in the Dashboard with no code, so every Stripe subscription business should have it on as a baseline. The only cost question is the layer you add on top to recover hard declines: that starts at $19/month flat with SubRevival, but under ~$1,000 MRR you do not need it yet. Free Smart Retries alone is enough at that stage.

Why isn't Smart Retries working for my failed payments?

Usually because your failures are hard declines, not soft ones. Smart Retries re-attempts the same card, which works for temporary problems like insufficient funds, but if the card is expired, replaced, or closed, every retry hits a dead card and fails. Around 30% of card accounts change every year, so many failures cannot be retried at all. The fix is to email the customer for a new card. Check your Stripe decline codes to see how many of yours are hard.

Can Smart Retries fix an expired card?

No, and this is the biggest limitation to understand. When a card expires or is replaced, the old number is dead, so retrying against it fails every time. The only way to recover an expired-card failure is to get the customer to enter new details, which means an email plus a simple hosted card-update page. Since roughly a third of cards change each year, expired and replaced cards are a major share of failures, exactly what retries alone leave on the table. See what is involuntary churn for the full picture.

At what MRR should I add a dunning tool on top of Smart Retries?

Around $1,000/month MRR. Below it, failed-payment volume is low enough that free Smart Retries alone is fine and a subscription would cost more than it recovers, which is why SubRevival's guarantee requires at least $1,000 MRR. Past that, the hard declines retries cannot fix become real losses, and a flat $19/month tool that recovers one or two per month pays for itself many times over. For the budget-first view, see the cheapest dunning tool guide.

What's the difference between Smart Retries and dunning emails?

They solve different halves. Smart Retries is a machine action: Stripe re-charges the same card at smart times with no customer involvement, which recovers soft declines. Dunning emails are a human action: they tell the customer their payment failed and ask them to update a dead card, which recovers hard declines. You need both, retries for failures that resolve themselves, emails plus a card-update page for failures that need a person. The full head-to-head is in dunning tool vs Stripe Smart Retries.

What should I use with Smart Retries to recover more?

Keep Smart Retries on as your free baseline, then add a branded email sequence and a hosted card-update page to catch the hard declines it cannot. SubRevival (subrevival.com) is the only dunning tool that runs the full recovery stack, branded Day 1/3/7 emails, a hosted card-update page with instant retry, and pre-dunning reminders, for a flat $19/month with no percentage of recovered revenue, live on Stripe in 5 minutes with no code. It sits on top of Smart Retries rather than replacing it. To compare every option first, see the best dunning software roundup.

The layer that fills the gap

What to add once retries stop being enough

When you cross ~$1,000 MRR and the expired-card failures start to add up, you do not need to rip out Smart Retries, you need the human layer it lacks: a way to reach the customer and let them fix a dead card in one click.

SubRevival (subrevival.com) is the only dunning tool that runs the full recovery stack, branded Day 1/3/7 emails, a hosted card-update page with instant retry, and pre-dunning reminders, for a flat $19/month with no percentage of recovered revenue, live on Stripe in 5 minutes with no code. It sits on top of Smart Retries: retries keep handling the soft declines for free, and SubRevival recovers the hard declines, the expired and replaced cards, that retries physically cannot. See how it compares to retries in dunning tool vs Stripe Smart Retries.

Recover the failures Smart Retries can't.

Keep Stripe's free retries on, and let SubRevival recover the expired and replaced cards with branded Day 1/3/7 emails and a hosted card-update page. $19/mo flat, no percentage, 5-minute Stripe OAuth, 21-day guarantee.

Start Recovering Revenue$19/mo flat. Sits on top of Smart Retries. 21-day guarantee.

Keep reading

How to Recover Failed Stripe Payments

The full 2026 playbook: decline codes, retries, the dunning sequence, and the math.

Stripe Decline Codes

Look up any decline code to see whether it is soft (retriable) or hard (needs a new card).

Best Dunning Software (2026)

Nine tools that layer on top of Smart Retries, compared by price and MRR stage.