Before this series goes any further - before we talk about authorization rates, payment method breakdowns, dunning strategy, or any of the specific optimization levers that are coming - we need to establish this foundation. Because without it, every metric that follows is harder to interpret, and every benchmark comparison you make is potentially misleading.

Raw vs. Deduplicated - What’s the Difference?

A raw payment success rate counts every single transaction attempt. Every retry, every re-attempt, every duplicate request against the same underlying invoice. Each one gets tallied independently - a success or a failure in isolation.

A deduplicated payment success rate looks at the final outcome of a series of attempts tied to the same base transaction. Rather than counting each attempt individually, it groups them - typically by a shared identifier like an invoice ID, a payment intent, or an order reference number - and asks a single question: did this invoice ultimately get paid?

The definition itself hints at the mechanics. If you’re not already thinking about how your billing platform or processor groups transactions, now is the time to start.

Worth noting too: because deduplication is about final outcomes, it’s inherently tied to time. You can only accurately assess a deduplicated rate for a given period after your recovery window has fully closed. If your dunning process runs for 30 days, the deduplicated success rate for a given billing cycle isn’t a stable number until those 30 days are up. That window is baked into the calculation - which is exactly the point.

What This Looks Like in Practice

Let’s make it concrete.

Say you have 100 unique upcoming renewals in a given billing cycle. 70 succeed on the first attempt. The remaining 30 fail and enter the dunning process, where each gets up to 2 additional attempts - that’s 60 more attempts total.

Of those 60 dunning attempts, 15 succeed and 45 fail.

All in all, out of 160 total attempts, 85 succeeded and 75 failed.

Raw success rate: 53%. (85/160)

But here’s what actually happened: of those 100 unique invoices, 85 ultimately got paid - 70 on the first try, and 15 after multiple retries. 15 failed after every attempt was exhausted.

Deduplicated success rate: 85%. (85/100)

Same billing cycle. Same 100 customers. Two completely different pictures of your payment health.

That 53% raw rate isn’t wrong - it’s just answering a different question. The question your business actually cares about is simpler: out of 100 renewals, how many succeeded? That answer is 85. That’s your real number.

The 75 failed attempts buried in the raw figure are a mix of retries within invoices that eventually recovered, and retries within invoices that didn’t. Deduplication separates the noise from the signal.

What Goes Wrong When You Rely on Raw

This isn’t academic. Measuring raw rates instead of deduplicated ones leads to real, costly mistakes.

The more dangerous scenario: raw numbers can mask a real deterioration in payment health. If your retry volume is increasing - because more payments are failing on first attempt - your raw success rate might hold steady or even improve while your deduplicated rate quietly declines. The noise hides the signal until the problem is much harder to fix. By the time it’s visible in raw numbers, you’ve already lost significant revenue.

Another common scenario: you benchmark your 53% raw success rate against industry standards, conclude you have a serious payment problem, and start making changes - adjusting retry logic, overhauling dunning flows, escalating to your processor - when your actual deduplicated performance is 85% and perfectly healthy. You’ve spent time and resources solving a problem that didn’t exist.

In both cases, you’re flying blind. And in a subscription business, where payment performance compounds across every renewal cycle, flying blind is expensive.

Why This Matters for Everything That Follows

Payment success rate is the highest-level expression of payment health - but it’s made up of more granular metrics, each of which carries the same raw vs. deduplicated distinction. Authorization rates. Block rates. Merchant-initiated transactions (MITs) vs. customer-initiated transactions (CITs) performance. Even method-specific metrics like Apple Pay or card-not-present rates. The same principle applies to all of them.

As this series develops and we go deeper into each of those metrics, the raw vs. deduplicated lens will show up again. Setting this foundation now means we don’t have to relitigate it every time.

And beyond payments - this way of thinking touches retention and acquisition too. How you measure dunning recovery, renewal performance or conversation efficacy - the methodology carries over. We’ll get into that when those series come.

Questions Worth Asking Yourself

A few things to sit with after reading this:

• Are you currently tracking deduplicated payment success rates -or just raw?

• If you’re deduplicating, is it happening natively in your billing platform or processor, or is it a manual calculation your team is running?

• When you pull your payment success rate today, what number comes up - and do you know which version it is?

• If you were to run both numbers side by side for the last billing cycle, what gap would you expect to see?

Share your thoughts below - and if you’re working through this and want a second set of eyes, this is exactly the kind of foundational work I do at resubscribed.com

Adam Lifshitz

Founder & Principal Consultant, [re]subscribed

Want to work with me? Head over to resubscribed.com

I recently left to go out on my own, and when I was working through what I wanted to build, I kept coming back to the same observation - one I’d seen play out across multiple businesses, across different verticals, at different stages of growth.

There is a significant amount of revenue sitting inside many subscription businesses that isn’t being collected. Not because the product isn’t good enough. Not because the subscribers don’t want to stay. But because the operational infrastructure underneath the subscription - the payments stack, the retention mechanics, the acquisition flow - has gaps in it that nobody has gotten around to closing.

That’s what this newsletter is about.

What [re]subscribed covers

I’m going to write about three areas, and I’ll be direct about why I chose them.

Payments is the least glamorous and arguably the highest-ROI area in subscription businesses. Authorization rates, processing fees, fraud prevention, dispute management - these aren’t topics that get much airtime, but the delta between a well-optimized payments stack and a neglected one is real and quantifiable. I’ve seen businesses recover hundreds of thousands of dollars annually just by fixing things that were quietly broken.

Retention is where subscription businesses are really won or lost. Not just cancel save flows and churn metrics - though those matter - but the full picture: involuntary churn from failed payments, renewal economics, the transactional emails that are supposed to retain subscribers but sometimes do the opposite, the self-service gaps that make it easier to lapse than to stay.

Acquisition is where it begins - specifically the conversion side, not only the media buy. How a visitor becomes a paying subscriber. What’s happening on the purchase landing page, in the checkout flow, in the two minutes immediately after someone commits to a subscription. There’s often more room here than teams realize, and it doesn’t require spending more on acquisition to find it.

These three areas don’t operate in isolation. A better checkout flow improves the quality of the subscriber you’re acquiring. A better payments stack reduces the involuntary churn that undermines your retention numbers. The whole system is connected, and I’ll write about it that way.

Who this is for

If you’re running or working on a subscription business - media, e-commerce, SaaS, box subscriptions, it doesn’t matter - and you’re responsible for the numbers, this is for you. Whether that’s revenue, subscribers, churn, conversion, or all of the above.

I’m writing practitioner-to-practitioner. Not frameworks for their own sake. Not thought leadership at 30,000 feet. Specific diagnostics, specific benchmarks, specific things worth checking in your own business.

The goal isn’t to be comprehensive for the sake of it. It’s to be useful.

What’s coming

I’ll aim to publish weekly, alternating between payments, retention and acquisition, so there’s something relevant regardless of where your head is at.

If you’re running a subscription business and any of this resonates, subscribe. Forward it to someone who’d find it useful. And if something I write prompts a question or a reaction, reply - I read everything.

Let’s get into it.

Adam Lifshitz

Founder & Principal Consultant, [re]subscribed

Want to work with me? Head over to resubscribed.com