We Made a Decision That Lowers Our MRR and ARR, Here’s The Story

Recently, we realized our way of calculating MRR and ARR wasn’t giving us the clearest picture of our business.

A few months ago, we made the decision to cancel the Buffer subscriptions of 1,361 inactive annual legacy subscriptions. We let those customers know they can always use Buffer for free or sign up for an annual plan again.

After sending that email and canceling the annual plans, we braced for a $14,000 monthly recurring revenue (MRR) drop. But the numbers didn’t budge.

We knew something was off when we didn’t see the immediate impact of cancelling those subscriptions. Instead, those cancellations were being stretched across the next 12 months, tied to each customer’s renewal date.

That didn’t sit right with us. Those customers’ accounts had already been cancelled. Why should their revenue still be counted as if nothing had changed?

Here’s how we changed our calculations to get a clearer picture of Buffer’s finances and a faster feedback loop on how customer experience drives growth.

The change: recognizing churn immediately

Until now, when customers cancelled their Buffer subscription, we continued counting their revenue until the end of their paid period. For example, someone cancelling halfway through an annual plan would remain ‘active’ until the twelve months ended. This method is common in analytics tools, like Chartmogul, because API limitations make it hard to track cancellations immediately. We’ve put in the extra work to overcome that limitation, so our MRR and ARR now reflect cancellations in real time, making our numbers more accurate and responsive.

Going forward, we’re recognizing churn the moment it happens, at the exact point a customer churns. By definition, MRR is meant to reflect the future expectation of monthly recurring revenue. If a customer cancels today, they’re gone. The revenue isn’t “recurring” anymore.

This shift has an immediate impact: our reported MRR/ARR is lower. To put this into perspective, we reported that our closing numbers for July were $1.93M MRR ($23.1M ARR). Those numbers have now been adjusted to $1.84M MRR and $22M ARR.

Today in September, our MRR sits around $1.87M ($22.4M ARR). That’s below some of our recent milestones, like celebrating $23M in ARR and crossing 70,000 paid subscribers. But it’s also a more accurate, real-time reflection of Buffer’s revenue and customer count.

Recognizing churn immediately gives us a clearer picture of the business and a faster feedback loop on how customer experience drives growth. When customers leave, we see it right away. And when they stay, that loyalty shows up more clearly, too.

We’re still syncing the data, but going forward, you’ll see a dip on August 3rd in our transparent metrics when we cancelled those 1,361 inactive annual Buffer subscriptions.

Choosing smaller, more accurate numbers

“We’re doing this because we believe having this responsiveness baked into our metrics will serve us in providing a superior experience.” - Joel Gascoigne, Founder CEO of Buffer

The decision to recognize churn immediately wasn’t a correction or a fix to a mistake. It was a deliberate choice to move away from the default in favor of what we believe is a higher-quality, more transparent methodology.

It’s also a bold choice.

Many companies prefer delay recognizing churn until the end of a customer’s paid period. This makes the reported numbers look larger for longer. We’ve chosen the opposite: to reflect cancellations right when they happen. The result is smaller numbers, but ones that feel more accurate, transparent, and true to our customers’ experience. And because we’re independent, we have the freedom to report in the way we believe is most meaningful.

For us, this is about being genuinely customer-centric and shaping every aspect of how we operate so that it reflects the real experiences of our customers.

What this means going forward

• Our MRR and ARR charts will now move more responsively with customer behavior, both growth and churn.
• Some of our past milestones will look different (we’ll be updating the historical data on our Open page to reflect this methodology).
• Fluctuations may appear sharper, especially around month-ends or when multiple cancellations happen on the same day. We see this as a feature, not a bug: it gives us even more incentive to reduce friction and improve the product experience.

Staying true to transparency

We know this might feel unusual. It’s not common for SaaS companies to voluntarily adopt a methodology that lowers their headline numbers. But as soon as we realized this would improve how we use the numbers, we wanted to share it with you. We believe it strengthens the accuracy and transparency of our reporting, bringing us closer to our customers, which is ultimately our most important goal.

At Buffer, transparency has always been one of our guiding principles. That means sharing not only the highs of our journey, but also the changes we make along the way as we learn and grow, on our way to building the healthiest and most customer-centric company possible.

In the long run, we believe this change will make us a stronger, more resilient company. It gives us clearer insight into the impact of our product and customer experience work, and it ensures that when we celebrate future milestones, they’ll be rooted in the most accurate reflection of our business.