Dalal Street Blogs provides information for educational purposes only and does not constitute investment advice. Please consult a SEBI-registered investment adviser before making any investment decisions.
CAUTION: Dalal Street Blogs publishes market news and educational content only. We do not promote, recommend or endorse any stock, and nothing here is a solicitation to buy or sell securities. Please consult a SEBI-registered adviser before investing.
Results

Gross Revenue Retention: The Software Metric That Ignores Upsells and Shows the Leak

4104
By Dalal Street Blogs Editorial10 min read

Here’s a number that rarely makes the headline slide of a software company’s investor deck: gross revenue retention. The headline usually goes to net revenue retention, the figure that can sit comfortably above 100% and make the business look like it grows by itself. Gross revenue retention is the quieter sibling. It ignores every upsell and every price hike and asks one blunt question: of the recurring revenue you had a year ago, how much did you hang on to?

That makes it the better leak detector. If you’re reading the results or investor presentation of a SaaS company, including the Indian-founded ones that report in dollars, this is the metric that tells you whether the bucket has holes before the expansion revenue gets poured in on top.

In this article
  1. What gross revenue retention actually measures
  2. The two ways to calculate it
  3. A worked example with round numbers
  4. Why the gap between GRR and NRR is the real story
  5. Quarterly, monthly or annual: watch the period
  6. Definitions that can flatter the number
  7. Where GRR fits when you read results
  8. Questions worth asking on an earnings call
  9. The short version

What gross revenue retention actually measures

The SaaS Metrics Standards Board, a New York-based body that publishes standard definitions for subscription metrics, defines gross revenue retention (GRR) as the percentage of recurring revenue retained over a specific period. Its other name is gross dollar retention (GDR), so if a presentation uses that term, it’s the same idea.

Only two things can pull GRR down:

  • Churn, meaning customers who cancel or don’t renew.
  • Down-sell, or contraction, meaning existing customers who stay but cut their commitment: fewer seats, a cheaper plan, a smaller usage tier.

What it deliberately leaves out is the good news. Upsells, cross-sells and usage expansion from existing customers don’t count. That’s the whole difference from net revenue retention, which nets those gains against the losses. Because GRR never adds anything back, the Standards Board is explicit that it can never exceed 100%. A company quoting a GRR of 104% has either mislabelled its net figure or invented a new kind of arithmetic.

Net revenue retention tells you how fast the bucket is filling. Gross revenue retention tells you how big the hole at the bottom is.

The two ways to calculate it

The Standards Board describes two methods and makes clear which one it prefers.

The cohort method (the preferred one)

Take every customer who was active at the start of the measurement period, usually one year earlier. Note their monthly recurring revenue (MRR) then. Note their MRR now. Then, for each customer, take the lesser of the two numbers. The Standards Board calls this the “adjusted MRR”. If a customer grew, you cap them at their starting figure. If they shrank or left, you take the smaller number or zero.

GRR is then simply total adjusted MRR at the end divided by total MRR at the start. Capping each customer is the neat trick. It strips out expansion mathematically while still catching every rupee or dollar of shrinkage and churn, so you don’t need to track contraction separately.

Two practical points from the standard are worth keeping in mind when you read a company’s number. First, it should be built on recognised revenue, not bookings, billings or cash. Second, companies with large enterprise contracts and long implementation cycles may use annual recurring revenue (ARR) or contracted ARR instead of MRR. Using contracted ARR catches customers who sign, then cancel before the software is even live, which an MRR-based figure would miss.

The formula method

The shortcut version is: (beginning MRR minus churned MRR minus down-sell MRR) divided by beginning MRR, times 100. The Standards Board says this is common in businesses with lower contract values and relatively high churn, and it names B2C and telecom companies that benchmark on monthly churn as typical users. It also suits fast-growing firms where many current customers haven’t been around for a full year.

The catch is that some churn during the period may come from customers who weren’t in the starting base at all. Those losses end up divided by a denominator they never belonged to. The standard notes the error grows over longer measurement periods, which is why it calls the cohort approach the more accurate one.

A worked example with round numbers

Let’s build one. The figures below are hypothetical, chosen to keep the maths easy. Imagine a small software company with six customers a year ago, measured in annual recurring revenue in ₹ crore.

CustomerARR a year ago (₹ cr)ARR now (₹ cr)What happenedAdjusted ARR (lesser of the two)
A1014Expanded10
B88Flat8
C60Churned0
D129Down-sold9
E47Expanded4
F1013Expanded10
Total505141

Net revenue retention is 51 divided by 50, or 102%. On a slide, that reads as “our existing customers are growing”. Gross revenue retention is 41 divided by 50, or 82%. That reads very differently. The company lost ₹9 crore of the ₹50 crore it started with, ₹6 crore from customer C walking away and ₹3 crore from customer D cutting back, and only made it look fine because three customers bought more.

Run the formula method on the same set and you get the same answer here: 1 minus (6 + 3) divided by 50, which is 82%. The two methods match in this example because no new customer joined and left inside the year. In real life they often won’t.

Why the gap between GRR and NRR is the real story

Two companies can post the same net revenue retention and be completely different businesses. Take two hypothetical firms, both with NRR of 115% on a starting base of 100.

  • Company X loses 3 to churn and down-sell and gains 18 from expansion. Its GRR is 97%.
  • Company Y loses 20 and gains 35. Its GRR is 80%.

Same headline. Very different risk. Company X’s customers almost never leave, so its revenue base is sticky and the expansion is a bonus. Company Y has to win back a fifth of its base every year from the same customers just to stand still, and then sell them even more on top. If its sales team has a bad year, or customers start trimming software budgets, the expansion dries up first and the 20% leak is still there.

When expansion slows, net retention falls fast. Gross retention is the floor it falls towards.

This is why the Standards Board’s own description is worth reading closely. It calls GRR a critical metric because it directly measures the stability of a SaaS business’s revenue, and says companies with high GRR are more predictable because churn isn’t a significant drain. We’d put it more simply: GRR is the part of the revenue you don’t have to re-earn.

Quarterly, monthly or annual: watch the period

A GRR number means little until you know what period it covers. The standard says the typical period is one year, and that an annual figure is the most intuitive one to benchmark.

Shorter periods can be annualised by raising the result to a power: the fourth power for a quarter, the twelfth for a month. The Standards Board’s own example shows why this matters. A monthly GRR of 98.5% sounds superb, but annualised it becomes about 83.4%. The same 98.5% measured over a quarter annualises to about 94.1%.

Try it with a number of your own. A quarterly GRR of 96% annualises to 0.96 to the fourth power, which is roughly 84.9%. So a company boasting “96% gross retention” in a quarterly update is, on an annual basis, losing about 15% of its starting recurring revenue. Same business, very different impression. When you compare two companies, make sure both figures cover the same period, or convert them first.

The standard also warns that monthly numbers might not be appropriate for annualisation because they bounce around a lot. One unusually quiet month can make the annualised figure look far better than the business really is.

Definitions that can flatter the number

Because there’s no accounting standard forcing companies to report GRR, each one picks its own definition. That’s where you need to read the footnotes. A few things to check, most of them drawn from the nuances the Standards Board lists:

  1. Is it recognised revenue or bookings? The standard says recognised revenue. Bookings can include contracts that haven’t started, which hides churn that’s already been agreed.
  2. How is usage-based revenue treated? The Standards Board recommends excluding variable usage-based revenue from GRR, because it swings with usage rather than with retention. A company that includes it could see its GRR jump around for reasons unrelated to customers staying or leaving.
  3. What’s the win-back window? The standard says most companies define a win-back period of 30 to 90 days. A customer who churns and returns within it is counted as a renewal, not a loss and a new sale. A generous window can make churn quietly disappear.
  4. Are accelerated revenues from cancelled contracts stripped out? The standard says revenue recognised early because a contract was cancelled shouldn’t sit in either the beginning or ending figure. If it does, a cancellation can briefly look like retention.
  5. Which customers are in the cohort? The cohort method can’t see customers acquired during the year, and it isn’t much use for a young business with few customers older than a year. For a fast-growing company, ask what share of revenue the cohort actually covers.

If a company changes its definition from one year to the next, treat the old and new figures as two different metrics. Comparing them directly is how investors get fooled by a “jump” in retention that is really a change in the maths.

Where GRR fits when you read results

GRR isn’t something you’ll find in the audited profit and loss statement. It usually turns up in investor presentations, earnings call commentary or the shareholder letter, and companies differ in whether they disclose it at all. When it is there, we’d read it alongside a few other items.

  • Net revenue retention. The gap between NRR and GRR is the expansion engine. A widening gap with a stable GRR is healthy. A stable NRR propped up by rising expansion while GRR slides is a warning.
  • Deferred revenue. Money billed but not yet earned shows up as a liability on the balance sheet. A shrinking deferred revenue balance can be an early sign that renewals are slowing.
  • Deal sizes and contract values. Large new wins, often reported as total contract value, can mask a weak base. GRR tells you whether the old contracts are holding while the new ones are being signed.
  • Customer counts. The Standards Board also publishes a separate logo retention standard, which counts customers rather than revenue. If logo retention is weak but GRR is strong, small customers are leaving while big ones stay. That’s a different risk from losing one large account.

For the bigger framework of reading a results release line by line, our guide on how to analyse quarterly results of a company covers revenue, margins and the cash flow checks that sit around metrics like this.

Questions worth asking on an earnings call

If you get a chance to listen to management, or you’re reading a transcript, these are the questions that tend to get the most useful answers:

  • Is the reported GRR on a cohort basis or a formula basis, and over what period?
  • Is it based on ARR, MRR or contracted ARR?
  • Is usage-based revenue in or out?
  • How much of the drop was full churn and how much was down-sell? A customer cutting seats is a different problem from a customer leaving.
  • Is the churn concentrated in one segment, such as small businesses, or one region?

Vague answers to the first three are themselves information. A company confident in its retention usually explains its definition without being asked.

The short version

Gross revenue retention is the share of last year’s recurring revenue a company still has, counting losses from churn and down-sell but ignoring every bit of expansion. It can’t go above 100%. The cohort method, which caps each customer at their starting revenue, is the more accurate way to calculate it. Always check the period, because a monthly or quarterly figure looks much better than its annualised equivalent, and always check the definition, because usage revenue, win-back windows and bookings can all flatter it. Read it next to net revenue retention. The gap between the two shows how much of a software company’s growth depends on selling more to customers it is also losing.

Disclaimer: This article is for educational purposes only and is not investment advice. The worked examples use hypothetical numbers. Investing in equities involves risk of loss. Please consult a SEBI-registered investment adviser before making investment decisions.

Leave a Reply

Join the conversation

Commenting is open to registered readers. Sign up with just your email — no password needed, and it takes under a minute.

Sign up to comment Log in