SaaS KPIs That Explain Why Recurring Revenue Grew

SaaS KPIs show whether a subscription business is gaining customers, keeping the ones it has, and earning back the cost of growth. To see why recurring revenue changed, reconcile starting and ending monthly recurring revenue (MRR), a measure of the subscription revenue rate for one month under a consistent company policy. Then check whether the change came from new customers, existing-customer upgrades, or losses. Those movements point to different decisions.

In short: Build an MRR bridge, then use gross revenue retention (GRR) and net revenue retention (NRR) to read the same starting group of customers. Compare customer acquisition spending with gross profit from the customers it won. Investigate the movement that is material, unexplained and most urgent under your cash and customer commitments.

Start with a revenue bridge, not a metric list

Suppose a fictional B2B SaaS company, Atlas, starts April with $100,000 in MRR. By month-end it has $12,000 in MRR from new customers, $5,000 in upgrades by existing customers, $2,000 in downgrades, and $3,000 lost when existing customers cancel. In this simplified month, those upgrade and loss amounts also equal the net start-to-end changes for individual existing customers; no account reversed a change within April. Its ending MRR is:

$100,000 starting MRR + $12,000 new + $5,000 expansion − $2,000 contraction − $3,000 churn = $112,000 ending MRR.

The company grew MRR by 12%. The bridge reveals the source: all $12,000 of net growth came from new customers. Expansion exactly covered the $5,000 lost through downgrades and cancellations. A leader seeing only “MRR up 12%” might increase acquisition spend without investigating those existing-customer losses. Compare the losses with prior months and customer segments before calling them a worsening trend.

This is an illustrative example, not a real result or a benchmark. Atlas had no reactivations, currency movement or other adjustments in April.

In a real bridge, add separate lines for those movements under a written policy; otherwise the beginning and ending balances will not reconcile. Define what qualifies as recurring subscription MRR and exclude one-time setup fees. Record how annual plans, usage charges, discounts, refunds and reactivations enter the series.

MRR is a subscription run rate under that policy; it is not the same thing as recognized revenue, invoices sent or cash collected. Annual recurring revenue (ARR) is commonly the annualized run rate, often MRR × 12 for a simple monthly subscription base. That multiplication does not turn current MRR into a forecast of next year's revenue.

Read retention as a pair

Both retention measures start with the same customers who had positive subscription MRR at the beginning of the period. A cohort is that defined group followed over time. For April retention, freeze the customers with MRR on April 1. The ten customers Atlas won later in April are excluded from retention and appear in the new-customer payback example instead. An invoice collection delay does not itself remove a customer from this cohort under Atlas's stated policy.

Compare each April 1 customer's starting and ending MRR before adding the results. Use that customer's net start-to-end change for retention, even if the billing ledger shows several upgrades and downgrades during April. For GRR, cap each customer's ending MRR at that customer's own starting MRR before summing. For example, if one account goes from $1,000 to $900 and another from $1,000 to $1,100, GRR is $1,900 ÷ $2,000 = 95%; NRR is $2,000 ÷ $2,000 = 100%. The upgrade does not erase the other account's loss in GRR. For Atlas's starting cohort, the aggregate formulas are:

• GRR = (starting MRR − contraction − churn) ÷ starting MRR. Here, ($100,000 − $2,000 − $3,000) ÷ $100,000 = 95%.
• NRR = (starting MRR + expansion − contraction − churn) ÷ starting MRR. Here, ($100,000 + $5,000 − $2,000 − $3,000) ÷ $100,000 = 100%.

The 100% NRR says existing-customer MRR ended flat after expansion. The 95% GRR says Atlas still lost 5% of its starting recurring revenue from that cohort before upgrades. The two measures answer different questions. Reading only NRR would hide the loss; reading only GRR would hide the expansion. These definitions match ChartMogul's GRR and NRR formulas and SaaS Capital's cohort explanation.

Do not compare April's one-month retention with another company's twelve-month retention. State the window, starting cohort, contract mix, and whether price increases count as expansion. Review customer-count retention alongside revenue retention: losing many smaller customers can coexist with steady NRR when a few larger accounts expand. That pattern asks for a segment and customer-reason investigation, not an automatic conclusion that the product is healthy.

Check whether new growth pays back its acquisition cost

Atlas's $12,000 of new MRR came from ten new customers. Suppose $96,000 of sales and marketing cost is attributable to acquiring that group, and the new subscriptions have an estimated 80% gross margin. Gross margin is the share of subscription revenue left after defined costs to deliver and support the service. Customer acquisition cost (CAC) is the acquisition spend assigned to the won customers, divided by their number.

• New-customer CAC: $96,000 ÷ 10 = $9,600 per new customer.
• Monthly gross profit from the cohort: $12,000 × 80% = $9,600.
• Illustrative CAC payback: $96,000 ÷ $9,600 = 10 months at that monthly gross-profit level.

This is a simple steady-state estimate. It assumes the cohort keeps paying at that rate, its service cost and margin stay similar, and the $96,000 acquisition spend belongs to those ten customers. Actual cash recovery can differ because contracts, collection timing, expansion, contraction and churn change the cash and profit path. Stripe's CAC payback explanation uses acquisition cost divided by monthly profit from acquired customers.

For a group of won customers, assign the sales and marketing costs incurred to win them using a documented attribution window and cost policy. Shared spend may require an estimate. State how you treat salaries, commissions, campaigns, onboarding and partner fees; do not count the same onboarding cost in both acquisition spend and service costs. Do not divide this quarter's entire acquisition expense by an unrelated month's new MRR and call the result a cohort payback period.

Whether ten months is workable depends on Atlas's cash plan, contract length, margin and funding position. No single payback target fits every SaaS company. The next section shows which records Atlas should inspect under each cash condition.

Choose the first investigation when signals conflict

Use the bridge to identify the movements, then compare each with its own prior months and customer segments. Ask whether a change is material to revenue or cash, whether its cause is still unknown, and whether a team can act before more loss occurs. For Atlas, the cash constraint determines the first sheet to pull:

• If ten-month payback breaches the cash plan: Inspect the acquisition-cohort sheet first. Link attributable spend, the ten new customers, their new MRR, gross margin and observation date. Test whether cost attribution, margin or customer mix explains the gap before changing spend.
• If payback fits the cash plan: Inspect the April 1 account list first. Include starting and ending MRR, downgrade or cancellation reason, segment, activation status and account owner. Compare it with earlier months. If losses cluster among accounts that never reached the first useful product action, test the onboarding explanation; if they cluster in a segment Atlas intentionally does not serve, test qualification.

These are investigation priorities, not automatic instructions to cut spend or change the product. Neither one-month metric diagnoses the cause by itself. Record the decision in one line: “We will inspect [specific movement and cohort] because [amount, trend or cash constraint]; [owner] will compare [records] by [date], and we will change [decision] if [finding].”

Use earlier KPIs to explain the movement

Revenue and retention are outcomes. When one moves, add a measure that can explain it and name the action it could change.

If new MRR slowed: inspect qualified pipeline and conversion

Count opportunities that meet a written customer and deal definition, then track progression from qualified opportunity to closed customer by source and cohort. Define the denominator, for example eventual won deals from opportunities created in April ÷ all qualified opportunities created in April, and give that cohort enough time to close before comparing it with another. A lower win rate may come from a different buyer mix, weaker qualification, pricing, competition or sales execution. Review several lost and won deals before prescribing more outbound volume.

For an outbound channel, meetings booked alone can reward activity that never becomes qualified pipeline. Track relevant replies or meetings only alongside accepted opportunities and eventual customers. Alore's data-driven sales guide offers a broader approach to choosing a sales measurement from the decision it will inform.

If GRR fell: inspect activation and reasons for loss

Choose one product action that a customer must complete to get the value it bought, such as connecting its data source and completing a first live workflow. Define activation rate = customers in a signup or onboarding cohort who complete that action within a stated time ÷ all eligible customers in that cohort. Segment the rate by acquisition source, plan and customer size. Then compare those cohorts with later downgrades and cancellations. A low activation rate is a diagnostic lead, not proof of cause; interview or inspect the affected accounts before changing onboarding.

Use both customer churn (starting-cohort customers lost ÷ customers at the period start) and gross revenue churn ((starting-cohort MRR lost through cancellations and contractions) ÷ starting MRR). They diverge when accounts have different contract values. Keep cancellation and downgrade as separate components within gross revenue churn because each may need a different response. State how any within-period reactivation is treated.

If NRR rose while GRR fell: inspect who expanded

Separate upgrades from price changes and usage growth. See whether expansion is spread across a healthy set of customers or concentrated in a few large accounts. Read that distribution beside GRR and customer-count retention. If expansion from one segment covers many smaller losses, the next question is whether the lost segment is poorly served, poorly qualified, or intentionally outside the company's target market. The decision depends on account-level evidence, not the NRR percentage alone.

Which SaaS KPIs belong on the leadership dashboard?

For a subscription-led B2B SaaS business, keep three paired views on the core monthly page. Open two diagnostic views when a movement needs explaining. Each earns its place by answering a decision question.

Core monthly views

• MRR bridge and ARR run rate: How much did recurring revenue change, and which movement caused it? Keep ARR as a run-rate label, separate from recognized revenue and cash.
• GRR, NRR and customer retention: Are existing customers staying, paying less, or expanding? Use one starting cohort and one time window.
• New MRR, CAC and gross-margin payback: Is acquisition creating enough recurring gross profit to recover its cost on a viable timeline? Match the customers and spend.

Open when needed

• Qualified pipeline and cohort conversion: Is future new MRR constrained by relevant opportunities or by converting them? Define stages and allow for sales-cycle lag.
• Activation by customer cohort: Are new customers reaching the first value-producing action? Use it to investigate later retention, not to replace retention.

The exact set can change with the business model. Usage-priced products may need a committed-versus-consumed revenue view; a services-heavy company needs the subscription and services economics separated. Product engagement, net promoter score (NPS) and lifetime value (LTV) can answer narrower questions. LTV estimates depend on margin, expected customer lifetime, churn definition and segment, so label those inputs before using an LTV:CAC ratio as a spending rule.

The Rule of 40 adds revenue growth rate and profit margin to give one combined percentage. It is an investor shorthand, with variations in which growth and margin definitions are used. It cannot reveal whether Atlas's April growth came from new sales or whether expansion concealed downgrades. If you use it, state the period and definitions and retain the operating measures above.

Turn the dashboard into a monthly decision

At the close of each month, reconcile beginning and ending MRR first. Then review GRR and NRR for the same starting cohort and the customer-count view by segment. Finally, inspect acquisition economics for a completed new-customer cohort and only the leading measures needed to explain a material change. Record one decision, its owner, and the next observation that would confirm or disprove it.

For fictional Atlas, the April 1 account list is the next useful output if the ten-month estimated payback fits its cash plan. It will show which accounts produced the $5,000 loss and whether those losses differ from prior periods or segments. If payback breaches the cash plan, the acquisition-cohort sheet is equally urgent. The numbers locate the investigation; customer and cost records determine the fix.