The skill: Calculating and tracking the financial metrics that tell you whether your SaaS business is healthy, growing efficiently, and fundable. Product analytics tell you if users get value. Financial metrics tell you if the business captures value. Most startups track these wrong.
Benchmarks
In a Nutshell
- MRR is the foundation. Monthly Recurring Revenue is the single number your entire financial picture builds on. Get this wrong and everything downstream is wrong.
- Track the four MRR components, not just the total. New MRR, Expansion MRR, Contraction MRR, Churned MRR. The waterfall is the story. The total is just the ending.
- NRR is the best single metric for SaaS health. Net Revenue Retention above 100% means you grow even if you stop acquiring customers. Below 100% means you are filling a leaking bucket.
- CAC must be fully loaded. If your CAC only includes ad spend, it is fiction. Salaries, tools, content production, events, everything that contributes to acquiring a customer goes in.
- LTV:CAC ratio is overrated. CAC payback period is underrated. A 5:1 LTV:CAC sounds great until you realize the payback takes 24 months and you run out of cash first.
- Burn Multiple tells you how efficiently you convert cash into growth. Net burn divided by net new ARR. Below 1x is elite. Above 2x is a problem.
- Always segment financial metrics. Blended CAC across channels is a vanity metric. Per-channel CAC is what drives allocation decisions.
- Revenue recognition is not optional. Bookings are not revenue. A $120K annual contract signed today is $10K MRR, not $120K. Mixing bookings and recognized revenue will mislead every decision.
The MRR Waterfall
The single most important financial view for a SaaS startup is the MRR waterfall. It breaks total MRR change into four components:
New MRR — Revenue from customers who started paying this month. First payment only. If a customer signed up last month and their first charge hits this month, it counts here.
Expansion MRR — Additional revenue from existing customers. Upgrades, seat additions, usage increases. This is the engine behind NRR above 100%.
Contraction MRR — Revenue lost from existing customers who downgraded but did not cancel. They are still paying, just less.
Churned MRR — Revenue from customers who cancelled entirely. Gone.
The formula: Ending MRR = Starting MRR + New MRR + Expansion MRR - Contraction MRR - Churned MRR
This waterfall tells you things the total cannot. If MRR grew $20K this month, that could be $25K new and $5K churned (healthy growth) or $50K new and $30K churned (growth masking a retention crisis). Same total, completely different businesses.
Defining Each Metric
MRR (Monthly Recurring Revenue) — The normalized monthly value of all active recurring subscriptions. Annual contracts are divided by 12. Quarterly contracts by 3. One-time fees, setup charges, and professional services are excluded. Always excluded: free trials, paused accounts, and pending cancellations that have not yet reached their billing period end.
ARR (Annual Recurring Revenue) — MRR × 12. Not the sum of annual contracts. This is a common mistake. ARR is a projection of your current run rate, not a sum of contract values.
Net Revenue Retention (NRR) — The percentage of revenue retained from a cohort of customers after a period, including expansion. Trailing 12-month calculation: take the MRR from customers who existed 12 months ago, see what those same customers pay today (including upgrades and downgrades, excluding churned), divide by their original MRR. A 120% NRR means that last year's $100 in MRR is now $120, without a single new customer.
Gross Revenue Retention (GRR) — Same as NRR but excludes expansion. It can never exceed 100%. It measures pure retention: of the dollars you had, how many did you keep? GRR below 85% means you have a leaky bucket that expansion alone cannot fix.
CAC (Customer Acquisition Cost) — Total sales and marketing spend in a period divided by new customers acquired in that period. Fully loaded means: salaries (sales, marketing, SDR teams), tools (CRM, ad platforms, email tools), ad spend, content production, events, and allocated overhead. Calculate per channel: paid CAC, organic CAC, outbound CAC. Blended CAC is useful for board reporting but useless for operational decisions.
LTV (Lifetime Value) — Average revenue per account divided by revenue churn rate. Or more precisely: ARPA × gross margin / monthly revenue churn rate. The gross margin adjustment matters. A $100/month customer with 70% gross margin generates $70/month in actual value. Without the margin adjustment, you will overestimate how much you can spend to acquire a customer.
LTV:CAC Ratio — LTV divided by CAC. The benchmark is 3:1 or higher. Below 3:1 means you are spending too much to acquire customers relative to what they are worth. Above 5:1 might mean you are underinvesting in growth. But this ratio hides a critical dimension: time.
CAC Payback Period — Months to recoup the cost of acquiring a customer. CAC divided by (ARPA × gross margin). This is more actionable than LTV:CAC because it directly maps to cash flow. A 3:1 LTV:CAC with an 18-month payback means you need 18 months of cash per customer before you break even. For a startup burning cash, this number determines whether you can actually afford to grow.
Burn Multiple — Net burn divided by net new ARR. If you burned $500K and added $400K in net new ARR, your burn multiple is 1.25x. David Sacks's framework: below 1x is amazing, 1-1.5x is great, 1.5-2x is good, 2-3x is suspect, above 3x is bad. This metric cuts through the noise of "growth at all costs" and asks: for every dollar you burn, how much recurring revenue did you create?
SaaS Quick Ratio — (New MRR + Expansion MRR) / (Contraction MRR + Churned MRR). Measures growth efficiency. A Quick Ratio of 4 means for every $1 lost, you add $4. Below 2 and you are fighting a losing battle.
Gross Margin — (Revenue - Cost of Goods Sold) / Revenue. For SaaS, COGS includes: hosting and infrastructure, third-party API costs, customer support team costs, onboarding and implementation costs, and payment processing fees. It does not include R&D, sales, or marketing. Software companies should target 70%+ gross margins. Below 60% and investors will question whether you are really a software company.
Rule of 40 — Revenue growth rate (%) + profit margin (%). A company growing 60% with -20% margins scores 40. A company growing 20% with 20% margins also scores 40. Both are healthy. This metric matters for fundraising and exits. It acknowledges that growth and profitability are substitutes: you can trade one for the other, but the combined score should be at least 40.
Do's and Don'ts
Common Calculation Errors
The formula is usually right. The inputs are where things break.
MRR inflation. Including one-time setup fees, consulting revenue, or implementation charges in MRR. These are real revenue but they are not recurring. A $5K setup fee is not $5K MRR. It is $5K of non-recurring revenue. Mixing them makes your MRR look better than it is and throws off every downstream metric.
Annual contract normalization. A customer signs a $24K annual contract. That is $2K MRR, recognized on day one of the contract. It is not $24K in the month they signed (that is bookings) and it is not $0/month until they renew (that is cash-basis, not MRR). Annual contracts must be divided by 12 and spread evenly. Quarterly contracts by 3.
Partial-month CAC. If a customer signs up on the 28th of the month, their first month of MRR is prorated. But their CAC is not prorated. The marketing spend that acquired them happened across the full month. This mismatch is small month-to-month but compounds in LTV:CAC calculations. Standardize: use full calendar months for both.
Survivorship bias in NRR. Some teams calculate NRR by looking only at customers who are still active. That excludes the churned revenue by definition and inflates the number. NRR must include churned customers in the denominator. If 100 customers paid you $100K last year and 80 of them pay you $90K today, your NRR is 90%, not 112.5%.
Gross margin misclassification. Engineering salaries are not COGS. R&D is not COGS. But the DevOps engineer who keeps the production infrastructure running might be. The support team that handles customer tickets is. The line between COGS and operating expense is judgment-dependent, and startups routinely classify too little as COGS, inflating gross margins. This matters because LTV calculations use gross margin, and an inflated margin produces an inflated LTV.
The Three Financial Dashboards
You need three views, not one.
Weekly operational dashboard. MRR waterfall (new, expansion, contraction, churned), Quick Ratio, trailing 4-week trends. This is what the team looks at in the Monday meeting. It should answer: are we growing, and is the growth healthy? No more than 6-8 metrics. Updated automatically from your billing system.
Monthly board dashboard. MRR, ARR, NRR, GRR, CAC by channel, LTV:CAC, CAC payback, Burn Multiple, runway. This is what investors and advisors see. It should answer: is this business efficient, retaining, and fundable? Calculated on the first of each month from finalized data.
12-month rolling view. Every metric above on a trailing 12-month basis. Month-to-month financial metrics are noisy. A single large churn event or a big enterprise deal can distort a month. The rolling view smooths that noise and shows the real trend. If your NRR is 115% this month but the 12-month trailing is 105%, the 115% is an anomaly, not a trend.
In Practice
A B2B SaaS startup at $80K MRR reported a "healthy" LTV:CAC of 4:1 to their board. The math looked right: $1,200 ARPA, 3% monthly churn, $100 blended CAC. But three things were wrong.
First, the CAC only included ad spend. When fully loaded with the two-person sales team, marketing tools, and content costs, true CAC was $800, not $100. LTV:CAC dropped to 0.5:1.
Second, the LTV calculation used revenue, not gross margin. Their gross margin was 65% (high infrastructure costs for a data-heavy product). Margin-adjusted LTV dropped from $40K to $26K.
Third, they calculated NRR excluding churned customers. The real NRR was 92%, not the 108% they reported. Their "expansion" was being eaten by churn they were not counting.
After correcting all three errors, the picture changed completely. They were spending $800 to acquire customers worth $26K over their lifetime (a legitimate 32:1 ratio, but over years), with a 12-month payback. The business was actually solid, but the 12-month payback meant they needed to manage cash carefully. The original 4:1 was not just wrong. It was hiding the cash constraint that almost caused them to over-hire.
Written with ❤️ by a human (still)