%20scilla.studio.svg)
You can benchmark your startup's growth with six numbers, a spreadsheet you already have, and about an afternoon (no analytics hire, no data warehouse, no SQL). Pull your retention, activation, LTV:CAC, CAC payback, K-factor, and MAU growth, line each one up against the B2B SaaS or consumer band below, and read the shape of the gap rather than the gap itself. That's the whole job. The rest of this guide is how to do it without fooling yourself.
Here's what I actually think: most founders skip benchmarking not because it's hard but because they assume it needs infrastructure they don't have. It doesn't. The numbers that tell you whether you're healthy are the ones you can count by hand from Stripe, your auth provider, and your product database. The trap is the framing: treating a benchmark as a target instead of a piece of context.
Benchmarking is comparing your real numbers to a known range so you can tell the difference between "we have a problem" and "this is normal for our kind of product." That's it. It is not a scoreboard, and a single number in isolation will lie to you almost every time.
Two rules before any number goes on the page:
So pick now. Do people buy your product to do a job at work, often paid by a company? That's B2B SaaS. Do individuals download or sign up for themselves? That's consumer. Every band below comes in both flavors.
You need exactly six. For each, I'll give you the plain definition, where to find the raw number, and the benchmark band. Then we'll talk about how to read them together, which is the part most guides skip.
Retention is the percentage of a cohort still active some number of days after they first used the product. It's the single most honest signal you have, because no amount of paid acquisition can fake it.
How to pull it without a data team: take everyone who first did the core action in a given week (call it the cohort), then count how many of them came back on day 1, day 7, day 14, and around day 90. A pivot table on an export of user_id, first_active_date, last_active_date gets you most of the way.
| Retention | B2B SaaS | Consumer | What it tells you |
|---|---|---|---|
| Day 1 | 50 to 70% | 20 to 30% | Users who activated once usually return at least once more. Lower suggests onboarding or value-clarity problems. |
| Day 7 | 40 to 60% | 8 to 15% | Early habit formation or workflow relevance. |
| Day 14 | 35 to 55% | 4 to 8% | The drop should be flattening by now; a steep decline signals a weak core loop. |
| Day 90 | 25 to 35% | 1 to 4% | Long-term product value. Below 20% is a red flag for B2B; above 5% is exceptional for consumer. |
Sources: Pendo Product Benchmarks, Amplitude, Mixpanel (B2B); Adjust, AppsFlyer, UXCam (consumer).
The part most guides skip: the absolute day-90 number matters less than whether the curve flattens. A retention curve that keeps falling toward zero is a leaky bucket no acquisition budget can fill. A curve that drops and then goes flat, even at a modest level, means you've found a group of people for whom the product genuinely sticks. Look for the flattening. (More on this in why retention drops after day 1 and the full retention rate benchmarks.)
Activation is the percentage of new users who reach the moment where they actually experience the product's value: completed onboarding, then did the thing the product is for.
How to pull it: define the one event that means "this person got it" (created their first project, sent their first invoice, whatever your "aha" is), then divide the users who hit it by all the users who signed up in the same window.
I'll be honest about these: activation is the one metric here I won't give you a hard table for, and that's deliberate. The framing is well-established (Lenny's Newsletter, Userpilot, OpenView's PLG benchmarks all converge on it), but the per-number bands you see quoted around the web vary wildly with how each product defines its "aha" moment, so a single number would be false precision dressed up as a benchmark. Here's the directional read instead:
Use these to ask "are we roughly in the neighbourhood, and is our Day-7 activation well below our onboarding completion?", not to grade yourself to the percentage point. The gap between completing onboarding and reaching value is the signal worth chasing. (See activation rate for the deeper treatment.)
LTV:CAC is the ratio of the lifetime value of a customer to what it cost to acquire them. It answers: for every krona (or dollar) we spend getting a customer, how much do we get back over their lifetime?
Formula: LTV ÷ CAC. You can pull CAC from total sales-and-marketing spend ÷ new customers in the same period, and a rough LTV from average revenue per account × average customer lifetime (in the same time unit). No data team required; this lives in your finance spreadsheet.
| LTV:CAC | B2B SaaS | Consumer |
|---|---|---|
| Healthy range | 3:1 to 5:1 | 2:1 to 4:1 (ideal ≈ 3:1) |
| Comment | ~3:1 is the minimum healthy baseline. Below 2:1 is structurally risky. | Consumer growth is faster but less durable; below 2:1 burns cash. |
Sources: Bessemer State of the Cloud, OpenView SaaS Benchmarks, a16z, Wall Street Prep (B2B); Adjust, AppsFlyer, a16z (consumer).
Here's the interesting part, and it's the most counterintuitive line in this whole article: an LTV:CAC above 5:1 is often a warning sign, not a trophy. It usually means you're underinvesting in growth, leaving acquisition on the table that you could profitably buy. A "too good" ratio can mean timidity dressed up as discipline. (And yes, that means your "great" number might be the thing holding you back.) Full breakdown in LTV:CAC ratio.
CAC payback period is how many months it takes for a customer's gross margin to repay what you spent acquiring them. It's the cash-flow companion to LTV:CAC. The ratio tells you whether the unit economics work; payback tells you how fast you get your money back.
| CAC payback | B2B SaaS | Consumer |
|---|---|---|
| Healthy range | 6 to 12 months (SMB / self-serve) 12 to 24 months (enterprise sales) |
1 to 6 months |
| Comment | Under 12 months is strong. Longer is acceptable only with very high retention and expansion. | Consumer products are expected to recoup CAC fast; over 6 months usually fails at scale. |
Sources: OpenView, KeyBanc SaaS Survey (B2B); AppsFlyer, Mobile Dev Memo (consumer).
Read LTV:CAC and payback together. A 4:1 ratio with a 30-month payback can still starve you to death: the lifetime value is real but you're floating every customer's acquisition cost for two and a half years. If your payback looks scary, why your CAC payback period is too long walks through the usual culprits.
K-factor (the viral coefficient) is how many new users each existing user brings in through sharing, invites, and other viral loops before their influence runs out. K = 0.3 means every 100 users generate 30 more.
| K-factor | B2B SaaS | Consumer |
|---|---|---|
| Typical range | 0.1 to 0.3 | 0.3 to 0.7 (rarely > 1) |
| Comment | B2B virality is usually weak; above 0.3 is unusually strong unless you have built-in collaboration loops. | Even 0.5 is very strong; above 1 is true viral growth and extremely rare. |
Sources: Reforge, Andrew Chen.
Two honest notes. First, most products are not viral, and that's fine. A K-factor of 0.2 with great retention beats a flashy 0.6 that never compounds. Second, K-factor without cycle time (how fast the loop completes) overstates your virality. A high K with a slow loop grows slowly; a modest K with a fast loop can outpace it. We dig into that distinction, and why measuring K well means instrumenting every loop separately (K = outputs per user × signups per output, summed across loops) rather than eyeballing one referral number, because cross-loop attribution is genuinely hard to instrument, in how to improve your K-factor without referral programs.
MAU growth rate is the month-over-month percentage change in your monthly active users. It's the headline "are we growing" number, but it's also the easiest to misread, because growth can come from acquisition you're renting rather than retention you own.
How to pull it: count distinct active users this month, divide by last month's count, subtract one. One row in a spreadsheet.
There isn't a single universal "good" MAU growth band the way there is for retention or unit economics; healthy growth rate depends heavily on your stage and starting size. What matters more is the engine underneath it: the cohort model that powers our benchmark tool treats MAU growth as the output of two levers you can benchmark, monthly retention (M1) and K-factor, feeding a stable base of long-tenured users plus each month's new cohorts. If your MAU is climbing but retention is weak, you're refilling a leaky bucket, and the growth will reverse the moment you stop spending. See MAU growth rate for how to separate rented growth from owned growth.
Pulling six numbers is the easy 80%. Reading them as a system is the 20% that decides whether benchmarking actually changes a decision. A few rules:
Retention before efficiency. Check retention first. If retention is below band, fix that before you touch CAC, payback, or K-factor, because every other metric is built on top of users who stay. Cheap acquisition into a leaky product just loses money faster.
Look for the flattening, not the absolute. This applies to retention and to any cohort curve. A curve that drops then plateaus means you've found genuine fit with someone. A curve that keeps sliding means you haven't yet.
Numbers interact, so never read one alone. A "great" 5:1 LTV:CAC can mean underinvestment. A "great" MAU growth rate can be masking a retention hole. A "decent" K-factor with a 60-day cycle can be near-worthless. The single most common benchmarking mistake is grading one number, declaring victory or panic, and missing the metric next to it that changes the story entirely.
Your business model moves the goalposts. Freemium, self-serve, and enterprise-sales motions have genuinely different "good." A 24-month payback is alarming for self-serve and completely normal for enterprise. Match the band to your motion, not the average of all motions.
No growth hacks; those don't survive contact with reality. Concrete levers, honestly framed:
You don't need to build any of this by hand. The free benchmark tool charts your retention, LTV:CAC, payback, and K-factor against the B2B SaaS and consumer bands above, and models your MAU growth from monthly retention (M1) and K-factor (so you see the shape of your curves, not just the headline numbers) in a couple of minutes. Plug in the six numbers from your spreadsheet and read the gaps in context: benchmark.scilla.studio.
Do I really need a data team to benchmark growth? No. The six metrics that matter (retention, activation, LTV:CAC, CAC payback, K-factor, and MAU growth) can all be pulled from data you already have (Stripe, your auth provider, a product-database export) using spreadsheet pivot tables. A data team makes it faster and more granular, but it isn't a prerequisite for the basics.
Which metric should I benchmark first? Retention. It's the most honest signal of whether people actually want the product, and every efficiency metric (CAC, payback, LTV) is built on top of users who stay. Check retention before you optimize acquisition.
Are these benchmarks targets I should aim for? No. They're context, not targets. A band tells you where comparable products land so you can spot where to look closer. Hitting it doesn't mean you've won; missing it means investigate, not panic.
Can I compare my B2B product to consumer app benchmarks? No. B2B SaaS and consumer apps have fundamentally different retention shapes, viral mechanics, and payback expectations. Pick the profile that matches how people buy your product (work tool vs. individual signup) and only compare within it. See B2B SaaS growth benchmarks and consumer app benchmarks.
What's a good day-90 retention rate? For B2B SaaS, roughly 25 to 35% is a strong signal of long-term value, and below 20% is a red flag. For consumer apps, 1 to 4% is normal and anything above 5% is exceptional. But the flattening of the curve matters more than the single number: a curve that plateaus, even low, beats one that keeps sliding.
