%20scilla.studio.svg)
A product benchmark tells you where you stand relative to other products. It does not tell you where you should land. The moment you turn a benchmark into a target ("we need Day-7 retention of 50% because that's the B2B average"), you've quietly swapped your product's real goal for someone else's median. That's the mistake this whole article is about, and avoiding it is the single most useful thing a benchmark can do for you.
Here's what I actually think: most teams don't get burned by not knowing the benchmarks. They get burned by trusting one number too much, out of context, and steering toward it. Benchmarks are directional references; they orient you. Targets are commitments; they bind you. Treating the first like the second is how good teams end up optimizing a metric while the business underneath it stalls.
A benchmark is a directional reference: the typical range a comparable product sees for a metric, used to interpret your own number.
A target is a number you commit to hitting, derived from your own goals, constraints, and customer reality.
The benchmark answers "is this normal?", not "is this what we should aim for?" The target answers the second question. Same shape (a number), completely different job.
Quick test for which one you're holding: if the number changes when you switch from B2B to consumer, from self-serve to enterprise, from a daily-use tool to a quarterly-use one, it's context. Real targets come from your model, not the industry's average. (And yes, that means two healthy products can sit at opposite ends of the same band for entirely good reasons.)
Every range you'll see, including the ones in our own tool, is an industry average across a messy population of products. The KPI benchmark canvas the tool draws on says it in the first line: these are "directional references, not absolute targets."
Take K-factor (your viral coefficient: how many new users each user brings before their influence runs out). For B2B SaaS the typical range is 0.1 to 0.3; anything above 0.3 is unusually strong unless the product has built-in collaboration loops (Reforge, Andrew Chen). For consumer apps it's 0.3 to 0.7, and above 1.0 is extremely rare. So a B2B founder who reads a consumer growth blog, sees "aim for K above 0.5," and sets that as the target has just committed the team to a number that's abnormal for their category. The benchmark would have told them that, if they'd read it as context instead of a goal.
The metric that looks great in isolation is the one most likely to mislead you. The classic: LTV:CAC, the ratio of customer lifetime value to acquisition cost. The healthy range is 3:1 to 5:1; below 2:1 is structurally risky, and above 5:1 often means you're underinvesting in growth (Bessemer State of the Cloud, OpenView SaaS Benchmarks, a16z). Read that again: a "great" 7:1 ratio can be a warning sign, not a trophy. It usually means there's cheap demand you're not buying.
You only catch that by reading LTV:CAC alongside its sibling metric, CAC payback period (how many months of margin it takes to earn back acquisition cost). For SMB/self-serve that's 6 to 12 months; for enterprise sales, 12 to 24 months (OpenView, KeyBanc SaaS Survey). A 5:1 ratio with a 20-month payback in a self-serve business is a different animal than a 5:1 ratio with a 7-month payback. The benchmark for either number, alone, can't see that. The context can: both numbers, read together against your business model. (We pull this thread further in LTV:CAC looks great but you're not growing.)
The part most guides skip: an industry average is, by construction, the middle of a distribution that includes struggling products. Aiming for the median is, quite literally, aiming to be average, which is a strange thing to put on a roadmap. Steering toward the average can mean steering down from where a strong product already sits, or steering up toward a number your customers' actual usage cadence will never support.
Retention makes this concrete. B2B SaaS Day-7 retention typically runs 40% to 60% (Pendo, Userpilot); a consumer app's Day-7 is 8% to 15%, with many apps below 10% (AppsFlyer, Amplitude). A consumer team that imports the B2B "40% to 60%" as a target has set themselves up to feel like they're failing at 12%, when 12% is genuinely good for their category. The benchmark, read as context, protects you from that. Read as a target, it manufactures false failure.
Here's the same data the tool uses, laid out as context. Notice the Comment column; that's where the "how to read it" lives. A range without its caveat is exactly the half-truth that gets turned into a bad target.
| Metric | Benchmark | How to read it (context) | Sources |
|---|---|---|---|
| LTV:CAC | 3:1 to 5:1 | ~3:1 is the minimum healthy baseline. >5:1 often signals under-investment in growth; <2:1 is structurally risky. | Bessemer, OpenView, a16z |
| CAC payback | 6 to 12 mo (SMB/self-serve) 12 to 24 mo (enterprise) |
<12 mo is strong. Longer is fine only with very high retention and expansion. | OpenView, KeyBanc |
| K-factor | 0.1 to 0.3 | B2B virality is usually weak. >0.3 is unusually strong unless you have collaboration loops. | Reforge, Andrew Chen |
| Day-1 retention | 50 to 70% | Activated users usually return once more. Lower hints at onboarding or value-clarity problems. | Pendo, Amplitude |
| Day-7 retention | 40 to 60% | Early habit / workflow relevance forming. | Pendo, Userpilot |
| Day-14 retention | 35 to 55% | The curve should be flattening by now; a steep drop signals a weak core loop. | Amplitude, Mixpanel |
| Day-90 retention | 15 to 25% | Long-term value signal; Pendo keeps ~39% of users at month one, so the curve is well below the early-day bands by month three. <10% is a red flag for B2B. | Pendo |
| Metric | Benchmark | How to read it (context) | Sources |
|---|---|---|---|
| LTV:CAC | 2:1 to 4:1 (ideal ≈3:1) | Consumer growth is faster but less durable. <2:1 burns cash; >4:1 often means scale is constrained. | Adjust, AppsFlyer, a16z |
| CAC payback | 1 to 6 mo | Consumer products are expected to recoup CAC quickly. >6 mo usually fails at scale. | AppsFlyer, Mobile Dev Memo |
| K-factor | 0.3 to 0.7 (rarely >1) | >1 is true viral growth and extremely rare. Even 0.5 is very strong. | Andrew Chen, Reforge |
| Day-1 retention | 20 to 30% | <20% = weak first impression; >30% = top quartile. | Adjust, Statista |
| Day-7 retention | 8 to 15% | Many apps fall below 10%; >15% is excellent. | AppsFlyer, Amplitude |
| Day-14 retention | 4 to 8% | A steep drop is normal; flattening matters more than the absolute number. | Mixpanel, Amplitude |
| Day-90 retention | 1 to 4% | Above 5% is exceptional for consumer. | AppsFlyer, Adjust |
Activation rate is the one place to be extra careful: the tool ships activation bands for onboarding completion and activation, but those numbers don't yet have per-number sourcing the way retention and unit economics do. Treat activation benchmarks as the loosest context of the lot: useful for orientation, not for any kind of commitment.
Two rules that come straight off the bottom of the canvas: don't compare B2B and consumer directly (they follow different growth mechanics) and look for curve flattening, not just absolute numbers. A retention number that's "below benchmark" but has clearly stopped falling is often healthier than one sitting inside the band and still dropping. (More on reading the shape of the curve in retention rate benchmarks and B2B vs consumer growth.)
Three jobs, in order:
1. Sanity-check, not scoreboard. First question is always "is this number normal for a product like mine?" If your B2B Day-1 retention is 22%, the 50% to 70% band isn't telling you to "hit 50%." It's telling you something is wrong upstream, probably onboarding or value clarity, worth investigating before you touch any growth lever.
2. Interpret pairs, not points. Never read LTV:CAC without payback period. Never read K-factor without cycle time: a K of 0.6 with a two-week loop beats a K of 0.8 with a 60-day loop, because growth rate is K divided by cycle time, not K alone. (We work that math in full in the K-factor article.) The context lives in the relationships between metrics, never in one cell of a table.
3. Set your own target from your own model. This is the step benchmarks can't do for you. And honestly? It's the step most teams skip, because pulling a number off an industry table feels like rigor and building your own model feels like work. Your target for Day-1 retention should come from what your activated users need to do to get value and how often their job recurs. Then you use the benchmark to check whether that target is wildly out of line with reality. A cohort growth model is the honest way to derive targets, because it forces you to connect retention, new-user inflow, and viral compounding into one forward projection instead of chasing isolated numbers. (That's exactly what the cohort growth model does.)
Notice the order: the benchmark bookends the process, sanity-checking at the start and reality-checking at the end. It never sits in the middle as the goal. The goal is yours.
The hard part isn't finding benchmark numbers; it's reading them in context without quietly turning them into targets. That's the whole reason we built the free benchmark tool: it charts your retention curve, K-factor, and unit economics against the B2B and consumer bands side by side, keeps each number next to its caveat, and shows you the shape of your curve rather than a single pass/fail verdict. A couple of minutes, no signup gymnastics.
Check your numbers against the benchmarks → benchmark.scilla.studio
Use it the way it's meant to be used: to understand where you stand, then go set your own targets.
Are industry benchmarks useless, then? No, they're essential context. They tell you whether a number is normal for your category and flag when something is off. They just don't tell you what to aim for. Use them to interpret your metrics, not to set your goals.
What's the difference between a benchmark and a target? A benchmark is a directional reference: the typical range comparable products see. A target is a number you commit to, derived from your own product goals and customer reality. Benchmarks orient you; targets bind you. Don't confuse the two.
Why can a "good" LTV:CAC ratio be a bad sign? The healthy range is 3:1 to 5:1. Above 5:1 often means you're underinvesting in growth: there's demand you could profitably buy but aren't. A 7:1 ratio can signal a missed growth opportunity, which is why you read it alongside CAC payback period rather than alone.
Should a consumer app aim for B2B retention numbers? No. B2B Day-7 retention runs 40% to 60%; consumer Day-7 is 8% to 15%, with many apps below 10%. They follow different growth mechanics, so importing a B2B target into a consumer product manufactures false failure. Compare yourself only to your own category.
How do I set a real target if not from the benchmark? Derive it from your own model: what your activated users must do to get value, how often that job recurs, and how those numbers compound in a cohort growth simulation. Then use the benchmark to check the target isn't wildly unrealistic. The target comes from your product; the benchmark just reality-checks it.
Sources: Bessemer State of the Cloud, OpenView SaaS Benchmarks, a16z, KeyBanc SaaS Survey, Reforge, Andrew Chen, Pendo Product Benchmarks, Amplitude, Mixpanel, Userpilot, Adjust, AppsFlyer, Statista, Mobile Dev Memo. Benchmark ranges consolidated in the tool's own KPI canvas and used throughout the benchmark tool.
