Lean Startup Metrics: Glossary for Founders

Start building anything new — an app, a side project, a small business — and the same words start showing up on every call: retention, LTV, MVP, DAU, burn rate, PMF. Other founders use them. Articles assume you already know them. Nobody stops to explain, because asking feels riskier than nodding along.

This is the cheat sheet I wish someone had handed me. Plain definitions, opinionated notes on what actually matters at each stage, and a section on what changes between mobile apps and web apps — because most glossaries online ignore that the two play by very different rules.

Benchmarks below come from real industry reports — AppsFlyer, Mixpanel, ChartMogul, David Skok, and a few founder-canon essays — all retrieved May 2026. They shift year to year, so treat them as anchors, not absolutes.

How to read this glossary

Most metrics fall into one of six buckets, and they line up with the lifecycle of a startup. Read it in order the first time. The terms build on each other.

1. Building & funding — MVP, PMF, runway, burn, bootstrapped
2. Acquisition — CAC, CPI, channels, virality
3. Activation — first value, aha moment
4. Engagement — DAU, MAU, stickiness
5. Retention & churn — D1/D7/D30, cohorts
6. Revenue & unit economics — LTV, ARPU, MRR, payback

Then two sections that founders ask about constantly:

• Mobile vs. web — what actually changes
• The three metrics that matter before everything else

Building & funding

These are the words you'll hear before anyone has used the product yet: how lean to launch, how to know it's working, and how long the cash holds out. Get the first three right and most of the glossary becomes optional reading until product-market fit shows up.

MVP (Minimum Viable Product)

The smallest version of your product that delivers your core value to a real user. Not a prototype. Not a wireframe. Not a slide deck. Something a stranger can actually use to do the thing your product is for. The point isn't to launch cheap; it's to learn whether the problem is real before you spend a year building.

An MVP that nobody hates is usually too big. If you're not slightly embarrassed by what you ship, you waited too long.

You'll also hear MLP — Minimum Lovable Product as a counterpoint, popularized in a 2013 essay by Aha! co-founder Brian de Haaff: "instead of doing the bare minimum, you should be thinking about what it will take for customers to love you, not tolerate you." Same shipped product, higher bar. Viable means it works. Lovable means the first ten users tell a friend. In a category with a comfortable incumbent — which is most of them now — MLP is the more useful framing pre-PMF, but only after MVP-level signal proves the problem is real.

Product-Market Fit (PMF)

The point where users come to you instead of you chasing them. In numbers: signups grow without ad spend, weekly users keep coming back, and people get angry the moment the product breaks. PMF isn't subtle — when you have it, you know.

The cleanest gut-check is the Sean Ellis test: ask users "How would you feel if you could no longer use this product?" If 40% or more answer "very disappointed," you have early PMF. Ellis derived the threshold from studying around 100 startups and seeing the 40% line consistently separate the ones that scaled. It's a heuristic, not a peer-reviewed law, but it has aged well.

Default alive or default dead

PMF tells you whether the product works. Default alive vs. default dead tells you whether the company will live long enough to find out. The framing is Paul Graham's October 2015 essay, which poses a single question: "Assuming their expenses remain constant and their revenue growth is what it has been over the last several months, do they make it to profitability on the money they have left?" Default alive means yes. Default dead means no, and the only thing keeping you upright is the next round.

Pivot

A directional change that keeps one foot planted. You change the customer, the problem, or the channel — but you keep the team and what you've learned. A pivot isn't "we threw it all out." That's a restart.

Bootstrapped

Funded from revenue, savings, or consulting income — not from investors. Bootstrapped founders trade speed for ownership. You move slower, but you never have to explain a metric to someone who isn't building the thing.

Runway

The number of months your company can operate before it runs out of cash.

Runway (months) = Cash in bank / Monthly net burn

If you have $300k and you're net-burning $50k/month, you have 6 months of runway. Y Combinator's seed fundraising guide recommends raising "enough to reach profitability or your next major milestone, usually 12–18 months out," which means you start the next round well before the runway light goes red. Three months of runway is a panic, not a plan.

Burn rate

How much cash leaves the business each month.

• Gross burn — total monthly outflows (salaries, hosting, tools, ads).
• Net burn — gross burn minus revenue. This is the number that drives runway.

A pre-revenue startup's gross and net burn are the same. The day they diverge is a milestone worth celebrating.

Pre-seed / Seed / Series A

Funding stages, in rough order of company maturity. Pre-seed is "we have a deck and a demo." Seed is "we have early users and signal." Series A is "we have evidence this is a real business." Round sizes vary wildly by geography and sector, so don't anchor too hard on numbers from a US TechCrunch article.

Acquisition

Acquisition is everything that gets a stranger to your front door. Two questions matter: what does each new user cost, and which channels actually pay back once you measure retention by source. Cheap users who leave a week later aren't growth — they're a tax.

CAC (Customer Acquisition Cost)

Total amount you spend to acquire one paying customer.

CAC = (Marketing spend + Sales spend) / New customers in same period

Be honest about what counts as spend: ad budgets, the salaries of the people running ads, content, agencies, free trials you're subsidizing. CAC that excludes salaries is a vanity number.

CPA / CPI (Cost Per Action / Cost Per Install)

Channel-level acquisition cost. CPA is generic — cost per signup, per lead, or per click-to-app-store. CPI is mobile-specific: cost per install on the App Store or Play Store. CPI is almost never CAC, because most installs never convert into paying customers.

Channel

The source category through which a user found you: organic search, paid social, content, referrals, App Store search, and so on. Track CAC, retention, and LTV per channel. Channels that look cheap on CAC alone often have terrible retention.

K-factor (viral coefficient)

How many new users each existing user brings in.

K = Invites sent per user × Conversion rate per invite

K above 1 means the product grows on its own. K below 1 means every cohort needs paid acquisition to refill it. True K above 1 is rare. Be skeptical when you see it claimed.

Organic vs. paid

Organic = users who arrive without you paying for the click (SEO, word of mouth, App Store browsing, press). Paid = you paid for the impression or click. Healthy startups have a mix. A 100% paid business is a treadmill.

Activation

Acquisition without activation is just a list of people who saw your homepage. Activation is the first time the product actually works for someone — the moment they get the thing they came for. Most founders skip defining it because pinning it down is hard, but skipping it means you can't tell whether your funnel is leaking or just empty.

Activation

The first moment a user gets real value from your product. Not "signed up." Not "opened the app." Something concrete: created their first project, sent the first message, got their first export. You have to define this for your product before you can measure it.

On a tool we shipped recently, activation only moved when we cut signup from four screens to one — defining the event was the easy part; finding the friction in front of it took weeks.

Aha moment

The action that, once a user takes it, makes them dramatically more likely to stick around. Slack is the classic example: in a First Round Review interview, Slack co-founder Stewart Butterfield said any team that exchanged 2,000 messages had "really tried" Slack — and 93% of those teams were still using it. Find your version by comparing users who stayed against users who left, and looking at what the ones who stayed did in their first session.

Engagement

Engagement metrics tell you how habitual the product is, not how big it is. An app with 10,000 monthly users and 40% stickiness is healthier than one with 100,000 monthly users and 3% stickiness — the second is a leaky bucket dressed up in big numbers. Pick definitions that survive contact with a skeptical investor.

Active user

A user who did something that matters in a given window. The thing that matters is yours to define. Opening the app is a weak definition; sending a message or completing a task is a strong one. Be honest. You will be tempted to pick the loosest definition.

DAU / WAU / MAU

Daily / Weekly / Monthly Active Users. Same definition, different windows. DAU answers "how many people used the product today?" MAU answers "how many people used the product in the last 30 days?"

Stickiness (DAU/MAU ratio)

What fraction of monthly users come back daily.

Stickiness = DAU / MAU

Fred Wilson, USV co-founder, in his 2011 30/10/10 essay, gives the cleanest published heuristic: of your registered base, ~30% should be monthly active, ~10% weekly active, and ~10% daily active. As a stickiness ratio that lands around 20–30% for solid consumer products, 50% or more for utility-grade apps like messaging, and under 10% for products that are once-a-week tools at best. Under 10% can be perfectly fine for a tax app. It's not fine for a social app.

Session length / Session count

How long a single visit is, and how many sessions a user has per period. Useful, but easy to misread. Long sessions can mean engagement or confusion — pair them with task completion to know which.

Retention & churn

Retention is the leading indicator of whether your product works. Churn is the same number read from the other end. Acquisition without retention is a leaky bucket — you can pour water in forever and never fill it. Per AppsFlyer's 2024–2025 retention benchmarks, an average mobile app loses about 75% of users by Day 1 and roughly 95% by Day 30, so the bar is lower than most founders assume.

Retention rate

The percentage of users who come back N days after their first visit. The single most important leading indicator of whether your product works. Concretely: if 100 people sign up on Monday and 25 are still around the next day, your D1 retention is 25%.

D1 / D7 / D30 retention

Cohorted retention checkpoints — the percentage of new users who return on day 1, day 7, and day 30 after first use. Approximate consumer mobile-app benchmarks, drawn from AppsFlyer's 2024–2025 benchmark series and Mixpanel's 2024 Product Benchmarks Report:

The shape of the curve matters more than any single number. A curve that flattens — even at a low number — means PMF. A curve that decays to zero means no PMF, no matter how high D1 looks.

Cohort

A group of users who share a starting point, usually their signup week or month. You analyze retention by cohort so you can tell whether the changes you ship actually improve retention, or whether you just got better users this week.

Churn

The opposite of retention, mostly used for paying customers. Monthly churn of 5% sounds small until you compound it: at 5% monthly, you lose roughly 46% of your customers in a year.

Annualized churn ≈ 1 − (1 − monthly churn)^12

Per ChartMogul's SaaS Benchmarks Report, enterprise SaaS aims for monthly logo churn around 1% or below, while SMB-focused SaaS often runs 3–7%. For consumer subscriptions, under 5% monthly is workable. ChartMogul also notes SMB cohorts churn roughly 8x faster than enterprise — which is why the same product can have wildly different churn numbers depending on who you sell to.

Revenue & unit economics

Unit economics answer one question: is this a business or a science project? The two numbers that drive every conversation are LTV and CAC, and the ratio between them is the single most important sentence in your investor update once you start spending on growth. Get them wrong by counting only ad budget, and you'll celebrate a number that's quietly negative.

LTV (Lifetime Value)

The total revenue you expect from one customer over the entire time they use your product. The rough formula for a subscription business:

LTV ≈ ARPU / Churn rate

If your average user pays $20/month and you churn 5% monthly, LTV is roughly $400. LTV is an estimate, not a fact. David Skok cautions that early-stage LTVs are unreliable because the retention denominator is too thin to trust. Treat them with appropriate skepticism.

LTV/CAC ratio

The single most important unit-economics metric for any business that pays to acquire customers. David Skok's "SaaS Metrics 2.0" is the canonical source for the rule of thumb, which has held up for over a decade:

• Under 1:1 — you lose money on every customer. Existential problem.
• 1:1 to 3:1 — you might be a business, but margin is thin.
• 3:1+ — healthy, scalable.
• Over 5:1 — either great or you're under-investing in growth.

ARPU (Average Revenue Per User)

Total revenue divided by total active users in a period. Watch the denominator: ARPU per paying user is very different from ARPU per all active users — and "ARPU" alone usually means the latter.

ARPDAU / ARPPU

ARPU broken out per daily active user (ARPDAU) or per paying user (ARPPU). These come from the mobile gaming world and matter a lot for ad-supported and freemium apps.

MRR / ARR

Monthly Recurring Revenue and Annual Recurring Revenue — the predictable, contracted portion of subscription revenue. Excludes one-time fees. The unit of measurement of every B2B SaaS conversation.

Conversion rate

The percentage of people who move from one step to the next in your funnel. Visitor → signup, signup → activated, free → paid, trial → paid. Always specify which conversion you mean.

Payback period

How many months until the revenue from a customer covers the cost of acquiring them.

Payback (months) = CAC / Monthly gross profit per customer

Sub-12-month payback is healthy for B2B SaaS. Consumer can be much shorter — or much longer if you're playing the LTV-tail game.

North Star Metric, AARRR, and vanity metrics

Once you're tracking the metrics above, the next question is which one to actually optimize for. Three frameworks shape that conversation: North Star picks the single number that captures user value, AARRR maps the funnel from acquisition through revenue, and vanity metrics are the trap — numbers that move but don't drive any decision.

North Star Metric

The single metric that best captures the value your product delivers to users. Amplitude's North Star Playbook, co-written by Sean Ellis and Amplitude's John Cutler, lists Airbnb's nights booked and Spotify's time spent listening as canonical examples. The point isn't to ignore everything else — it's to make sure the team has one number that, if it goes up, the business is genuinely better.

AARRR (Pirate metrics)

Dave McClure's framework for organizing the funnel: Acquisition, Activation, Retention, Referral, Revenue. Coined in his 2007 Startup Metrics for Pirates deck, it's still useful as shared vocabulary — especially when your team has engineers, designers, and growth folks who all use slightly different words for the same stages.

Vanity metrics

Numbers that go up and to the right but don't change any decision you would make. Total signups since launch. App downloads. Press mentions. Twitter followers. They feel good in a board deck and tell you almost nothing about whether the product is working. The opposite of a vanity metric is a metric tied to a cohort and a behavior — like D7 retention of users from organic search, last 4 weeks. That's a number you can actually act on.

Mobile vs. web — what actually changes

The metrics above apply to both, but how you measure them and the realistic benchmarks shift a lot. The biggest practical gap is re-engagement: web lets you email a churned user; mobile mostly doesn't.

Web lets you re-engage churned users; mobile mostly doesn't. That changes which metrics deserve obsession. On web, you can afford to lose a user this week and win them back with email next month. On mobile, an uninstall is usually permanent — so D1 and D7 retention matter disproportionately more.

The three metrics that matter before everything else

If you had to pick three signals for a brand-new app, track activation rate, D7 retention by cohort, and qualitative feedback from 10 user calls a week. Pre-seed and seed-stage products tend to drown in numbers; these three give you the whole story within 90 days. Skip the rest until they're stable.

1. Activation rate. Of the people who sign up, what percentage reach the first real value moment? If this is below ~25%, your top of funnel is leaking faster than your acquisition can refill it. Fix activation before you spend a dollar on ads.

2. D7 (or W1) retention by cohort. Do users come back a week later? Day 1 lies — most people open something the day they sign up. Day 7 tells you whether the product earned a habit.

3. Qualitative signal from 10 users a week. Not a metric. A practice. Talk to ten users every week. The reason every metric above is squishy is that numbers tell you what — users tell you why. You need both.

Once you have signal on these three, layer in CAC, LTV, and channel-level retention. Not before.

FAQ

What metrics should I track for a new startup app?

For a new app pre-PMF, track three things: activation rate (% of signups who reach first value), D7 retention by signup cohort, and qualitative feedback from 10 weekly user conversations. Add CAC and LTV once you start spending on acquisition. Skip vanity metrics like total signups or downloads.

What's a good retention rate for a new mobile app?

Realistic consumer-app benchmarks are roughly 20–30% D1, 10–15% D7, and 4–7% D30, per AppsFlyer's 2024–2025 retention benchmarks. The shape of the curve matters more than any single number — a curve that flattens means product-market fit, even if the absolute numbers are modest.

What's the difference between LTV and ARPU?

ARPU is average revenue per period (usually monthly). LTV is total revenue over a customer's lifetime. For subscription businesses, LTV ≈ ARPU ÷ churn rate. ARPU answers "how much do we make per user this month?" LTV answers "how much will we make from this user before they leave?"

Are MVP and prototype the same thing?

No. A prototype demonstrates that something could work — usually to internal stakeholders or designers. An MVP is a real product, used by real users in the wild, that delivers genuine value (however small) and lets you learn. A prototype is a sketch; an MVP ships.

What's a healthy LTV/CAC ratio?

3:1 or higher is the standard rule of thumb, codified in David Skok's "SaaS Metrics 2.0". Below 1:1, you're losing money on every customer. Between 1:1 and 3:1, the business is fragile. Above 5:1, you're either exceptional or under-investing in growth. Always measure the ratio with payback period — a great LTV/CAC with a 5-year payback can still kill a startup.

Bootstrapped vs. funded — which is better?

Neither is universally better. Bootstrapped means slower growth, full ownership, and no fundraising distraction. Funded means faster growth, dilution, and more stakeholders. The right choice depends on the market, the speed at which a winner emerges in your category, and your personal preference for control vs. velocity.

What's the difference between DAU and MAU?

DAU (Daily Active Users) is how many unique users used your product on a given day. MAU (Monthly Active Users) is the same metric over a 30-day window. The ratio DAU/MAU — called stickiness — tells you what fraction of monthly users come back daily. Fred Wilson's 30/10/10 heuristic is the cleanest published benchmark, with stickiness above 20% considered solid for consumer products.

Other field notes from the workshop are over on the blog. The about page has the rest of the story.