✍️ Essay
✓ Link checked
Free
Beginner
Why we picked it
This is the essay that forces the honest question underneath your idea: are you building a growth company or a good small business, because they are different DNA and require different lives. Graham is blunt that a barbershop is not a startup no matter how new it is, and that clarity helps you choose on purpose instead of drifting. There is nothing wrong with either path, but you should pick the one you actually want before you spend years on it.
From
Paul Graham
by Paul Graham
~20 min read
- A startup is defined by fast growth, not by being new or funded, so a business that cannot grow fast is a different (and often fine) choice, just not a startup.
- Growth needs two things at once: something many people want, and a way to reach them at scale, if either is missing the idea caps out as a niche.
- Deciding whether your idea can grow beyond a niche is really deciding what kind of company, and what kind of years, you are signing up for.
Open
paulgraham.com →
📄 Article
✓ Link checked
Free
Beginner
Why we picked it
Elman helped grow Twitter, LinkedIn, and Facebook, and his rule is exactly your short answer in practice. Decide the core action, decide how often a real user should do it, then measure how many people hit that bar. His Twitter example (seven visits in a month predicts they stay) is the model for naming a metric and a threshold before you ship.
From
Medium
by Josh Elman
- Ask three things: are people using it, doing the core action, at the frequency you expected
- A good metric predicts whether someone keeps coming back, not just that they showed up once
- Total signups and pageviews tell you almost nothing about whether it works
Open
joshelman.medium.com →
📄 Article
✓ Link checked
Freemium
Intermediate
Why we picked it
Your short answer tells founders to set a rough threshold before shipping, and this gives you real benchmarks so your threshold is not a guess. It collects activation numbers across many products so you know whether your result is good, bad, or average. Use it to decide honestly, a week later, whether the feature earned its place.
From
Lenny's Newsletter
by Lenny Rachitsky
- A threshold is only useful if you compare it against real benchmarks
- Activation rates vary widely by product type, so match yourself to similar businesses
- Set the number before launch so you cannot rationalize a bad result afterward
Open
lennysnewsletter.com →
📄 Article
✓ Link checked
Freemium
Intermediate
Why we picked it
Retention on the exact flow you shipped is one of the signals your short answer names, and this is the hands-on guide to measuring it. It explains cohorts, retention curves, and the difference between a curve that flattens (people keep using it) and one that decays to zero. This is how you tell a launch from a working feature.
From
Lenny's Newsletter
by Olga Berezovsky
- Group users by when they started, then track how many come back each period
- A curve that flattens means real retention, a curve that hits zero means you have nothing
- Look at the flow the feature touches, not just overall app retention
Open
lennysnewsletter.com →
✍️ Essay
✓ Link checked
Free
Intermediate
Why we picked it
Chen frames the exact trap you are in: pouring signups into a bucket that leaks, where more acquisition just wastes more water. He lays out scenarios that separate an onboarding leak from a deeper product leak, which is the fork in your question. It is the reason 'fix activation before you touch acquisition' is the right order of operations.
From
Andrew Chen
by Andrew Chen
10 min read
- Acquisition is pointless until the bucket stops leaking
- You cannot A/B test or notify your way out of bad retention
- Diagnose whether the leak is early activation or ongoing value
Open
andrewchen.com →
📖 Book
✓ Link checked
Paid
Beginner
Why we picked it
Ries gives you a working method for reading weak signals instead of guessing: build a small test, measure how real people respond, and learn fast enough to change course before a year is gone. Its most useful idea for this question is validated learning and the pivot-or-persevere call, a concrete way to decide whether an idea is worth continuing. Read it as a discipline for catching a dead-end early, not as a growth-hacking manual.
From
theleanstartup.com
by Eric Ries
~330 pages
- Validated learning means progress is measured by what you have actually confirmed with customers, not by how much you have built.
- The build-measure-learn loop is meant to shorten the time between an assumption and honest feedback on it.
- Pivot or persevere is a scheduled, evidence-based decision, so you are not drifting on an idea by default.
Open
theleanstartup.com →
📖 Book
✓ Link checked
Paid
Intermediate
Why we picked it
The book behind the "One Metric That Matters" discipline, which is exactly the mindset you need to judge whether a ship worked instead of drowning in vanity numbers. It ties the right metric to your business type and stage, so you know which signal to watch after a launch. Read it as the reference that makes the rest of this list make sense.
From
O'Reilly / Lean Series
by Alistair Croll and Benjamin Yoskovitz
440 pages
- Pick the One Metric That Matters for your stage and business model, and let it settle the "did it work" argument.
- Vanity metrics feel good but do not tell you a shipped thing is working: draw a line to real behaviour.
- Match the metric to where you are (empathy, stickiness, virality, revenue, scale), because the right signal changes as you grow.
Open
leananalyticsbook.com →
📄 Article
✓ Link checked
Free
Intermediate
Why we picked it
Turns the vague feeling of product-market fit into a number you can move. Ask users how they would feel if they could no longer use the product, then track the share who say 'very disappointed'. Under 40 percent means keep working. A test you can run on an idea long before you scale it.
From
First Round Review
by Rahul Vohra
~20 min read
- The 40 percent 'very disappointed' benchmark for product-market fit.
- Segment to your high-expectation customers and build for them.
- Make the fit score a metric you improve quarter by quarter.
Open
review.firstround.com →
✍️ Essay
✓ Link checked
Free
Beginner
Why we picked it
Cutler names the exact trap your short answer warns about: shipping feature after feature while nobody measures whether any of them worked. The twelve signs are a mirror you can hold up to your own week. Sign one is the whole point: teams that never measure impact and treat shipping itself as the win.
From
Cutle.fish
by John Cutler
- Not measuring the impact of what you ship is the top warning sign
- Success theater around launches replaces honest talk about impact
- Constant feature churn without learning is the default failure mode
Open
cutle.fish →
✍️ Essay
✓ Link checked
Free
Advanced
Why we picked it
Cagan draws the exact line at the heart of your question: output is what you shipped, outcome is the result it produced, and teams confuse the two constantly. He is honest that measuring outcomes is harder than counting features, which is why most teams quietly avoid it. Read it to commit to the harder, truer scoreboard.
From
Silicon Valley Product Group
by Marty Cagan
- Output is what you build, outcome is the difference it makes
- Feature factories measure shipping because outcomes are harder to face
- Sign your team up for a result, not a list of things delivered
Open
svpg.com →
✍️ Essay
✓ Link checked
Free
Advanced
Why we picked it
Tavel gives you a ladder for what working actually means: users completing the core action, coming back, and eventually pulling others in. It teaches you to define the one core action a feature should drive and measure repeat use of it, not surface adoption. A sharp framework for judging whether engagement is real or just curiosity.
From
Medium
by Sarah Tavel
- Define the single core action your product is built around
- Real engagement is core action, then retention, then self perpetuation
- Measure repeat use of the core action, not one time interest
Open
sarahtavel.medium.com →
✍️ Essay
✓ Link checked
Free
Advanced
Why we picked it
Winters, who led growth at Pinterest and Grubhub, reframes PMF as the point where users stop leaving, and shows how to read it off cohort retention curves rather than satisfaction scores. It is the most rigorous treatment of measuring fit as a trend over time, exactly the framing in your short answer. Best once you have some usage data to analyze.
From
Casey Accidental (Casey Winters)
by Casey Winters
~3,500 words
- PMF is when they stop leaving, measured by retention flattening
- Satisfaction surveys mislead, revealed retention behavior does not
- Use cohort analysis to watch fit as a trend, not a snapshot
Open
caseyaccidental.com →
✍️ Essay
✓ Link checked
Free
Intermediate
Why we picked it
If your worry is whether the product itself is the problem, the honest signal is the shape of your retention curve, and Balfour explains how to read it. A curve that keeps sliding to zero means no fit yet, a curve that flattens for some segment means you have found fit for that group. He frames fit as a progression through survey signal, engagement, and retention rather than a single yes or no, which keeps you from over reading one week of churn.
From
brianbalfour.com
by Brian Balfour
- A retention curve that flattens (levels off) for some segment is the clearest product side signal of fit, one that never flattens is not.
- Pair the curve with engagement data and qualitative survey signal, no single metric decides it.
- Fit is not a permanent verdict, markets move, so treat the diagnosis as ongoing.
Open
brianbalfour.com →
