Startup = growth: Paul Graham’s weekly growth rate, and how to use it

Unofficial. Not affiliated with or endorsed by Paul Graham or Y Combinator. Every teaching links to the original essay or talk.

1. 1

   What exactly is a startup?

   Graham opens "Startup = Growth" (September 2012) with a distinction: not every newly founded company is a startup. His definition: "A startup is a company designed to grow fast." He adds that a startup does not have to work on technology or take venture funding.

   His comparison is a barbershop and a search engine. A barbershop is not designed to grow fast; a search engine is. That is why he says the only essential thing about a startup is growth, and everything else we associate with startups follows from it.

   Key idea

   If you are not trying to grow fast, you may be running a fine business, but not a startup in Graham’s sense.

2. 2

   What growth rate counts as good?

   During YC, Graham writes, growth is measured per week. His benchmarks: "A good growth rate during YC is 5-7% a week." If you can hit 10% a week you are doing exceptionally well. If you can only manage 1%, it is a sign you have not yet figured out what you are doing.

   What to measure: revenue is best. For startups that are not charging yet, the next best is active users. In a footnote he adds that these weekly targets are for startups that have launched or can launch soon, not, say, a company building a new database.

   YC partner Adora Cheung went further in "How to Set KPIs and Goals" (Startup School, 2019): pick one primary metric you would bet the company on, prefer revenue, define "active" precisely if you use active users, and keep three to five secondary metrics, including retention, so growth on top does not hide users leaving at the bottom.

   • 1% a week: you have not found it yet. Talk to users and change the product.
   • 5–7% a week: good, by YC’s standard.
   • 10% a week: exceptional.
   • Measure weekly, on revenue if you charge, or on precisely defined active users if not.
3. 3

   Why do small weekly differences matter so much?

   Because growth compounds. Graham’s table: 1% a week is 1.7x in a year, 5% is 12.6x, 7% is 33.7x and 10% is 142x. A company making $1,000 a month and growing 1% a week makes about $7,900 a month four years later. At 5% a week, he writes, it would be making $25 million a month.

   The same maths is why he tells founders in "Do Things that Don’t Scale" (2013) not to dismiss tiny numbers: with 100 users, 10 more next week is 10% growth, and at that rate you reach about 14,000 users in a year.

   Weekly growth  Multiple (1+g)^52  Revenue after 1 year  Doubles every
    1%             1.68x              ₹83,900 a month       ~70 weeks
    5%            12.64x              ₹6.32 lakh a month    ~14 weeks
    7%            33.73x              ₹16.9 lakh a month    ~10 weeks
   10%           142.04x              ₹71 lakh a month      ~7 weeks
   
   Check it yourself: 1.05^52 = 12.64, so ₹50,000 x 12.64 = ₹6,32,000.

4. 4

   How does a growth target make decisions for you?

   Graham’s practical advice: "pick a growth rate they think they can hit, and then just try to hit it every week." He says you can then use growth like a compass for almost every decision. Should you spend two days at a conference? Should you hire another programmer? Ask whether it helps you hit this week’s number.

   It also shrinks a bewildering job into a single problem. He goes further: optimising for growth can discover startup ideas, acting as evolutionary pressure on what you build. In a footnote he rules out trickery, such as buying users for more than their lifetime value, and says fooling investors that way would only hurt you.

   In "Startups in 13 Sentences" (2009) he adds the warning that comes with this: you make what you measure, so be careful what you measure. Graham suggests putting a chart of your numbers on the wall. Cheung goes further: draw the next 10 weeks of target growth, post it where you will see it every day, and ask every week what the biggest obstacle to hitting this week’s target is.

   1. Pick one number↓→
   2. Set a weekly target↓→
   3. Ask: what helps hit it?↓→
   4. Do that↓→
   5. Review every week
5. 5

   Where does growth actually come from?

   Graham describes the growth curve of a successful startup as an S: a period of slow or no growth while the startup works out what it is doing, then fast growth, then a slowdown as it matures. YC’s 2017 pocket guide, written up by Geoff Ralston, puts it as "Growth is the result of a great product not the precursor", and adds: don’t scale your team or product until you have built something people want.

   So when the number is flat, the fix is usually the product and the users, not ads. Go back to the earlier lessons: talk to users, do things that don’t scale, and make a few people love what you make.

   Key idea

   A weekly growth number tells you whether you have made something people want. It does not make it for you.

6. 6

   What is ramen profitable, and why aim for it?

   In "Ramen Profitable" (July 2009), Graham defines it simply: a startup makes just enough to pay the founders’ living expenses. In "Startups in 13 Sentences" he says that once you cross into ramen profitable, it completely changes your relationship with investors.

   His reasons: investors take advantage of founders who obviously need money, so not needing it gets you better terms. It shows you can get paid and keep costs down. It lifts morale, because survival becomes the default. And you can raise money when you choose instead of stopping work to do it.

   His warnings: the biggest danger is that it turns you into a consulting firm, since, as he notes, it is pretty easy to make $3,000 a month consulting. And ramen profitable "is not the destination"; growth is. It just keeps you alive long enough to find it.

   • Add up your team’s real monthly living costs. That is your ramen number.
   • Track monthly revenue against it every week.
   • Charge early. Paying users also tell you more than free ones.

In his own words

Straight from the source

What Paul Graham or YC actually wrote or said, taken apart, then what to do with it this week.

Paul Graham

Cofounder of Y Combinator; YC partners later repeated this essay’s weekly benchmarks in Startup School talks.

Startup = Growth (essay, September 2012)

1 · Understand

A startup is a company designed to grow fast, so growth is the one essential thing. Measure it weekly, aim for a rate, and let that rate guide your decisions.

2 · Break it down

• Benchmarks during YC: 5–7% a week is good, 10% exceptional, 1% a sign you have not figured it out.
• Measure revenue if you can; active users if you are not charging yet.
• Compounding turns small weekly differences into huge yearly ones: 5% a week is 12.6x a year.
• Tricks like buying users at a loss do not count, and hurt you in the end.

3 · Use it

Pick a weekly target you think you can hit, write it where you see it, and test every plan this week against it.

Adora Cheung

Cofounder of Homejoy and a Y Combinator partner.

How to Set KPIs and Goals (Startup School talk, 2019)

1 · Understand

Choose one primary metric you would bet the company on, usually revenue, set weekly growth goals for it, and review it every week with a few supporting metrics.

2 · Break it down

• Revenue is the best primary metric for most startups; she urges founders to get paid.
• If you use active users, define "active" precisely. A signup is not a user.
• Keep three to five secondary metrics, and watch retention so new users do not hide lost ones.
• Each week, ask what the biggest obstacle to hitting the target is, and work on that.

3 · Use it

Write down your primary metric and its exact definition today, then put it next to your weekly target.

5 terms

Words to know

Startup

In Paul Graham’s definition, a company designed to grow fast. Growth is the one essential thing.

Weekly growth rate

How much your key number grows each week. YC’s benchmark from Graham: 5–7% is good, 10% exceptional, 1% means you have not figured it out.

Primary metric

The one number you track growth on. Graham says revenue is best and active users next best if you are not charging yet.

Compounding

Growth on top of growth. 5% a week is not 260% a year but about 12.6x, because each week builds on the last.

Ramen profitable

Making just enough to cover the founders’ living expenses, so the company can survive without new investment.

1 video · plays here

Watch

Hand-picked from the people who explain this best. The video loads only when you press play.

Do this week

Put it into practice

Set up your weekly growth number in 20 minutes

1. 1Choose one number: revenue if you charge anything, otherwise active users with a strict definition (for example, used the core feature at least twice this week).
2. 2Make a sheet with three columns: week, number, growth % (this week divided by last week, minus 1).
3. 3Fill in every past week you can reconstruct, even if the numbers are tiny.
4. 4Pick a weekly target you think you can hit, as Graham advises. If you have no idea, start with 5%.
5. 5List what you plan to do this week. Next to each item, write how it helps hit the target. Cut anything that doesn’t.
6. 6Add your ramen number (founders’ monthly living costs) as a line on the same sheet.

Read Paul Graham’s "Startup = Growth"

3 questions · instant feedback

Check yourself

Pick an answer to see if you are right, and why.

Question 1

Question 1 of 3

Your paid users went from 200 to 202 this week. By Graham’s benchmarks, what does that tell you?

AYou are doing exceptionally wellBAbout 1% a week: a sign you have not yet figured out what you are doingCExactly YC’s good rangeDNothing, weekly numbers do not matter

Question 2

Question 2 of 3

You charge nothing yet. You can track signups, page views or users who used the core feature this week. Which fits Graham and YC best?

ATotal signups since launchBPage viewsCActive users, defined precisely, such as people who used the core feature this weekDSocial media followers

Question 3

Question 3 of 3

You are ramen profitable doing custom projects for three clients and are tempted to take on more clients. What danger does Graham name?

ANone, profit is the goalBTurning into a consulting firm, which does not grow like a product companyCInvestors will refuse to meet youDYou will grow too fast

Lesson 6 of 17

Done? Keep going

Mark this lesson complete to keep your place. Progress is saved on this device, no login.

Mark lesson complete

Up next · Lesson 7 · 14 minWhat makes a great founderPaul Graham’s answer has stayed the same for over fifteen years: past a basic level of intelligence, determination matters most. Add flexibility, imagination, a little naughtiness, earnestness and a cofounder you are close to, and you get the founder he calls "relentlessly resourceful". Most of it can be trained, and YC judges it from stories of what you actually did, not from how confident you sound.← Previous: Launch early and talk to users

People also ask

Questions people ask

What is a good weekly growth rate for a startup?+

Paul Graham wrote in 2012 that a good growth rate during YC is 5–7% a week, 10% a week is exceptional, and 1% a week suggests you have not yet figured out what you are doing.

What should a startup measure growth on?+

Revenue, according to Graham. If you are not charging yet, use active users, and define "active" precisely. YC partner Adora Cheung also recommends watching a few secondary metrics such as retention.

What does ramen profitable mean?+

It is Paul Graham’s term (2009) for a startup that makes just enough to pay the founders’ living expenses. It means you no longer depend on raising money to survive, which improves your terms with investors.

6 sources · checked Nov 2026

Sources and further reading

• Paul Graham: Startup = Growth (September 2012)
• Paul Graham: Ramen Profitable (July 2009)
• Paul Graham: Startups in 13 Sentences (February 2009)
• Paul Graham: Do Things that Don’t Scale (July 2013)
• Geoff Ralston, YC’s Essential Startup Advice (YC blog, September 2017)
• Adora Cheung: How to Set KPIs and Goals (Y Combinator, 2019)

Related on this site

Launch early and talk to usersDo things that don’t scaleDefault alive or default deadHow to raise moneyFounder’s office guide