Week 7 of 12 · Phase 2: Grow

Week 7: prioritise hard and hit this week’s growth number

35 min to read · the doing takes the week0/5Start the challenge to see your dates →
This week in short

Past the middle of the batch, office hours keep coming back to one question: did you hit your number, and if not, why not? This week you grade every task by its impact on your weekly goal, the way YC partner Adora Cheung teaches, cut one distraction, and aim for 5–7% growth.

Week 7 challenge · Grow

The one-number week

Set this week’s target at 5–7% above last week’s value of your one number. Work only on tasks you graded high impact, and cut one recurring distraction for the whole week.

You pass when
A Sunday update with last week’s value, this week’s target, this week’s actual, your graded task list showing what you finished, and the distraction you cut. You pass if you hit the target; if you missed, name the biggest blocker and the fix you will try next week.
Why YC founders do this
Paul Graham’s “Startup = Growth” (5–7% a week is good during YC), Adora Cheung’s YC Startup School talk “How to Prioritize Your Time” (2019), and the group office hours described on YC’s “What Happens at YC” page.
How to do it this week ↓

YC does not publish a day-by-day calendar. We rebuilt the structure from YC’s pages, partner talks and founders’ accounts; every claim is labelled documented or reported. Unofficial, not affiliated with Y Combinator.

Inside YC · around week 7

What happens inside YC this week

Documented by YC, its partners or Paul Graham

YC’s “What Happens at YC” page says partners hold group office hours every two weeks and one-on-one office hours as often as founders want, and that what founders discuss depends on their stage and where they are in the batch. Each group is split into sections of 6 to 10 companies.

Paul Graham described the loop in 2022: founders bring the problems they are worried about, partners help find the one that matters most, and founders try fixes “at a resolution of a week or less”, then report back. His summary: “Speed defines startups. Focus enables speed. YC improves focus.” In “Startup = Growth” (2012) he says YC measures growth per week, and that 5–7% a week is good.

Founders’ accounts match. Nathan Leung (Harbor, S26), writing six weeks in, says his subgroup is 9 healthcare companies, that group office hours run every two weeks with founders sharing what has and hasn’t worked, and that outside the weekly events “the YC program has largely been hands-off”. Katia Damer (Prolific, S19) writes that their first office hours with group partner Jared Friedman found growth rate to be the main bottleneck, and that he pushed for a goal “so ambitious that you could barely imagine it working, but is theoretically possible.”

In March 2025, YC CEO Garry Tan told CNBC that “the whole batch is growing 10% week on week” for Winter 2025, without saying which metric. YC’s public pages do not describe a formal mid-batch review, so we don’t claim one.

2 frameworks · explained here in full

This week’s frameworks

Real vs fake progress, graded by impact and complexity

Adora Cheung, YC partner (Startup School, 2019)

Cheung’s first test: real progress moves your main metric, usually revenue or active users. Fake progress feels productive but doesn’t: conferences, awards, networking events, polishing a metric that isn’t the main one. YC’s recap of the talk quotes her: “If it contributes to real startup progress, you should consider doing it, and if it doesn’t, then don’t.” Most real tasks turn out to be a version of talking to users or building the product.

Keep one running list of ideas, but don’t act on each one as it arrives; she says switching every time causes whiplash. Once a week, set a goal on your main metric. Then grade each task twice. Impact: high, medium or low, meaning how likely it is to help hit this week’s goal compared with the other tasks. Complexity: easy (under a day), medium (one to two days) or hard (many days).

Do high-impact easy tasks first, then high-impact medium ones. Avoid low-impact hard ones. Take on only as many as you can finish well. Her example is a SaaS founder whose weekly goal is five new paying customers: visiting 10 prospects they have been introduced to ranks at the top, and most programming tasks rank near the bottom.

To check yourself, she suggests writing down every hour of a past week with the impact you expected and the impact you got, and a weekly update with the goal, whether you hit it, the biggest blocker, and what you learned. Group similar work into long blocks (a version of Graham’s maker’s schedule), and decide quickly: “Time is often wasted in indecisiveness.”

  1. 1Write this week’s goal as a number on your one metric.
  2. 2Put every task idea into one list. Don’t start any of them yet.
  3. 3Grade each task’s impact on this week’s goal: high, medium or low.
  4. 4Grade each task’s complexity: easy (under a day), medium (one to two days), hard (longer).
  5. 5Order the list: high impact and easy, then high impact and medium. Cut low impact and hard.
  6. 6Commit to only what you can finish, and block your days (for example, mornings building, afternoons with users).
  7. 7On Sunday, compare the impact you expected with what happened, and write it in your update.
Illustrative: a graded list for a clinic booking tool
Goal: weekly active clinics 60 → 64 (+6.7%)

High impact, easy     call the 6 clinics that went quiet last week
High impact, easy     ask the 5 happiest clinics for one introduction each
High impact, medium   build the SMS reminder 4 clinics asked for
Medium impact, easy   fix the calendar sync bug (2 complaints)
Low impact, hard      redesign the dashboard            → cut
Low impact, medium    apply for a startup awards list   → cut (fake progress)

A weekly growth target, checked at the resolution of a week

Paul Graham, “Startup = Growth” (2012) and “What I’ve Learned from Users” (2022)

Graham writes that YC measures growth per week, partly because there is so little time before Demo Day and partly because early startups need frequent feedback. “A good growth rate during YC is 5-7% a week.” 10% is exceptional. 1% is “a sign you haven’t yet figured out what you’re doing.”

A weekly target turns “we’re making progress” into a yes or no answer every Sunday. A miss is information: it tells you which problem to bring to office hours. Graham’s point about office hours is that founders often arrive with a small problem and leave having found the bigger one behind it.

Founders don’t all agree that a week is the right unit. Katia Damer of Prolific (S19) wrote that their customers took months to convert, so weekly experiments were not meaningful and pushed short-term thinking. If your sales cycle is long, set the weekly target on a leading number you control, such as qualified demos or pilots started, and check revenue monthly. Our lesson on the essay is at /paul-graham/startup-equals-growth.

  1. 1Take last week’s actual value and multiply by 1.05 and 1.07. That is your target range.
  2. 2Write the target where your cofounder sees it before Monday noon.
  3. 3Midweek, check your pace. If you are behind, ask which high-impact task moves the number fastest.
  4. 4On Sunday, write hit or miss. If you missed, name the single biggest blocker.
  5. 5Take that blocker to your own office hours: a mentor or two other founders, with one specific question.
Illustrative: one week of the loop
Last week: 60 active clinics
Target: 63–64 (+5% to +7%)
Actual: 61 (+1.7%)
Biggest blocker: 6 clinics went quiet when their free month ended
Fix for next week: call each clinic before day 25 of the free month and agree a paid start date
1 talk · play here

This week’s talks

Adora Cheung - How to Prioritize Your Time
Y Combinator · 17 min

Cheung grades tasks live. Watch “Grading impact” and “Grading complexity”, then grade your own list.

5 assignments · saved on this device

Your week

Do these before next week. Each one says why it matters and what done looks like.

0 of 5 done

  • Graham calls 5–7% a week a good growth rate during YC.

    Done when: Target written down before Monday noon, with last week’s value next to it.

  • Cheung’s grading turns a long to-do list into a short, ordered one.

    Done when: A list where every task has an impact grade and a complexity grade, and a line under the ones you will do.

  • Cheung’s fake progress, and YC’s pocket guide: avoid conferences unless they are the best way to get customers.

    Done when: The distraction named in your update, with the hours it freed.

  • Cheung expects most founders to find more low-value hours than they thought.

    Done when: Three days of hourly notes, with the low-value hours marked.

  • YC runs group office hours every two weeks, and founders describe them as the place to share what has and hasn’t worked.

    Done when: The session held, and one piece of advice you will act on this week written down.

Office hours · run it yourself

Questions a YC partner would ask you now

Answer each in one or two sentences, out loud first, then in writing. If an answer is vague, that is this week’s problem to fix.

  1. 01
    Why a partner asks this

    Cheung puts this at the top of the weekly update, and Graham describes YC’s loop as trying fixes within a week and reporting back.

  2. 02
    Why a partner asks this

    Graham (2022): founders often come in with a small problem, and the conversation finds the bigger one behind it.

  3. 03
    Why a partner asks this

    Cheung’s real-versus-fake progress test.

  4. 04
    Why a partner asks this

    YC’s pocket guide: startups can only solve one problem well at any given time.

  5. 05
    Why a partner asks this

    Damer (Prolific, S19) reports group partner Jared Friedman pushing for a goal you can barely imagine working but is theoretically possible.

Weekly update · one number: target vs actual, growth %

Write this week’s update

Week 7: last week’s value, this week’s target and actual, growth %, the tasks you finished with their impact grades, the distraction you cut, your biggest blocker and next week’s target.

Your update, ready to send
Our company: week 7 update

Our number: ?
This week's focus: one number: target vs actual, growth %
What we did: -
What we learned: -
Biggest obstacle: -
Ask: -

Your number also goes into the growth tracker on the program page. Saved on this device.

Checkpoint before week 8

A target 5–7% above last week, a graded task list, one distraction cut, and an update showing target against actual with your biggest blocker. Graham calls 5–7% a week good in “Startup = Growth” (2012); hitting it once matters less than knowing why you did or didn’t.

Questions about week 7

What happens at YC group office hours?

YC says partners run them every two weeks, with groups split into sections of 6 to 10 companies. Founders’ accounts, such as Nathan Leung’s from Spring 2026, describe each company sharing what has and hasn’t worked and getting advice from partners and peers. YC does not publish a fixed agenda.

Is there a formal mid-batch review at YC?

YC’s public pages don’t describe one. They describe group office hours every two weeks, one-on-one office hours on request and weekly events. If a site claims a specific mid-batch review, ask for its source.

What if my growth is about 1% a week?

Graham says 1% is a sign you haven’t yet figured out what you’re doing. Go back to week 6: check whether your cohorts flatten and who would be very disappointed, and spend next week on the product and on those users rather than on new channels.

Sources for week 7