Raising
When to raise
Raise when capital buys a milestone, with 6–9 months of runway left — not when you are merely anxious.
Fundraising is a sales process with a weird inventory: you sell equity (or SAFEs) in a company that does not yet have a liquid market. The hard parts start before the first intro: deciding when to raise and how much.
Paul Graham's durable advice still frames the work: either you are fundraising or you are building — do not drip "maybe we'll raise" for a year. Raise when capital buys a milestone that meaningfully changes the company's value.
The big idea
Runway first. Know cash, burn and months left. See burn and runway. A common pattern: start a process with 6–9 months of runway so you are negotiating from strength, not desperation. Raising with under 3 months left collapses power into the investor's hands and invites bridges and down rounds.
Concept
Runway is cash ÷ net monthly burn (roughly). If you have $1.8M in the bank and burn $150k/month, you have ~12 months. Fundraising itself consumes founder time and often increases burn (travel, counsel, diligence asks), so plan as if the process costs 1–2 months of focus.
Milestones, not vibes. Write the post-money state you need: for example "$2.5M ARR, 3 AEs ramped, NRR >110%" or "shipped v2, 40 design partners paying." The raise amount should cover 12–18 months to that state with a buffer for hiring slip.
How much. A practical seed/Series A estimate: headcount you will hire × fully loaded monthly cost × months of runway, plus non-people spend. PG's old rule of thumb — roughly $15k fully loaded per person per month × 18 months — is still a useful ceiling sketch (as of mid-2020s actual burn varies a lot by city and seniority; use your real numbers). Err slightly low in public messaging so progress looks like progress: saying "$2.5M" and later expanding beats saying "$5M" and stalling at $1.5M.
Soft criteria that repeatedly predict painful processes: no crisp one-liner, metrics that move sideways for two quarters, founders who disagree on the story in the room, or a cap table that cannot absorb a standard round without a cleanup. Fix those before you burn your A-tier intros.
Walking through a headcount budget
Plan: hire 6 engineers and 2 GTM over 18 months. Fully loaded cost average $18k/person/month.
People cost ≈ 8 × $18k × 18 = $2.59M
Add $400k infra/legal/misc → ~$3.0M raiseAnnounce "$2.5–3M" and treat $3M as the stretch if demand appears.
Instrument. Early: often post-money SAFEs or a small priced seed. Later: priced rounds with a term sheet. Do not invent custom securities without counsel.
Common mistakes
- Raising because anxiety peaked. Capital should buy a named milestone.
- Starting with under 3 months of cash. You will take whatever terms appear.
- Announcing $5M and stalling at $1.5M. The public number became a failed round.
Check yourself
Cash is $1.8M, burn is $150k/month. How much runway, and is that a good time to start a process?
About 12 months. Yes — 6–9 months is the usual start window; 12 months is strength.
Why announce $2.5–3M rather than $5M if the model says $3M?
A smaller public target that you beat looks like demand. A $5M target you miss looks like a failed round.
Key takeaways
- Raise to hit a milestone with 12–18 months of new runway; start before you are desperate.
- Runway ≈ cash ÷ burn; the process itself costs a month or two of focus.
- Size the round from headcount × loaded cost × months, plus a buffer.
- Fix the story and the cap table before you spend A-tier intros.
Further learning
Next, who you are raising from: target lists and intros.