Ops Handbook for Pre-Seed Teams

A handbook that scales from 1 to 5 without rewriting. Use as-is for the first six months; revisit at headcount 6.

Why pre-seed teams need ops at all

Most pre-seed founders defer ops as "Series A problems." That works until your second hire, then it breaks. The cost of writing this stuff down at 3 people is ~2 hours per topic. The cost of fixing it after a misalignment incident at 6 people is ~2 weeks of trust-rebuilding plus a hire who'll quit.

The 10 SOPs below are the minimum. Each one prevents a specific class of failure I've seen kill momentum at pre-seed teams.

1. Communication norms

• Async by default. Writing forces clarity. Slack threads beat meetings for 80% of decisions.
• Channel discipline: Slack for ephemeral. Notion/Linear for durable. Email for external only.
• Response SLAs: Slack DM within 4 working hours. @-mention within 1 hour. Email within 24 hours. No expectation outside working hours unless tagged urgent.
• Working hours visible in calendar. No-meeting blocks honoured by everyone.

2. Decision rights (RAPID-lite)

For a 3-5 person team, full RAPID is overkill. Use a 3-tier model:

• Reversible (one-way reversible): anyone can ship. No review needed. Examples: copy tweaks, A/B tests, retry logic.
• Two-way doors: one written paragraph in the team channel before shipping. 24-hour async review window. Examples: a new endpoint, a pricing experiment, a feature flag.
• One-way doors: written proposal, 48-hour async review, founder signs. Examples: hires, contracts, infra migrations, any commitment over $5k.

Write the policy down. Everyone references the same doc.

3. Tools stack (opinionated, low-switching-cost)

• Linear for product, engineering, and design tasks.
• Notion for durable docs (specs, retros, decisions log).
• Slack for chat.
• Loom for async video walkthroughs.
• Stripe for revenue.
• Vercel + GitHub for code and deploys.
• 1Password for secrets (not Notion, not Slack).
• Cal.com or Calendly for booking.

Don't add tools for problems you don't have. The most common pre-seed ops failure is tool sprawl — three places to look for the same thing.

4. Recurring rituals

• Monday 15-min standup: each person says (a) what they're working on, (b) one blocker, (c) one ask. No status reading. Notes go in Notion.
• Wednesday customer pulse review: 30 min, founder-led. What did customers say this week? What are we changing?
• Friday 15-min retro: what worked, what didn't, one experiment to try next week.
• Monthly investor update on the same calendar day (e.g., 1st of the month). Sent same day every month builds trust.
• Monthly cap-table check: 5 minutes to confirm the cap table matches reality. Catches small errors before they compound.

5. Hiring pipeline

• Sourcing: founder-led for the first 5 hires. No recruiters. No hiring@ inbox.
• Application response: 48-hour SLA, even for "no." A canned-but-warm rejection beats silence every time.
• Process: 30-min intro → take-home or paid trial → onsite (4 hours, panel) → references → offer.
• Offer: written, with equity grant terms, vesting schedule, cliff, and a deadline (typically 7 days). No verbal offers.
• Onboarding doc: one Notion page per hire with first-week goals, who-to-meet, and where-to-find-everything.

6. Customer feedback loop

• Every founder-conversation goes into a Notion database within 24 hours.
• Tag by ICP segment, theme, and quote-worthiness.
• Review at the Wednesday pulse meeting. Pick one thing to ship in response.
• The customer who asked sees the ship within 30 days. Loop closed.

7. Incident response

• On-call rotation: even at 3 people, rotate weekly. Founder isn't always on-call.
• Alert routing: PagerDuty (or free alternative) → on-call's phone. No alerts to email.
• Incident channel: #incidents. Start a Loom or recording on Severity-1.
• Postmortem: blameless, written within 48 hours of resolution. Filed in Notion.

8. Spending and finance

• Single corporate card (Brex / Mercury / Ramp). All subscriptions on it.
• $500 single-purchase threshold: anything over needs founder approval.
• Monthly burn review: 15 min on the 1st. Cancel one thing.
• Receipts and invoices: route to accountant's email automatically. No manual filing.

9. Data and privacy

• Single source of truth per data type (customers, leads, financials). One place to look.
• Access by role: founders see everything; engineering sees code + infra; ops sees customers + finance.
• Quarterly access review: 15 min to confirm everyone has only what they need.
• PII handling: documented one-pager. Where customer data lives, who can access it, how it's deleted.

10. Departures (yes, even at 3 people)

• Exit conversation within 48 hours of notice. Founder runs it.
• Knowledge transfer doc required before last day.
• Access revocation checklist: 15 items. Don't trust memory.
• Goodbye Slack message from the leaver, on their last day. Keeps the door open for future.

What to skip

• A full employee handbook (write it at headcount 10).
• A career framework (write it at headcount 15).
• 1:1 templates (use a free Notion template until you outgrow it).
• A performance review process (use lightweight 360s once a year).

The handbook above gets you to seed without ops becoming the bottleneck. The next step is a real People function — usually at headcount 15-20.