How to Raise a Seed Round in 2026: The SF Founder's Playbook
Last updated: July 2026
Raising a seed round in 2026 looks nothing like it did in 2021. Rounds are smaller, expectations are higher, and AI has compressed what investors expect you to have built before you pitch. This is the plainspoken guide, written from what actually happens in SF right now.
What a seed round means in 2026
A seed round is the first "priced" or SAFE-based check meant to get you from a working prototype to real revenue. In 2026 the median SF seed is $2.5M-$4M at a $12M-$20M post-money cap. Pre-seed is $500k-$1.5M at a $6M-$10M cap. Series A now requires $1M+ ARR growing 3x, which is why seed rounds are being stretched to 24 months of runway instead of 18.
Founders often conflate pre-seed and seed. The practical difference: pre-seed buys you the right to build, seed buys you the right to grow. If you have no users and no revenue, you are raising pre-seed no matter what you call it.
See our SAFE notes guide for the mechanics of caps, discounts, and dilution math.
What you need before you start
Investors in 2026 want to see traction before writing seed checks. The bar, roughly:
1. A working product real users pay for or use daily.
2. A repeatable acquisition channel, even if small ($10k MRR growing 20% MoM is enough).
3. A cofounder or a technical lead. Solo non-technical founders raise, but at 2x the difficulty.
4. A clear "why now" that AI or a specific market shift makes possible.
5. Three months of runway to survive the raise itself.
If you are missing two of these, focus on building first. Fundraising when the metrics aren''t ready wastes six weeks and burns warm intros.
The raise timeline
A well-run seed raise in SF takes 6-10 weeks from first pitch to wire. Compressed rounds (2-3 weeks) happen when you have a lead already committed or heat from a demo day.
Weeks 1-2: prep. Deck, data room, target list of 40-60 investors, warm intros lined up. Practice the pitch on 5 friendly founders before any investor sees it.
Weeks 3-5: first meetings. Aim for 5-8 first meetings per week. Batch them: same day, same neighborhood. Investors move faster when they hear you have other meetings.
Weeks 6-8: partner meetings and diligence. The good funds bring you to a partner meeting within 2 weeks of first meeting. If a fund is dragging past 4 weeks, they''re not going to lead.
Weeks 9-10: term sheet, close, wire. Once you have a lead, the round often fills in a week. Give a "we''re closing Friday" deadline to the follow-on checks.
For the full breakdown of what each week looks like, see our SF fundraising timeline.
Who to pitch: the SF active seed list
The most active SF-based seed funds writing checks in July 2026:
- Multi-stage funds doing seed: Sequoia Arc, a16z Speedrun, General Catalyst, Founders Fund seed, Greylock seed.
- Dedicated seed/pre-seed: Hustle Fund, Pear VC, South Park Commons, Character VC, Founders Inc, Unpopular Ventures, Boost VC, Uncorrelated, Long Journey.
- AI-specific: Conviction, Radical Ventures, Air Street Capital, AIX Ventures.
- Solo capitalists / small funds: Elad Gil, Lachy Groom, Sarah Guo (Conviction), Nat Friedman + Daniel Gross (AI Grant), Basis Set.
Angels writing $25k-$100k are more numerous and often faster than funds. See our full filterable directory at /investors for check sizes, thesis, and portfolio.
The pitch deck that works in 2026
10-12 slides. No cover slide theater.
1. What you do (one sentence, then a screenshot).
2. Why now (the AI or market shift that made this possible in 2026 that wasn''t in 2023).
3. Traction (revenue chart, user growth, retention curve). Numbers only, no vanity metrics.
4. Product (2-3 screens, or a 30-second Loom link).
5. Market (specific, not "$400B TAM"). Bottom-up math wins.
6. Business model (price, unit economics, path to $10M ARR).
7. Team (why you three, why now, why not someone else).
8. Competition (honest, with your wedge).
9. Ask ($X for Y months to hit Z milestone).
10. Appendix (optional detailed metrics, hiring plan, roadmap).
The two slides that decide the meeting are traction and team. Every other slide is defense.
The first meeting: what to actually do
Investors decide in the first 5 minutes. Open with the traction number, not your origin story. If the number is small, frame the growth rate ("we went from $0 to $30k MRR in 8 weeks").
Bring the demo. Not a video of the demo, the actual product on your laptop. In 2026 investors are exhausted by decks and want to see the thing work.
Ask them at the end: "What would you need to see to be interested?" and "Who else should I be talking to?" These two questions do more work than the entire deck.
Common mistakes that kill rounds
- Raising too early. No revenue and no distinct insight equals no round in this market.
- Talking to associates instead of partners. Associates can only say no. Get to the partner meeting or move on.
- Optimizing for valuation. A $20M cap that closes beats a $30M cap that doesn''t. Dilution matters less than survival.
- Serial rejection blindness. After 15 no''s, stop pitching and rework the deck. The market is telling you something.
- No lead strategy. You need one fund willing to write 40%+ of the round and set terms. Chasing 30 small checks with no lead is death.
- Missing the "why now." In 2026, if your pitch could have worked in 2020, investors assume someone already tried it.
After the round closes
Wire hits. Do these five things in the first week:
1. Update your cap table and file 83(b) elections if any new equity was issued.
2. Send a founder update email to every investor with the round announcement and 30/60/90 day milestones.
3. Set a monthly investor update rhythm (short, honest, asks specific).
4. Extend runway conservatively. Assume your next raise takes 12 months, not 6.
5. Hire the one role that unblocks growth. Not five roles. One.
Related guides