How to raise your first round

I'm the CEO of Sauna. I raised Wordware's $30M seed, led by Spark Capital, in about seven days, and over $40M across my companies. This is everything I'd tell a founder raising from angels or going for their first VC round, in the order you'll need it.

Run an actual process

The biggest mistake I see is founders not running an actual process. They take meetings as they come, one at a time, so every investor sets their own pace.

Waiting is rational for an investor: every week they wait, they learn more about you for free. Your job is to give them a reason to decide this week.

Here's the timeline, counted by how much cash you have left in the bank:

When	What
6 months left	Build a list of 200
5 months left	Rewarm friends, line up intros
4 months left	Write the deck, wait for a trigger
Kickoff week	Release every intro at once
Weeks 2 to 6	Meetings, updates, commitments
Closed	Sign the SAFEs, back to building

Hard stop at four months. If it isn't closing, go back to building, keep the burn very low, and learn the skills yourself.

Before you start

Incorporate. If you haven't, use Clerky and take the full formation package (currently $819).
Size the round. Raise 18 months of runway plus a 20% buffer: about 12 months of focused building and 6 months to raise the next one.
Know what you'll give up. Most seed rounds sell 10 to 20% of the company. Try to stay under 25%.
Have the SAFE ready. When an angel says yes, they should be able to sign it the same day.
Clear your calendar. While the process runs, raising is your full-time job.

Why momentum decides everything

More investors talking to you is momentum. More of them saying yes is social proof. Social proof makes the next investor want in, because others already are. That loop only starts if you meet as many investors as possible in as short a time as possible, which is why everything below is about timing.

Build the list

6 months of cash left.

Aim for about 200 intros, which means tracking around 300 names. Most will pass, and that's fine.

A process like this usually runs 60 to 80 meetings, sometimes 100. An investor who doesn't do your stage or your space is a wasted meeting, so the list decides most of the outcome. A spreadsheet is enough:

Firm	Partner	Why them	Warm path	Stage	Next step
Fund A	The one partner you want	Led two seeds in your space this year	A founder they backed	Intro asked	Release in kickoff week
Fund B	Partner	Thesis post that reads like your pitch	Your old cofounder	1st meeting	Send update Friday

The stages: not contacted, intro asked, 1st meeting, partner meeting, diligence, offer, pass.

What makes an investor a fit

Stage. They write checks at your stage, today.
Check size. Their usual check fits your round.
Leads or follows. You need a handful who lead seeds, as well as followers.
Space. Recent deals in your category, or a thesis post that reads like your pitch.
Activity. They've invested in the last 12 months.
The partner. At a big firm, the one partner who backs companies like yours.

Where the names come from

Your friends, advisors and existing angels. Ask who they'd send you to.
The investors behind 10 to 20 companies one step ahead of you. They already believe in your space.
Curated lists by stage and sector (see the databases).
Portfolio pages of funds you like, to find the right partner.
X and LinkedIn, for what each partner is excited about right now.

Find the warm path to every name

An intro from one of their portfolio founders beats any other path. So grow your founder network well before you raise: one or two founders a week for nine months is more than 70 people who could refer you.

Run the finished list through a warm-path tool (Happenstance or Village, below) to see who you know who knows each investor, and aim for two or three intros per partner.

Sequence it in batches

Friends and family. Almost no work, and you start with money already in.
Angels from your industry. They signal that people who know the space are in.
Well-known angels. Lots of social proof and a big network, but harder to land, so bring them in once a few others have said yes.
VC firms. Slow, big checks, and the only ones who can lead. Your angels make them more likely to say yes.

For a VC-led round, the batches feed the horse race.

Let Sauna build the first draft

Paste this into Sauna:

Build me a list of 200 seed investors who've led or joined rounds in {your space} in the last 18 months. For each: the partner to target, one line on why they fit, and who in my network could intro me. Put it in a sheet with the columns Firm, Partner, Why them, Warm path, Stage, Next step.

The investor databases

Start with the free ones and add a paid tool only if they run dry. Skip the bulk cold-email databases for a seed round: an email from a list everyone has bought reads like spam.

Start here, free

Signal by NFX (free). Curated lists for 82 sectors and 24 regions, including Top AI Seed and Angels, Scouts & Solo-Capitalists. Filter by stage and sector and move the ones with recent deals into your tracker. Sign in with Gmail to see paths through your own network. The big lists include inactive investors, so check recent deals first.
VC Sheet (free). More than 100 hand-picked sheets covering 5,000+ investors. Start your lead list from Funds that lead seeds, then cross it with the sheet for your sector. Some sheets are old; check the update date.
OpenVC (free, $99/mo Premium). About 18,000 investors with check size, reply rate and how they want to be approached. Check every name before outreach. Thin on the top funds.
Visible Connect (free). A cross-check for names you found on Signal and VC Sheet. Profiles are thin.

Find the warm path

Happenstance (free, $24/mo Pro). AI search across your Gmail, calendar, LinkedIn, X and contacts. Search each firm or partner; the closest person goes in the Warm path column.
Village (free plan). Maps your whole team's email, calendar and LinkedIn. Useful when your cofounders and angels know different people.

Paid data

Crunchbase Pro (from $49/mo). Pick 10 to 20 companies one step ahead of you and export everyone who backed their seed. Those are your best-fit names. Round data lags announcements.
Foundersuite (free, from $745/yr). About 216,000 investors plus an investor CRM. It mixes in LPs and family offices, so filter hard.
Tracxn Lite (free, limited). Look up who backed the companies investors will compare you with.
Harmonic, Specter and PitchBook. The data investors use themselves. Priced on request, and too much for most seed rounds.

Prices and counts checked on 12 Oct 2026.

Rewarm your friends and line up the intros

5 months of cash left.

Rewarm your connections. Remember the person from school who's now well connected? Ask them for a chat. Ask for nothing yet.
Send your list to your friendlies, the people who can intro you, and ask who they know. Then hold the intros until kickoff week.
Line up two or three intros to every partner. One intro is easy to ignore; three from different people is a signal.
At a big firm, pick one partner and go after them.

No connections? Join an accelerator. Roughly best to worst: Y Combinator; then Techstars, 500 Global, Berkeley SkyDeck, PearX and Plug and Play; then Entrepreneur First, Antler and On Deck.

Cold email is a last resort, for when there's truly no warm path. Use the three-bullet format below: what you do, why it's exciting with a number, and what you want.

Silicon Valley etiquette

Everyone wears a hoodie, and the rules are still strict. Break one and nobody tells you. You just don't get the second meeting. This is my version of Romain Serman's Silicon Valley Etiquette.

Early is on time. Arrive three minutes early. Running late? Text 20 minutes before with your new time. More than ten minutes late, offer to reschedule.
Reply the same day. Next day is fine; anything later and you're slipping. Investors assume you'll run the company the way you run your inbox.
Intros are double opt-in. Ask your friend first. They ask the investor. Only when both say yes does the intro happen. Make it easy with a forwardable email.
Three bullets, five lines. What you do, in two lines. Why it's exciting, with a number, in two lines. What you want, in one.
Ask for 15 or 30 minutes. An hour is for second meetings. Propose the times yourself and end on time. The night before, confirm, send a short agenda and your phone number. Keep 30 minutes free afterwards to update the deck and tracker and send the follow-up.
Listen more than you pitch. Research what they've backed and ask about it. If the investor ends up talking more than you, your odds go up.
Never leave without a next step. The goal of a first meeting is the second meeting. Follow up the same day, even after a no, and keep the no's updated: some join when the round is closing.
Close the loop. Thank whoever introduced you the same day, then tell them how it went, even when it went nowhere.
Use numbers. Delete "world-class", "revolutionary" and "AI-powered". One real number does more than three superlatives.
Tell one story everywhere. Hiring, selling or raising, it's the same pitch, specific enough that someone can repeat it to their partners.
Never screw anyone over. Don't break a verbal agreement, don't invent a competing offer, and part on good terms after a no. Investors talk to each other.
Learn the translation. "This is awesome" means nice to meet you. "Keep us posted" means no, for now. "Come back when you have a lead" means no. "Let's get you in front of my partners" is real interest. The only feedback that counts is the next meeting on the calendar.

What never to do in front of an investor

Shade the truth in any way. Talk so much they can't get a word in. Follow up slowly, or break an agreement. Bring detailed financials to a seed meeting. Quote a market size you can't justify. Pretend to know something you don't. Ask for an NDA. Negotiate on the spot. Take a no personally.

Write the deck

4 months of cash left.

Investors read decks on a phone, in a few minutes, and for anything with AI in it they want to see the product working. This is Marcin Pogroszewski's method, rebuilt for that.

Close the slides app. Answer eight questions in one sentence each: the problem, how many people have it, how you solve it, why now (for AI, what changed in the models), why you, how many you've helped, how many more in 12 months, and how much money gets you there.
Make an AI argue with you until it runs out of objections.
Kill the red flags.
Make eight slides, one answer on each: Problem, Market, Solution, Why now, Team, Traction, Plans, Ask. On Solution, show a 60-second recording of the product working.
One slide, one thought. Add the single strongest data point to each.
Order it like an ad: surprise me, make me understand, make me more greedy than fearful, tell me the next step.
Make it pretty last. AI design tools do this in minutes; they can't do steps 1 to 6 for you.
Send it as a tracked link (DocSend or similar), so you see who opened it and where they stopped.

The prompt for step 2:

You're a seed partner who has seen a thousand AI decks this year and backed very few. Here are my eight one-sentence answers. Attack each one: what's vague, what's made up, what you'd ask in the meeting. Then tell me which answer would make you pass, and why.

Red flags

The problem feels made up.
The market math is sloppy, or quietly dishonest.
It's a vitamin (nice to have). You want an aspirin.
New paint on an old car: only a bit better than what people use, so nobody switches.
The experience you highlight has nothing to do with the problem.
Your users are friends and family. Test on strangers.
You're selling $100 for $90.
It's unclear how you reach your 12-month goal, or why that takes this much money.
For AI: the next model release makes you unnecessary.
For AI: your margin disappears once you pay for inference.

If you need more than eight slides

YC's longer version: company and tagline, vision, problem, customer, solution and why now, market, landscape and your insight, traction, business model, team, a three-to-five point summary, and the raise. Make it work as a leave-behind a partner can show their partners. Spend a day or two on version one, pitch five to ten friendlies, and update it after nearly every meeting.

On decks, from X

Nikita Bier: his polished deck got average interest. Handing investors the prototype to try worked far better.
Andrew Chen: get the two-minute pitch right first. Open with a surprising why now, and describe the team with hard numbers.
Garry Tan: if you can't tell the story to a friend at a bar, it isn't ready.
Garry Tan: keep pilots, bookings, revenue and recurring revenue apart on your traction slide.
Andrew Chen: why investors pass on AI startups.
Garry Tan: his answer to "you're just a wrapper" is evals built on hard-won knowledge of how your customers work.

Then let Sauna keep it current:

Every Monday until the round closes, update the traction slide with last week's numbers, and tell me who opened the deck so I know who to follow up with.

Wait for a trigger

Don't pull the trigger too early. Wait for something that shows you're taking off. Before Wordware's seed, a couple of million people used our Twitter personality app, and we had a great Product Hunt launch.

Stack a few of these at once: a launch, a viral moment and a big customer in the same month. You can engineer them.

Paul Graham on why raising before you're attractive enough is a mistake, and Karri Saarinen on raising from a position where you don't need the money.

Pick a strategy, then release every intro at once

Kickoff week.

Push all your intros out in the same week, then run one of three plays from Ted Wang's Horses, rabbits and poker. They're mostly for a VC round; if you're new to this, a party round of angels can be the better choice.

The horse race. Pick 6 to 8 firms (many more if you aren't hot). Release every intro the same week so they move at the same pace, and slow down whoever races ahead. Aim for two offers in the same week. One offer and five first meetings is no leverage.

The rabbit. Start one firm a week early: one that already likes you, and that you'd love to work with. When the next firm offers a first meeting in eight weeks, you can say, "I've got a partner meeting at another firm next week." Suddenly, Thursday is free. Only say it if it's true.

Heads up. Poker with two players left. Tell one VC you know really well, "Make me a great offer and I'll take it." They need to know there's no second bid. If they ask what a great offer is, don't answer. Never negotiate against yourself.

The plays blend: a soft heads up looks like a rabbit, and a rabbit turns into a horse race once the timing lines up. No one for the rabbit or heads up? Run the horse race.

The forwardable email

Week 1.

You know somebody who knows the investor. Ask them (a LinkedIn message is fine) if they'd do the intro. When they say yes, send an email they can forward without editing.

Subject: Intro request: {your company} to {investor}

Hey {investor},

I'm {name}, {one line that makes you credible}. {Friend} said you'd be a great investor for us.

{What just happened, with a number.}

{What you do in two sentences. One "think X for Y" line helps.}

Here's the deck: {tracked link}

Would you like to chat?

{Name}
{Title, company} · {one credential}

Here's the one I actually sent for HeyDaily, my second company:

I'm Filip and I am a second-time entrepreneur, {Friend} recommended you as a great investor :) I have been bootstrapping a project for the last couple of months and now the sales exploded and we have 68k MRR projected based on contracts signed with 21 influencers and a 5.1mm people reach.

Credible in one line, the numbers early, a "think X for Y" line, one link, one ask. Elad Gil and Trae Stephens have good threads on the same idea.

Close the angels and small VCs

Weeks 2 to 6.

If the call went well and they seem in, ask. It can happen on the first call, but it's usually the second or third. If they aren't in by the third conversation, it's probably a waste of time. You want the money, and a commitment you can show the next investor.

A closing conversation usually goes something like this:

Angel: Wow, this all sounds great. I have a couple of comments, but I really like it!

You: Awesome. What's your typical check size for a company like mine?

Angel: I put $350k into a company at this stage before, but I'd encourage you to raise as little as possible.

You: Great, so can I count you in for $350k in this round?

Angel: Well, that's not what I meant. You'd need to get the round together first. Plus a few other conditions.

You: OK, so if I meet those conditions, you're committed for $350k?

Now wait for the answer. It will feel uncomfortable, and they may invent more conditions.

Angel: Yeah, that sounds reasonable. What about valuation?

You: We're thinking $20M.

Angel: Oh no, that sounds way too high.

You: Is there a number where you'd come in straight away?

Angel: $12M is something I could work with.

You: OK, so I hear a commitment of $350k at a $12M valuation. Thank you for your faith in us. Mind if I share your commitment with the people already in?

Typical angel checks are more like $15k to $150k. Whatever the number, confirm it in writing the same day: they're in for that amount, the offer stands for 48 hours, and they'll fund within 10 business days. They reply yes, and you send the SAFE.

When someone offers a big check

Don't say yes in the room, or there may be no room left for the investor you really want. Ask what allocation they want and say you'll come back within two or three days. Email everyone you've met, including the ones who passed, that the round closes in 48 hours. Then decide allocations: ask your cofounders and current investors, and call founders they've backed.

After a yes, ask for help

Every committed investor is a recruiter. Need more meetings? Ask who else they'd introduce you to. Need to convince someone in your pipeline? Ask who they can call. Oversubscribed? Ask for their read on the other investors.

Keep the heat on

Create triggers. It's in every investor's interest to wait, so give people a reason and a date to decide by.
Get to a no. A pile of maybes is fake traction. Near the end, push every maybe to an answer. Many fall apart, and some join.
Send every small win to your whole pipeline. If I hire someone good or have one good call, I somehow turn it into big news they need to hear.
When it heats up, get founder friends to WhatsApp the right partners. More scarcity, more hype.
Roll the round. If VCs aren't throwing money at you, close SAFEs as you go, and move the cap up as you need the money less.

The update email:

Subject: Quick update 🚀

Hi {name}, a quick one since we spoke. {One win, with a number}. {A second win}. {Who's committed since: a fund, an angel}. We're moving fast and plan to close by {date}. Happy to send more if useful.

"Come back when you have a lead"

It's the politest no in venture. From an angel, "we need a lead first" is normal. From a VC, it's usually a pass.

The lead wants most of the round to hit their ownership target, and everyone else piles in behind them. So "come back with a lead" costs the VC nothing. If you show up with a name they respect, they're in. If you don't, they've passed without ever saying no. If a VC says it, go back to your list.

Documents and terms

Use a SAFE. Pre-money if you can get it; post-money is the standard. Start from YC's SAFE documents.
Skip pro-rata if you can. Aaron Harris explains why: Pro rata is a bad term.
Pick a cap. When I first wrote this, my answer for a first US raise was an $8M cap. Caps have moved since, so ask two or three founders who closed this quarter, pick one number, and raise it as demand grows.
Send an investor questionnaire with the SAFE, so you know each investor is allowed to invest in you.
Have a lawyer read everything. I can't take any responsibility.

On a post-money SAFE, the investor's share is their check divided by the cap: $1M at a $10M cap buys 10% of the company. Add up every SAFE in the round to see your total dilution before you sign the last one. Same terms for everyone in the same round.

The mistakes I see most

Not being decisive, and not running a proper process.
Thinking you'll present and people will throw money at you. Negotiation and scarcity do most of the work.
Hoping for a lead.
Raising for more than four months. Go back to building, keep the burn very low, and learn the skills yourself.

Let Sauna run the process

Sauna is the thing I devote my life to building. It's a proactive AI assistant for work, and it can run this whole process for you: the list, the intro asks and every update to your pipeline. You take the meetings and make the calls. Try it with this as your first message:

I'm raising a {seed} round of {amount} for {company}, closing by {date}. Read this guide, then run the process with me: build the list of 200, draft the intro asks for my friendlies, keep the tracker current, and send updates to the pipeline when I approve them.

Resources

A guide to seed fundraising, Geoff Ralston at YC. How seed rounds work, from timing to documents.
How to build your seed round pitch deck, YC. What each slide is for.
Fundraising kit, Shizune. A step-by-step guide, templates and a checklist.
Horses, rabbits and poker, Ted Wang. The original three strategies.
Silicon Valley etiquette, Romain Serman. The longer version, from 2016.
Lessons from raising $100M, Adit at Reducto. Capital efficiency as leverage.
Billion dollar pitch decks. Real decks from companies that made it.

Filip Kozera is the CEO of Sauna. Follow along on Instagram or LinkedIn. Not legal advice: have a lawyer read anything you sign.