I spent four years at a big law firm drafting the documents that decide who gets rich when a startup sells. Then I left, founded a company, and signed a term sheet I only half understood. That gap — between writing the paper for clients and signing it as a founder — is the reason this site exists. Because the first investor meeting you'll ever take is not a conversation. It's a negotiation where the other side has done it two hundred times and you've done it zero, and the score gets recorded in a document you'll live with for a decade.

Here's the part nobody tells you early enough: by the time a term sheet lands in your inbox, most of the outcome is already decided. The valuation gets the headline. The structure gets the money. And structure is set by everything you did — or didn't do — in the months before that meeting: how you incorporated, who owns the IP, what your SAFE actually says, whether your co-founder's stock vests. Investors know this. So they diligence your paperwork before they negotiate your price, and every loose end they find becomes leverage against you.

This page is the overview I wish someone had given me over coffee, before I learned it the expensive way. It's long. Read it anyway. Each section links to a deeper guide when you're ready.

The Term Sheet That Taught Me to Read

My company's first priced round came in at a valuation I was proud of. I remember the number because I repeated it to my parents. What I didn't repeat — because I didn't understand it — was the rest of the paragraph: a liquidation preference with participation, a cumulative dividend, and an anti-dilution clause I'd skimmed because the word "weighted" made it sound moderate. Eighteen months later we sold the company for a price that would have made the founders comfortable under a clean term sheet. Under mine, the preference stack ate first, the participation ate second, and common — my stock, my co-founder's stock, our early employees' options — split what was left. It wasn't fraud. It was arithmetic I'd agreed to in writing.

And that's the pattern, over and over. Founders negotiate the number on the front page. Investors negotiate the definitions in the back. A "standard" term sheet is standard the way a casino's house rules are standard — everyone's used to them, and they still favor the house. You don't need to become a lawyer. You do need to understand the five or six provisions that actually move money, because no one at that table is paid to explain them to you honestly. Our term sheet guide walks through every clause line by line, but let's cover the big ones here.

Paper the Company Before You Pitch It

Investors fund Delaware C-corporations. Not LLCs, not S-corps, not "we'll convert later." A Delaware filing starts at an $89 state fee for the certificate of incorporation, and the whole thing — registered agent, bylaws, initial board consents, stock purchase agreements — is a solved problem that costs a few hundred dollars and a week of attention. Do it before you take a dollar. Every week you operate as a handshake partnership is a week of revenue, code, and promises that have to be untangled later at lawyer rates.

When you get your founder stock, file your 83(b) election within 30 days. Not 31. Thirty. Miss it and the IRS treats your vesting shares as income as they vest, at whatever the company is worth by then. I've seen a founder face a six-figure tax bill on paper gains in a company that later died. The form is one page. Set a calendar reminder the day you sign your stock purchase agreement.

Vesting is the other one. Founders resist it — "it's my company" — and investors insist on it, and here's the thing: the investors are right. Standard is four-year vesting with a one-year cliff, and it protects you from the co-founder who quits in month three and keeps a third of the company forever. Put it on your own stock voluntarily, before anyone asks. It signals you understand the game, and it keeps a dead cap table from killing your next round.

Then assign the IP. Every founder, every early contractor, every advisor who touched the product signs a proprietary information and invention assignment agreement. The company must own the code — not the people who wrote it. This is the first thing diligence checks and the first thing that blows up deals, and it's cheap to fix early and brutal to fix late. The details are in our incorporation walkthrough and the IP assignment guide linked from it.

The SAFE Is Not the Safe Part

You'll probably raise your first real money on a SAFE, and you'll probably misread it. A SAFE is not a loan and it's not equity — it's a promise to convert into shares later, at terms set by the valuation cap and discount you negotiate today. Say you raise $500,000 on a post-money SAFE with an $8M valuation cap. That's 6.25% of your company, gone, the moment it converts — before the new investor's money dilutes anyone. Raise $1.5M on that same cap across three SAFEs and you've sold nearly 19% before your Series A even prices.

The trap is that SAFEs feel free. No board seat, no interest, no maturity date, close in an afternoon. So founders stack them — one in March, two in June, another in the fall — each with a slightly different cap, and nobody adds them up until the conversion math lands in a cap table model at the priced round. I've watched a founder realize, mid-negotiation, that his SAFEs plus the new option pool plus the Series A would leave him under 50% of his own company. He'd raised less than $2M total. Run the waterfall before you sign each one, not after. Our SAFE guide has the dilution math with a chart that shows exactly where the percentages go.

The Preference Stack Decides Who Eats

A 1x non-participating liquidation preference is the market standard, and it's fair: the investor gets their money back or converts to common and takes their percentage, whichever is larger. In a decent exit, they convert, everyone shares pro rata, fine. The trouble starts with the variants. Participating preferred takes the money back and the percentage — double-dipping, and in a middling exit it can cut the common payout by a third or more. Multiples above 1x do the same thing faster. Neither is "standard," whatever the email says.

Do the arithmetic on every offer. Take a company that raised $10M at a 1x participating preference and sells for $25M. The investor takes $10M off the top, then converts-ish — takes their 40% of what's left, another $6M. Common splits $9M on a $25M exit. Non-participating, the same investor converts and takes $10M, common splits $15M. Same sale price, same headline valuation, a $6M swing that lives entirely in one adjective.

Anti-dilution is the other silent mover. Full ratchet says: if you ever sell a share cheaper, my price resets to that price — as if the down round happened to my whole position. One desperate bridge at a low price can vaporize the founders' stake. Broad-based weighted average is the tolerable version: it adjusts the conversion price by a formula that accounts for how much was actually sold cheap, so a small down round costs everyone a little instead of costing you everything. If an investor pushes full ratchet, ask why their model assumes you'll miss. Then negotiate it down or walk.

Board composition deserves the same arithmetic. A two-seat common, one-seat investor board after the seed round sounds balanced until the "independent" fifth seat shows up — and the independent is the investor's former partner. Protective provisions sound protective until you read them: a veto over new debt, over budgets, over hiring executives, over selling the company at all. Each one is a lever. You'll never remove them all and you shouldn't try; investors deserve protection on the existential stuff. But every veto you hand out is a door you have to knock on later, hat in hand, at the exact moment you need to move fast. Count them before you sign.

The Quiet Clauses That Move Companies

The clauses nobody puts in the pitch deck are the ones that bite at 2 a.m. Right of first refusal means before you sell a single share to anyone, the company and then the investors get to take the deal instead. Co-sale rides along: if you find a buyer for your shares, the investor can stuff their shares into your sale and crowd you out. Together they make your founder stock nearly illiquid, which is precisely the point. Drag-along goes the other direction — if the board and the preferred approve a sale, you can be dragged into it, at a price the preference stack chose. Read the drag threshold before you assume a majority means a majority.

And the no-shop. You sign a term sheet with a 45-day exclusivity clause, and now you can't talk to anyone else while your runway burns. The investor knows your bank balance — it's in the diligence package. Every week of slow diligence is a week of leverage, and the term sheet you signed on the good week gets "retraded" on the bad one. Cap exclusivity at 30 days, keep a parallel conversation warm, and never let one document be your only option. That's not cynicism. It's Tuesday.

Common Stock Is Not a Participation Trophy

Everything above sits on top of one fact: common eats last. Founders and employees hold common. Investors hold preferred, and preferred is a ladder of claims standing between you and the exit check. Your employees' options are priced off the common value — the number that gets crushed first when preferences participate, when ratchets reset, when a recap "cleans up" the cap table. Protecting the common isn't sentimentality. It's protecting the only people who can't reprice their risk with a new fund next year.

So fight for the boring wins. Keep the preference at 1x non-participating. Keep the pool shuffle — the investor's trick of sizing the option pool pre-money so it dilutes you, not them — as small as your hiring plan actually requires, with the spreadsheet to prove it. Get the drag set at a threshold that includes common. And ask every "founder-friendly" fund to mark up your term sheet in writing. The friendly ones will. The adjective does the work the paper won't.

One more thing, because nobody believes it until it happens to them: build the diligence folder before you're asked. Charter, bylaws, every stock purchase agreement, every SAFE, every signed PIIA, your cap table as a spreadsheet with the grant dates. When a fund asks for it, sending it in twenty minutes instead of twenty days changes the whole tone of the process. It says your house is in order, and it quietly removes the excuse for the "we found some issues, so about that price" conversation. Our due diligence guide has the full checklist, and it's shorter than you fear.

What I'd Do Before Your Next Investor Coffee

One week, five moves. First, confirm you're a Delaware C-corp with the IP assigned — every founder, every contractor, signed and filed. Second, check your 83(b) receipts; if you can't produce them, neither can the IRS audit you'll eventually invite. Third, model your SAFEs converting at the cap, all of them stacked, and write your post-conversion percentage on an index card. Fourth, read your vesting schedules and make sure every founder is on one. Fifth, open the guides on this site and read the two that match your next ninety days — term sheets if you're raising priced, the SAFE guide if you're not.

None of this requires a law degree. It requires treating the legal structure of your company as part of the product — the part that decides who owns the product. The investors sitting across the table built their returns on founders who didn't. Don't be the return.