Travis Kalanick built Uber. By the time it went public, he reportedly owned less than nine percent of it. Two years earlier he had resigned as CEO after major investors demanded a leadership change.
What funding looks like in the news is not what it looks like inside the company. The headlines show the check size and the valuation. They leave out the dilution, the terms, the board, and the clock that starts ticking the day the money hits the bank.
Most founders end up with less ownership and less control than they expected, and they sign up for all of it without realizing what they signed until they are already living with the consequences. The valuation is the number that gets celebrated. It is also, at the early stages, the number that matters least. Here is how a funding round actually works, and where the real cost is hiding.
Key Takeaways:
Pick the company you are building first. Venture-scale and durable-profitable are both valid. Taking outlier-return money and then deciding to run a profitable business at your own pace is the expensive mistake.
Dilution is a percentage, not a share count. The company prints new shares for investors. You keep your shares and lose percentage every round.
Early valuations are agreed, not calculated. Competition sets the number. It is the part founders overweight the most.
The terms decide who controls the company. Liquidation preference, protective provisions, pro rata, and drag-along matter more than the headline when things go wrong.
A clock starts the day the money lands. You agreed to grow fast enough to raise the next round on the investors' timeline, not yours.
Two kinds of companies, one expensive mistake
Before any of the funding details matter, there is a choice almost every founder makes without realizing they are making it. You can build a venture-scale company, with large rounds, high valuations, and an eventual sale or IPO. Or you can build a durable, profitable company that makes money, pays the founders well, and grows steadily for years while the founders own most of it. Both are completely valid.
The mistake is wanting one while taking money built for the other. You cannot take a check from a fund built around outlier returns and then decide later that you would rather just run a profitable business at your own pace. By signing for that money, you agree to grow fast enough to sell or go public on a timeline the investors set. What changed recently is that building a profitable company without outside money got much easier. AI cut the cost of building software, and solo-founded companies went from about a quarter of new startups on Carta in 2019 to over a third today. Skipping the raise is more realistic now than it has ever been.
The rounds are stages, not a lump sum
Funding comes in stages, because no investor wants to fund years of a company they have not seen progress from yet. Preseed is the idea-and-prototype stage, and in 2026 it has split in two. Many rounds are still under two hundred fifty thousand dollars, while the larger ones average closer to one and a half million. Most are not priced rounds at all. They are SAFEs or convertible notes that let you take money now and set the valuation later, usually with a cap somewhere between ten and fifteen million.
Seed comes once you have a working product, early users, and signs people want it. Median seed valuations hit a record around twenty-four million in late 2025, and AI companies carry a premium of roughly thirty-eight percent on top of that at Series A and beyond. Series A used to follow a simple rule, one million in annual recurring revenue and you could raise. That number is not enough anymore. Most investors now want to see two to three million unless your growth is exceptional. Each stage has its own rules, and pitching a Series A meeting like it is preseed gets you a pass.
Dilution is a percentage, not a share count
When investors put money in, they do not buy your shares. The company creates brand-new shares and gives those to the investors. If you and a co-founder started with one hundred shares split fifty-fifty, those shares stay yours. The company prints twenty new shares for the investor on top of the hundred that already existed. You still hold fifty, but fifty out of one hundred twenty is forty-one percent. You went from half the company to forty-one percent without selling a single share. That is dilution, and it happens every round.
Facebook is the cleanest example. Peter Thiel was the first outside investor, and in 2004 he put in five hundred thousand dollars for roughly ten percent, valuing the company at just under five million. The founders did not sell anything. The company made new shares, and everyone else's percentage shrank to make room. By the IPO, every original founder owned a fraction of what they started with on paper, but the company was worth a hundred billion, so that fraction was worth tens of billions. That is the deal you are agreeing to: you give up percentage for the chance to build something much bigger than you could fund alone.
Early valuations are agreed, not calculated
Founders fixate on valuation more than almost anything else. It goes in the press release and gets compared with other founders. At the early stages it is also the number that matters least, because it is not calculated. There is no formula, because there is often no revenue and barely a product to plug into one. The valuation becomes a negotiation about how much risk the investor takes now versus how much they can make later. If five investors are competing for your round, you can push it up. If one is mildly interested, you cannot. That is almost all of what sets an early valuation.
The terms are where control actually lives
The valuation gets the press release. The terms decide who controls the company, and that is the part most founders miss. A founder might celebrate raising five million at a twenty-five million valuation over the same amount at eighteen million, but if the higher number came with worse terms in the contract, it can be the worse deal. There are four terms worth knowing cold.
A liquidation preference decides who gets paid first in a sale. The founder-friendly version is one-times non-participating, which means investors get their money back and that is it. The version to watch out for is participating, where investors get their money back and a share of whatever is left. Protective provisions give investors veto power over big decisions like selling the company, raising again, or issuing shares, so the founder is technically CEO but cannot do those things without approval. Pro rata rights let early investors keep their percentage by putting more into later rounds. Drag-along rights mean that if a majority wants to sell, the smaller shareholders are forced to sell too. All four affect economics, ownership, or control, and they are written for the bad moments, not the good ones.
The clock starts the day the money lands
Once the round closes and the team is celebrating, a clock starts that most founders do not hear. Investors wrote the check because they expect you to grow fast enough to raise the next round, on a timeline they are already counting down. Hiring goals and growth targets get set against the size of the round. A few years ago founders raised about eighteen months of runway and that was enough. In this market eighteen months is tight, and many founders now raise twenty-four to thirty months so they are not pitching with an empty bank account.
This is where the real cost shows up. Dilution is one part. The bigger part is that you sold a piece of the company in exchange for putting yourself on a deadline. You cannot decide to slow down, because if you miss the milestone the next round comes with worse terms, or it does not come at all. One of the most common ways venture-backed startups die is running out of money between rounds, then taking an emergency bridge at a lower valuation, more dilution, or worse terms. Experienced founders do not just ask what valuation they got. They ask what they gave up to get it.
The 2026 split, and the move
The market has split in two. Regular startups now have to prove more before they can raise, with real paying customers, growth that has not slowed, and retention. The old approach of raising on a good idea and figuring out the rest later does not work the way it used to. On the other side, the best AI companies raise at valuations that look nothing like normal startup numbers, with Series A AI medians running roughly thirty-eight percent above non-AI. The market did not get easier. It got more uneven.
The deals changed too. Around ninety-two percent of preseed rounds now use SAFEs, which feels like avoiding the valuation conversation but only delays it. When the SAFE converts at the next round, the cap and discount you negotiated up front decide how much you give away. And the third shift is that not raising at all is a real option now in a way it was not three years ago. The founders who understand all of this raise from the right investor, at the right size, with terms that do not corner them two years later. The ones still running the 2022 playbook give up far more of their company than they need to. The only question is whether you figure that out before you sign or after.
Most of this is outside your control. The part that is not is preparation. A clean cap table, terms you actually understood before you signed, and a record an investor can verify are what keep the headline number from quietly becoming a worse deal. That is the work that pays off in the moments the term sheet was written for.
FAQ
What is dilution, exactly?
When investors fund your company, the company creates new shares and gives them to the investors. Your share count does not change. Your percentage of the company does, because there are now more total shares. You can go from owning half the company to owning forty-one percent without ever selling a share.
Why does everyone say valuation matters least at the early stages?
Because a preseed or seed valuation is not calculated, it is agreed. There is often no revenue and barely a product to put a number on, so the figure is a negotiation about risk and competition. If several investors are competing for your round, the number goes up. If one is mildly interested, it does not.
Which terms should a founder watch most closely?
Four. The liquidation preference, which decides who gets paid first in a sale. Protective provisions, which give investors veto power over major decisions. Pro rata rights, which let early investors keep their percentage in later rounds. And drag-along rights, which can force a sale. These affect economics, ownership, and control more than the headline number does.
Is a higher valuation always better?
No. A round at a higher valuation with a participating liquidation preference or heavy protective provisions can leave a founder worse off than a lower valuation with clean terms. The valuation gets the press release. The terms decide the outcome.
How much runway should I raise in 2026?
More than the old eighteen-month rule of thumb. Fundraising takes longer than it used to, so many founders now raise twenty-four to thirty months of runway so they are not pitching with their bank account close to empty. The clock starts the day the money lands.
Educational only. Not legal, tax, accounting, or investment advice. Deal Box is a technology and advisory platform, not a broker-dealer, placement agent, investment adviser, or custodian, and earns from issuers, never from investors.