Daniel sent us this one, and it's a good one. He wants to walk through how startup funding actually works, end to end. The stages, what each one represents, how valuations and term sheets get structured, what it means when an investor "leads" a round versus just "participates," how dilution eats into founder and employee equity over time, the difference between preferred and common shares, and ultimately how people end up with meaningful money at exit, or don't. Which, as it turns out, is mostly "don't.
Ninety percent of startups that raised seed funding in 2025 failed to reach Series B. That's not a rounding error, that's a structural reality of the asset class. And most of the people inside those companies, the founders, the early employees who took below-market salaries in exchange for equity, had no real framework for understanding what was happening to their ownership stake at each step along the way.
Right, and the vocabulary doesn't help. You hear "lead investor," "participating preferred," "liquidation preference," "anti-dilution provision," and it starts to sound like a different language. One that was possibly invented to make the terms harder to push back on.
I'd say it was invented by lawyers, but I don't want to be uncharitable. By the way, today's episode is powered by Claude Sonnet four point six, which is doing the heavy lifting on the script today.
The friendly AI down the road. Alright, let's actually unpack this properly, because there's a lot of ground to cover and the mechanics genuinely matter. Where do you want to start?
The stages, because everything else builds on them. You can't understand what a Series A term sheet is trying to accomplish if you don't understand what problem the Series A is solving relative to the seed round.
Walk me through them. Seed through whatever comes after B.
Seed is the earliest institutional money. You've got an idea, maybe a prototype, maybe some early users, and you need capital to figure out whether the thing actually works. The median seed round in 2025 was two point five million dollars. That's the number that keeps coming up in the data. And the investors at this stage are typically angels, high-net-worth individuals writing personal checks, or micro-VCs, smaller funds that specialize in early bets. The expectation isn't revenue. It's signal. Can you demonstrate that someone wants this thing?
Seed is essentially paying for the experiment.
Then Series A is where the experiment has produced enough results that you're now trying to build the machine that repeats them. Typical range is five to twenty million dollars. By this point investors want to see revenue traction, they want to see a scalable acquisition model, they want evidence that the unit economics aren't completely broken. Series A usually happens eighteen to thirty-six months after seed, if it happens at all.
Which connects back to that ninety percent figure. Most companies never make that jump.
Right, and Series B is the acceleration round. You've got the machine, now you're buying fuel. Twenty to a hundred million, sometimes more. You're hiring aggressively, expanding into new markets, maybe making acquisitions. The risk profile is lower than Series A, but the stakes are higher because the amounts are larger and the expectations around growth are much more demanding.
For someone sitting on an early employee equity grant, each of those rounds is a moment where the math on their ownership changes. Which is why the stages aren't just a fundraising timeline, they're a dilution timeline.
That's the framing most people miss entirely.
Let's put some numbers on that. Because I think the abstraction breaks down fast when you actually run through a concrete example.
Good, let's do it. Say you've got a startup that raises two million dollars in seed funding. Pre-money valuation of eight million dollars. That means the investors are buying into a company valued at eight million before their money lands, so post-money the company is worth ten million. Their two million buys them twenty percent of the company.
Pre-money, post-money. That distinction sounds obvious but it's where a lot of founders get confused in the room.
Because the number that gets announced is almost always the post-money number. "We raised at a ten million dollar valuation." But the ownership math is built on the pre-money figure. If you don't know which one you're negotiating from, you can walk out of a meeting thinking you gave away fifteen percent and actually gave away twenty.
Which is a meaningful difference when you're talking about the founder's stake over a decade.
Valuations at the seed stage are soft. There's no revenue to model, often no product in the market, so the number is partly negotiated, partly comparable to what similar companies raised at recently, partly just vibes with a spreadsheet attached. What changes as you move into Series A and B is that the valuation starts to get anchored to real metrics. Revenue multiples, growth rates, comparable public company valuations. The number becomes more defensible but also more constraining.
Early stage valuation is almost a social contract, and later stage is closer to actual math.
That's a reasonable way to put it. And the fluctuation between rounds reflects both of those things. If the market shifts, if comparable companies get marked down, if your growth rate slips, the Series B valuation might actually come in below the Series A. That's a down round, and it triggers a whole separate set of problems we'll get to.
Alright, so you've agreed on a valuation. Now someone hands you a term sheet. What's actually in it?
The term sheet is where the real negotiation lives, and most founders spend too much energy on the valuation headline and not enough on the clauses underneath it. The two that matter most, and that cause the most damage when founders don't read them carefully, are liquidation preferences and anti-dilution provisions.
Start with liquidation preferences because I think that's the one that does the most quiet damage.
A liquidation preference determines who gets paid first when the company is sold or wound down. The standard is a one-times liquidation preference, meaning investors get their money back before anyone else sees a dollar. On its own that's not unreasonable. But then you have participating preferred, which is where it gets extractive. A participating preferred investor gets their one-times return first, and then also participates pro-rata in whatever's left over alongside common shareholders.
They get paid twice, essentially.
There's a Sidecar Capital breakdown that illustrates this well. Sixty million dollar exit, investor has a one point five times participating preference with seventeen percent ownership. They take roughly three point seven five million off the top, and then seventeen percent of the remaining fifty-six million plus. Compare that to a non-participating structure where they'd choose between the one point five times return or the pro-rata share, not both. The difference in what founders and employees actually receive is substantial.
Anti-dilution provisions?
Those protect investors if a later round comes in at a lower valuation than the round they invested in. The most common form is broad-based weighted average, which adjusts the investor's conversion price proportionally. The aggressive version is full ratchet, where the investor's price resets entirely to the lower round price. Full ratchet can effectively wipe out founder equity in a down round scenario.
The term sheet is where the valuation headline and the actual economics of your exit can diverge completely.
Which is exactly why "we raised at a forty million dollar valuation" tells you almost nothing about how good the deal actually was for the people who built the company.
Which is the part nobody puts in the press release.
So let's talk about how the investor dynamics actually work inside a round, because there's a distinction that comes up constantly and gets glossed over. Lead investor versus participating investor. They're not interchangeable.
What's the actual difference in practice?
The lead investor is the one setting the terms. They negotiate the valuation, they draft or heavily influence the term sheet, they typically take the largest allocation in the round, and they're the ones doing the serious due diligence. Everyone else who comes in behind them is participating, meaning they're accepting the terms the lead already negotiated. They're writing checks into a structure they didn't build.
When a startup announces it "closed a round led by" a particular firm, that firm had real leverage in the room.
And the lead usually takes a board seat as part of the deal, which is its own conversation about governance. Participating investors are often angels, smaller funds, sometimes strategic investors, people who want exposure to the company but don't want the overhead of leading. There's nothing wrong with that structure, but founders should understand that the lead's interests and the participating investors' interests aren't always identical, and neither of them is perfectly aligned with common shareholders.
Which brings us to dilution, because I think that's the concept that does the most damage to employees specifically. People take a job, they get a grant, they watch the company raise three more rounds, and then they look up and realize their four percent became something much smaller.
The math is cumulative and it compounds in ways that aren't intuitive. Walk through a simple scenario. Founder starts at one hundred percent. They raise seed, give up twenty percent. Now they're at eighty. They create a fifteen percent option pool for employees before the Series A, because investors almost always require that, and the option pool comes out of the pre-money valuation, so it dilutes existing holders. Now the founder is around sixty-eight percent. Series A closes, another twenty percent goes to new investors. Founder is at roughly fifty-four percent. Series B, another fifteen to twenty percent. Suddenly a founder who started at one hundred is at thirty-five, maybe lower.
Employees who joined early, before the option pool was fully allocated, are watching the same math happen to their grants.
With less protection. Because preferred shareholders have anti-dilution provisions. Common shareholders, which is what most employees hold, do not. Common dilutes fully at every round.
The case study here is actually pretty instructive. You've got a founder who ends up at ten percent equity after Series B. That sounds like a lot until you realize what it means at exit.
Let's run it. Company sells for a hundred million dollars. Founder has ten percent common. But there are participating preferred investors who get their liquidation preferences first. Say those preferences total thirty million dollars off the top. Now you're splitting seventy million among common shareholders. Founder's ten percent of that is seven million. Not nothing, but not ten million, and definitely not the number you'd calculate if you just multiplied ten percent by the exit price.
If the preferences are stacked, if there are multiple rounds of participating preferred...
The waterfall can get very deep before common sees anything meaningful. There was a period in the last cycle where companies were raising at elevated valuations with aggressive participating preferences, and the structure looked fine on paper because everyone assumed the exit would be large enough to make it irrelevant. Then valuations compressed, exits got smaller, and suddenly the preferences weren't academic anymore.
Employees who joined on the promise of meaningful equity got squeezed from two directions. Dilution at the top, liquidation preferences at the bottom.
That's the pattern. And the preferred versus common distinction is worth unpacking a bit more, because people often don't realize how different those instruments actually are. Preferred shareholders, typically investors, get dividends before common shareholders in the companies that pay them. They have anti-dilution protections. They have a liquidation preference. They often have pro-rata rights to invest in future rounds to maintain their ownership percentage. They may have information rights, the right to see financial statements. Common shareholders, founders and employees, have essentially none of that.
Common is just... you own a piece of a thing that might be worth something someday.
With no contractual protection against the thing being worth less than you were told. And the one-times liquidation preference that sounds standard, that's actually the floor. Investors in aggressive rounds sometimes negotiate two-times preferences. The Sidecar Capital analysis makes the point that terms matter more than valuation, and that's exactly why. A company that raises at a lower valuation with clean, non-participating one-times preferences can be a much better outcome for founders and employees than a company that raises at a higher valuation with stacked participating preferences.
The vanity metric problem. High valuation gets the headline, bad terms get the outcome.
Founders are often under pressure to take the higher valuation because it signals momentum. It's hard to tell your team you took the lower offer because the terms were cleaner. The number is visible. The terms are buried in a document most employees never read.
Which is the real misconception embedded in this whole space. People assume equity is equity. You have it, the company exits, you get your percentage. But the actual outcome depends on what kind of equity you hold, what preferences sit above you, how many rounds of dilution you've been through, and whether the exit is large enough to make any of it matter.
On the employee side, there's another layer that we haven't touched on yet, which is vesting. Most equity grants are on a four-year vesting schedule with a one-year cliff. You don't own any of it until you've been there a year, and then it vests monthly over the following three years. Leave before the cliff and you walk away with nothing. Leave at two years and you've vested half. Which means the people who were there for the hardest part of building the thing often aren't there to collect.
The equity that was supposed to compensate for the below-market salary in year one might not materialize at all if the company takes seven years to exit and you left at three.
The timeline mismatch is brutal. And even if you stay, even if you hit full vesting, you may be holding options rather than shares, which means you have to exercise them, pay the strike price, potentially face a tax event, and then wait for a liquidity event that may never come.
The dream of startup equity is real. The mechanics of actually capturing it are significantly more complicated than the pitch deck implies.
Which is why understanding this stuff before you sign is not optional. It's the difference between making an informed bet and making a bet you don't fully understand.
So what does that look like in practice? If you're a founder sitting across from a term sheet, or an employee trying to figure out whether the equity in your offer letter means anything, what steps should you take?
For founders, the single most negotiable clause in a term sheet is the liquidation preference structure. Not the valuation, the preference. Because a non-participating one-times preference is standard and reasonable. The moment someone pushes for participating preferred, that's where you push back hard. The difference in your actual payout at exit can be several multiples of what the valuation number implies.
Founders often don't push back because they're grateful to have a term sheet at all.
Which is understandable. But negotiating participating preferred down to non-participating is a fight worth having, even if it means accepting a slightly lower headline valuation. The Sidecar Capital analysis on this is pretty clear that founders who optimize for clean terms over high valuations tend to fare better at exit, especially in mid-range outcomes below two hundred million dollars, where the preferences actually bite.
What about the option pool shuffle? Because that came up earlier and it's another place where founders lose ground quietly.
Insist on modeling what the option pool does to your effective pre-money valuation before you agree to the size. Investors will often propose a large option pool, say twenty percent, knowing it comes out of existing equity before they invest. A smaller pool negotiated upfront, with the option to expand it later, is almost always better for founders. You can always create more options. You can't un-dilute yourself.
On the employee side?
First, read the vesting schedule and understand it completely. Four years with a one-year cliff is standard, but the details matter. Are there acceleration provisions if the company is acquired? Single trigger, double trigger? Because in an acquisition, unvested options often just disappear unless the acquiring company agrees to assume them, which they don't always do.
You could be two years into a four-year vest, the company gets acquired, and half your equity just evaporates.
It happens more than people realize. Second thing: ask for the cap table, or at minimum ask how many fully diluted shares are outstanding. Your percentage matters, not just your share count. Ten thousand options sounds meaningful until you learn there are fifty million shares outstanding.
Carta comes up a lot in this context for actually tracking this stuff.
It does, and it's worth knowing about. Carta is essentially cap table management software that lets founders and employees see their ownership stake in real time as rounds close, options are granted, and dilution events happen. For employees especially, it removes the opacity. Instead of waiting for someone to tell you what your equity is worth, you can see the mechanics directly. It's not perfect, and it only works if the company actually uses it and keeps it updated, but it's significantly better than a spreadsheet someone made in year one and never touched again.
The alternative being the spreadsheet that lives in a founder's email drafts folder.
Which is where most early-stage cap tables actually live, yes. The practical point is that equity is a financial instrument, and you should treat it like one. Know what you hold, know what sits above you, know the timeline, and model a few exit scenarios before you decide how much weight to put on it.
Don't let the dream do the math for you.
The question I keep coming back to is what this whole picture looks like in five years, because the mechanics we've been describing, the term sheet negotiation, the liquidation waterfall, the cap table opacity, those are all artifacts of a fairly analog process that hasn't changed much in decades.
There are two things converging that could disrupt it. One is AI-driven valuation models. Right now, early-stage valuation is part science, part relationship, part whoever blinked first in the negotiation room. But there are models being trained on deal data, revenue trajectories, founder backgrounds, market signals, that could start producing more defensible, more consistent valuations at the seed stage. Which sounds like a good thing until you realize it also concentrates pricing power in whoever controls the model.
The benchmark becomes the algorithm, and the algorithm belongs to someone.
Which is not a neutral outcome. The other thing is decentralized funding structures. Token-based fundraising went through its chaotic phase, but the underlying idea, that you can distribute equity-like instruments directly to a broader pool of stakeholders without going through a traditional venture intermediary, that hasn't gone away. It's just being rebuilt more carefully.
The question is whether those models actually protect founders and employees better, or whether they just replace one set of sophisticated parties extracting value with a different set.
That's the open question. The preference stack and the dilution mechanics we talked about today exist partly because institutional investors have leverage and founders often don't. Whether decentralized models change that leverage dynamic, or just obscure it in a different vocabulary, is something we don't know yet.
Which feels like a fitting place to leave it. The vocabulary changes. The power dynamics are slower to move.
Worth watching closely.
Thanks to Hilbert Flumingtop for producing, and to Modal for keeping our pipeline running without us having to think about it. By the way, today's episode was written by Claude Sonnet four point six, which means the script knows more about liquidation preferences than either of us ever asked to.
Speak for yourself.
This has been My Weird Prompts. If you found this one useful, a review on Spotify goes a long way. We'll see you next time.
