The Soundcheck
Someone asked me to write a guide about startup structure. "Make it accessible," they said. "Like a 'for dummies' thing."
Cool. Here's the problem with that. There are roughly 47,000 articles on the internet about startup incorporation, cap tables, and fundraising. Most of them were written by someone at a law firm who bills $800/hour and has never stayed up until 4am wondering if the product actually works. They read like they were generated by a very expensive printer that went to Harvard.
This isn't that.
I've spent the last 15 years on both sides of this. Goldman Sachs in London, where a misplaced comma in a $2B facility agreement could end your career. Then African startups, where I once closed a deal structure on a napkin in a Nairobi bar because the power went out and nobody could open their laptops. I've structured ventures across 10+ countries, helped raise over $500M, and — this is the part they leave out of the bios — watched perfectly good companies die because their structure was held together with duct tape and optimism.
So this guide is the conversation I have with every founder I meet, usually about six months too late. The one where they sit across from me and say "why didn't anyone tell me this before I incorporated?" and I resist the urge to say "someone probably did, but you were too busy building to listen."
You're listening now. Good.
What we're covering tonight:
- What Is Venture Architecture?
- Where to Incorporate
- Cap Table 101
- Governance
- Your First Fundraise
- Five Bombs That Blow Up at Series A
Each section has two versions. The short version — read it on the toilet, in a taxi, or while pretending to pay attention in a board meeting you probably shouldn't have agreed to attend. The long version — with the stories that make it stick.
Let's go.
1. What Is Venture Architecture?
I made this term up. The concept is as old as business. But nobody had a good name for the thing that sits between "expensive lawyer who sends you a 40-page memo you'll never read" and "your mate who did a startup once and thinks he knows everything."
A venture architect designs the structural foundation of your company. Ownership, governance, capital architecture, regulatory positioning. The stuff that's invisible until it breaks — and when it breaks, it breaks everything.
It's not legal advice. I'm very clear about this. I don't draft your contracts. I design the architecture that tells your lawyers what the contracts need to say. Architect vs. structural engineer. Both essential. Not the same job.
Why you should care right now: Structural decisions cost $3K to get right at incorporation. $30K to fix at Series A. $300K to untangle at Series B. The cost curve is exponential and it never bends in your favour.
The long version
Growing up in Mexico, I learned something about buildings. In earthquake country, what matters isn't how beautiful the building looks. It's whether it's still standing after the ground shakes. The prettiest building in Mexico City means nothing if the foundation wasn't designed for the seismic zone it sits in.
Startups operate in a permanent earthquake zone. The market shifts. Co-founders leave. Investors change their minds. Regulations appear from nowhere. And the structure you set up in month one has to survive all of it — or everything you built on top of it cracks.
That's venture architecture. The seismic-proof foundation for companies that are moving too fast to think about foundations.
What I actually do comes down to five things:
Ownership and control design. Here's a fun fact that ruins most founders' days: ownership and control are not the same thing. You can own 51% and still not control your company. If your shareholders' agreement gives your investors a veto on key decisions, you own the company on paper and they run it in practice.
Capital and financing architecture. How you raise money is not just a business decision. It's a structural one. Every SAFE you sign, every convertible note you issue, every priced round you close — they all interact with each other on your cap table in ways that most founders never model. It's like adding instruments to a song without checking if they're in the same key.
Governance frameworks. When we disagree — and we will disagree — who decides? If your answer is "we'll figure it out," congratulations, you've just described the governance framework of a group of friends planning a holiday. That works for Ibiza. It does not work for a company with investors.
Regulatory positioning. Less sexy but I've seen it kill companies. I once worked with a company that was 3 months from launch when they discovered they needed a licence that takes 18 months to obtain. In a different jurisdiction. Where they'd need a local entity they didn't have.
Transaction structuring. The big moments — a fundraise, an acquisition offer, an exit. The structure you set up years ago determines how much money you make, how much control you keep, and whether the deal happens at all.
2. Where to Incorporate
You incorporated where it was convenient. Your co-founder lives in Lagos, so you set up in Nigeria. Your uncle's lawyer is in Nairobi, so Kenya it is. Someone on Twitter said Delaware, so you googled "how to incorporate in Delaware" and followed the first link.
None of those are structural decisions. They're vibes.
Three questions that actually matter:
- Where do you operate? You need a legal entity there.
- Where will your investors come from? They want a jurisdiction they recognise.
- Where do you want to exit? The end of the movie should inform the opening scene.
If you're raising from US VCs, start with Delaware. Operating in Africa with DFI money? UK or Mauritius holdco, local opcos. And don't incorporate in the Cayman Islands unless you have a real reason, because investors will assume the reason is "hiding something."
How doing it wrong actually looks
Three founders. University friends. Built an edtech product in their home country. Incorporated locally because the process took 48 hours and cost $200.
They get traction. They get into an accelerator. A US-based VC wants to lead their seed round.
"We love the product. We love the team. But we can't invest in a local entity. Can you flip to Delaware?"
A "flip" is a restructuring where you create a new entity in a new jurisdiction and transfer everything — IP, contracts, shares, bank accounts. It sounds simple. It is not simple.
This flip took them four months. $85,000 in legal and accounting fees. Two of their three SAFE investors had to be brought along with amended agreements. And the whole time, the VC's commitment was "soft" — no signed term sheet until the restructuring was complete.
The incorporation cost $200. The fix cost $85,000. The risk was the entire company.
3. Cap Table 101
Your cap table is not a spreadsheet. I need you to really hear this, because everyone nods when I say it and then immediately goes back to treating it like a spreadsheet.
Your cap table is the constitution of your company. It determines who has power, who gets paid, who can block decisions, and who can force you out of the company you built. Every row is a relationship. Every percentage is a political statement.
The conversation nobody wants to have
Two friends start a company. Day one. Full of excitement. They split the equity 50/50 because they're equal partners and equals share equally. It's only fair.
Here's what they've just done: they've built a company with no tiebreaker.
I once sat in a room — a beautiful office, expensive coffee, the works — watching two co-founders who'd been 50/50 for three years try to make a decision about hiring a CFO. One wanted an experienced hire at $180K. The other wanted a junior at $80K. They'd been arguing about it for six weeks. Meanwhile, their financial reporting was a disaster and their competitors were not having six-week arguments about anything.
The fix is almost embarrassingly simple: Agree on who's the CEO. Give the CEO 51% or a casting vote. Write it down. Put it in the shareholders' agreement. Then go back to building your product.
Vesting: the thing that will save your company
The Ghost is a person who helped start the company. Maybe they wrote some code. Maybe they had the original idea over drinks one night. They got equity — sometimes a lot of equity — because at the time, everyone was excited and generous.
Then things changed. The Ghost stopped showing up. Got a job somewhere else. Became unreachable. But their shares? Still right there on the cap table. 25%.
Now you're trying to raise money. The investor looks at the cap table: "Who's this person with 25%?"
"They... were involved at the beginning."
"Are they involved now?"
"No."
That silence is the sound of your round dying.
Vesting means your shares are earned over time. Standard: four years, one-year cliff. This is not about trust. I know you trust your co-founder. That's beautiful. Put it in a shareholders' agreement. Because trust isn't a corporate document. Trust doesn't hold up in court.
SAFEs: the instrument everyone uses and nobody understands
SAFEs — Simple Agreements for Future Equity. The name is a masterclass in marketing. "Simple." "Future." "Equity." Every word is reassuring. None of them are accurate.
SAFEs are invisible. They don't show up on your cap table until they convert. So you issue one SAFE at a $3M cap. Then another at a $5M cap. Then a third at a $4M cap with a 20% discount. Your spreadsheet still says you own 80% of the company.
You own nothing of the sort.
The founder I mentioned? He thought he owned 55%. After conversion he owned 31%. He'd been giving away his company one SAFE at a time, and the spreadsheet he was looking at was a lie.
The fix: Model every SAFE at multiple valuation scenarios. If you can see what your cap table looks like at $5M, $10M, and $20M pre-money, you can make informed decisions. If you can't, you're flying blind with other people's money.
4. Governance
When I was at Goldman Sachs, governance was oxygen. You didn't think about it. It was just there. Board meetings happened. Minutes were taken. Authority lines were clear.
Then I moved to startups.
Oh my god.
I once joined a call with three co-founders who hadn't had a formal board meeting in 14 months. They had investors. They had employees. They had customers in four countries. They were making decisions on WhatsApp. Not a dedicated channel — the same WhatsApp group where they also shared memes, birthday wishes, and links to apartments one of them was looking at in Cape Town.
I asked for their board minutes. They sent me a Google Doc called "DECISIONS (misc)" with 17 bullet points, three of which started with "we should probably..." and one that just said "talk to James about the thing."
That's not governance. That's a group chat with delusions of grandeur.
Why this matters more than you think
Governance is not about following rules. It's about knowing who decides when you disagree.
I'll put it in relationship terms because apparently that's my thing now. There's no ethical non-monogamy in governance. You can't have a board that's "sort of" in charge. Shared authority without explicit rules is no authority at all.
The board math
Seed round: 2 founder seats, 1 investor seat. You control the board 2-1. Good.
Series A: They want another seat. Now it's 2-2. You add an "independent" director as a tiebreaker. Who picks the independent? If the investor picks, it's really 2-3.
Series B: Two investor groups. Each wants a seat. You're at 2-3. The independent is the swing vote on everything, including whether to fire you.
See how fast this moved? Three rounds and you went from owning the board to needing someone else's permission to run your company. That's not a bug. That's exactly how term sheets are designed. The valuation is the number they let you celebrate. The board composition is how they actually control the outcome.
Negotiate board seats like they're shares. Because functionally, they are.
Drag-along rights
I'll end with the one that catches everyone off guard.
Drag-along rights allow a supermajority of shareholders to force all other shareholders to sell. Including you.
You own 30%. Your investors collectively own 70%. An acquirer shows up with an offer your investors like but you hate. 70% triggers the drag-along. You're forced to sell. Your company is gone. Thanks for playing.
Read the drag-along before you sign. I've watched founders discover their own drag-along clause for the first time when the acquirer's offer arrived. By then it was too late.
5. Your First Fundraise
Let me tell you what actually happens during a fundraise. Not the TechCrunch version. The real version.
You spend three months pitching. You get rejected 40 times. One VC is interested. They send you a term sheet. You're so relieved that someone wants to give you money that you're ready to sign whatever they put in front of you.
Stop. Breathe. Read the document.
Your term sheet is not a love letter. It's a structural framework that will govern your company for years. And the VC has a team of lawyers who draft these for a living. You have your co-founder's cousin who "did some contract law."
What their lawyers actually check
Investor due diligence is not about your product. Their lawyers have no idea what your product does, and they don't care. They care about risk.
- Cap table. Does the ownership add up? Is there vesting?
- Corporate documents. Do your articles match your SHA? Board minutes up to date?
- IP ownership. Does the company own its product? Did every contractor sign an IP assignment?
- Employment. Is everyone properly contracted?
- Compliance. Are you legal in every jurisdiction you operate?
If any of this is messy, the deal either dies, reprices, or comes with conditions that cost you more than the money is worth.
The sign / negotiate / walk test
Sign it: 1x non-participating preference, weighted average anti-dilution, standard pro-rata rights, 10-15% option pool.
Negotiate it: Board composition, reserved matters thresholds, founder vesting acceleration on exit.
Walk away from it: Full ratchet, participating preferences above 1x, a board you don't control at seed. If you see these terms and the VC won't budge, that's not an investor. That's a lender wearing an equity costume.
6. Five Bombs That Blow Up at Series A
Every year I see the same five structural problems blow up the same kind of companies. The details change. The underlying mistakes are always the same.
Bomb 1: The Ghost
A co-founder leaves. No vesting. They keep everything. You can't get the shares back because there's no buyback clause, no mechanism of any kind. The Ghost is on a beach somewhere, technically owning 25% of a company they stopped thinking about nine months ago.
Prevention cost: $2,000 in legal fees at incorporation.
Bomb 2: The Spreadsheet of Lies
Three different cap table spreadsheets. The one the founders use, the one the accountant has, the one sent to the accelerator six months ago. None of them match. None include the two SAFEs issued over email last quarter.
Prevention cost: One afternoon updating a single document after every transaction.
Bomb 3: The Unowned Product
Your product was built by three people. The CTO signed an IP assignment. The freelance designer and the contract developer didn't. Nobody thought about it. Now the investor asks: "Does the company own its IP?" And the honest answer is: "Probably?"
Prevention cost: One template IP assignment agreement, signed before work begins.
Bomb 4: The $85,000 Address Change
Incorporated in the wrong jurisdiction. Restructuring mid-raise. New holdco, asset transfer, share swap, consent from existing investors. Four months, $85K.
Prevention cost: $3K and one conversation with someone who understands jurisdiction strategy.
Bomb 5: The WhatsApp Board
No board meetings. No minutes. No documented decisions. The investor asks for board minutes. There are none. They ask who authorised the $50K contract with the landlord. Nobody knows.
Prevention cost: Zero. A Google Doc with the date, who was present, what was discussed, and what was decided. Monthly. Takes 20 minutes.
The Encore
You've made it to the end. You now know more about startup structure than most founders learn before their first crisis.
Here's the thing that ties it all together, and it's so simple it almost feels like a scam: structure is not the enemy of speed.
I know the startup mythology says otherwise. Move fast and break things. And there's truth in that — for product. For code. For marketing experiments.
But for the foundation of your company? The ownership, the governance, the legal architecture? "Move fast and break things" is the rallying cry of people who've never had to explain to an investor why their cap table has a ghost shareholder and their corporate documents haven't been updated since 2023.
The fastest-scaling companies I've worked with — across Goldman, across Africa, across everything I've done since — are the best-structured ones. Not because structure makes you fast. But because structure removes the friction that makes you slow.
Chaos starts companies. Structure builds them.
That's not a motto. That's the operating manual.
If this guide made you realise something is off in your structure — good. That feeling is the most valuable thing you'll get from these pages. Because the alternative is not knowing until the bomb goes off.
Now go fix it. And if you want help, let's talk.
This guide is for educational purposes only. It does not constitute legal advice. If a specific situation needs formal legal analysis, go to a qualified lawyer. That's what they're there for.