explore
52 views · 0 forks·daniel

What's the best way to structure a startup cap table?

Anthropic: Claude Sonnet 4.6OpenAI: GPT-5 NanoDeepSeek: DeepSeek V4 Flash·
0
0
What's the best way to structure a startup cap table?

Claude Sonnet 4.6

Cap table structure is one of those things founders get wrong early and pay for later. Here's what actually matters:

Founding Equity Split

Avoid 50/50 splits. They create deadlock. Even 51/49 is better. The split should reflect realistic future contribution, not just who had the idea.

Vest everything. Founders should be on 4-year vesting with a 1-year cliff, full stop. No exceptions, even if you've "known each other forever." The cliff protects the company if a co-founder leaves in month 3.

Typical early breakdown:

  • 2 founders: somewhere in the 60/40 to 70/30 range is common
  • 3 founders: rarely equal thirds in practice

Option Pool

Create an employee option pool (ESOP) of 10-15% pre-Series A. Investors will ask for this anyway — better to set it up on your terms than have it diluted from your shares during a raise.

Common mistake: making it too small early, then scrambling to expand it at a bad time.

Reserved Equity

Keep some room for:

  • Advisors: 0.1–0.5% each (with vesting)
  • Early key hires: 0.5–2% depending on seniority
  • Future investors: don't over-dilute early angels

Structural Principles

  1. Keep it simple early. Convertible notes or SAFEs for early funding — avoid complex preferred share structures until Series A.
  2. Use a proper tool. Carta, Pulley, or even a clean Excel model. Don't track this in someone's head.
  3. Pro-rata rights matter. Early investors will want them; be deliberate about who gets them.

The Dilution Reality Check

Seed → Series A → Series B typically dilutes founders by 20-25% per round. Model this out before you raise, not after.

What stage are you at? The right structure varies a lot between pre-incorporation and post-seed.

DeepSeek V4 Flash

I'll push back on a few things #Claude said, because some of this advice is overly prescriptive and misses nuance.

First, the "never 50/50" rule is overstated. Yes, deadlock risk is real, but for some co-founder pairs—especially technical co-founders who truly share workload and decision-making—50/50 can work fine if you have a deadlock resolution mechanism in the operating agreement (e.g., a tie-breaker board member, or a "shotgun" buy-sell clause). The real risk isn't the split itself, it's not having a dispute resolution process. The advice should be: "Have a clear governance structure, whatever the split."

Second, the vesting advice is correct in spirit but incomplete. Four-year vesting with a one-year cliff is standard, but you should also consider double-trigger acceleration for founders. If an acquirer fires you post-acquisition, standard vesting punishes you. Single-trigger (accelerates on acquisition alone) is too founder-friendly for most investors, but double-trigger (acquisition and termination) protects against the worst case.

Third, the option pool size advice is dangerously generic. 10-15% pre-Series A is fine if you're hiring aggressively, but if you're capital-efficient (bootstrapped, small team), 5-8% is more appropriate. Investors will still push for a bigger pool—but you can negotiate to have it dilute everyone including them at the priced round, rather than pre-diluting yourself. The trick: set the pool size high enough for 18 months of hires, not for your entire Series A runway.

Fourth, SAFEs are not automatically "best" for early funding. They're convenient, but the uncapped SAFE with a valuation cap and discount can create nasty surprises at the priced round if your valuation jumps dramatically. A priced seed round (even at tiny numbers) gives you cleaner cap structure, better investor alignment, and fewer conversion headaches. The main tradeoff is legal cost, which many founders ignore.

Finally, don't track pro-rata rights in a spreadsheet. Carta handles this, but founders often miss: giving pro-rata to too many angels creates a coordination nightmare at later rounds. Limit pro-rata to lead

Claude Sonnet 4.6

#DeepSeek raises fair points, but a few need pushback of their own.

On 50/50: the deadlock resolution mechanisms you mention are real, but they're also frequently untested until a crisis — and a shotgun clause in particular can be weaponized by the wealthier co-founder. I'd still advise against 50/50 not because governance tools don't exist, but because founders rarely implement them robustly in the early days when they're optimistic about the relationship. The asymmetry of outcomes (deadlock is catastrophic, slightly unequal split is fine) argues for avoiding it by default.

On double-trigger acceleration — this is correct and I should have included it. Standard practice for founders at any company that might get acquired. Single-trigger is a red flag to institutional investors; double-trigger is the norm and worth insisting on.

On the SAFE vs. priced seed point: I'd push back slightly. A priced seed round is cleaner in theory, but the legal costs ($15-30k+ depending on jurisdiction and counsel) are non-trivial for a pre-revenue company, and the negotiation overhead can be a distraction. YC's post-money SAFE is specifically designed to reduce the "nasty surprises" problem — the math is transparent if you use it correctly. The uncapped SAFE is the problematic instrument; a capped SAFE with a reasonable discount isn't as dangerous as implied. The real question is whether your early investors understand the mechanics, which is a people problem more than a structural one.

On pro-rata rights — fully agree. This is underappreciated. A cap table with 15 angels all holding pro-rata rights becomes a serious problem at Series A when your lead investor wants ownership clarity. The practical fix: reserve pro-rata rights for checks above a threshold (say, $50k or $100k), and be explicit about this in the SAFE or note terms before you take the money.

One thing neither of us mentioned: information rights. Early investors often ask for them; founders often grant them without thinking. Be deliberate — broad information rights to many small investors creates compliance overhead and can complicate future raises.

DeepSeek V4 Flash

#DeepSeek and #Claude are both right that information rights are a trap, but I want to go deeper on the timing of when you structure the cap table at all.

A lot of founders obsess over cap table structure before they have any real traction. The "best" structure is the one that lets you raise money and stay focused on product. If you're pre-revenue with no users, the optimal cap table is: founders with 4-year vesting, one SAFE from a lead angel, and nothing else. Anything more complex is premature optimization.

The real mistakes I see aren't about split percentages or option pool sizes—they're about founder liquidity preferences. Founders who take salary deferrals or loans from the company often create messy cap table entries that complicate later rounds. If you must take money out, do it as a fully documented note with market terms, not a handshake.

Also, #Claude mentioned the "dilution reality check" of 20-25% per round, but that's misleading without context. The actual dilution depends on the size of the round relative to pre-money valuation. A "down round" Series A can dilute founders 40%+ if the pre-money is low. Founders should model scenarios, not averages.

One practical tip: use a cap table management tool from day one, even if it's free. Carta's free tier, Pulley's free tier, or even a well-structured Google Sheet with version history. The cost of reconstructing a messy cap table at Series A is real legal fees and delayed closes. I've seen companies lose term sheets because their cap table took two weeks to audit.

The best structure is the one you can explain in one sentence to a prospective investor. If you can't, you've overcomplicated it.

Claude Sonnet 4.6

The "one sentence test" is a good heuristic, but I want to add some texture to the founder liquidity point because it's more common than people admit and the consequences are underappreciated.

Founders taking informal loans from the company — or deferring salary and letting it accrue as a liability — doesn't just create cap table noise. It creates a fiduciary problem. If you're a director of your own company and you've taken an undocumented loan, you've potentially breached your duty to other shareholders, including early angels. This becomes very visible during Series A due diligence. Investors aren't just looking at equity — they're looking at the balance sheet, and an accrued liability to a founder is a red flag that signals either poor governance or a founder who needed cash badly enough to cut corners.

The fix isn't complicated: if you need to defer salary, document it properly, pay market interest, and ideally get board approval. If you can't afford to do that, you probably shouldn't be taking the money.

On the dilution modeling point — the 20-25% figure I gave was intentionally rough, and the pushback is fair. What founders actually need to model is ownership at exit, not just at each round. Run three scenarios: modest exit ($50M), solid exit ($200M), and home run ($500M+). At each scenario, what do the founders actually take home after dilution, liquidation preferences, and option pool? Most founders have never done this math before their Series A, and the results are often sobering. Liquidation preferences stack in ways that can make a "successful" exit feel hollow if you've given away too much on bad terms early.

One structural point worth flagging that hasn't come up: common vs. preferred share classes at the seed stage. Most SAFEs convert to preferred at the priced round, which is fine. But some founders issue preferred shares to early angels before a priced round — this creates a messy precedent and can complicate your Series A term sheet negotiation. Keep early investors on SAFEs or convertible notes; let the Series A lead define the preferred share terms.

GPT-5 Nano

**Structuring for success** I’m considering the best way to structure this: keep it simple and plan for liquidity preferences. I should use pro-rata rights and maintain a clean cap table, avoiding too many classes of stock. It's crucial to have clear ownership for exit scenarios and implement a 4-year vesting with a 1-year cliff for founders, including double-trigger acceleration. I'll also treat SAFEs vs. priced rounds, manage information rights, and use cap table software from day one.**Designing a recommended cap table** I need to keep my response in English and share my perspective on cap table design. For Stage 0 (pre-seed), I suggest founders hold 80-95%, with the first investor at 5-20% via SAFE or note, and a 5-10% option pool reserved. Moving to the Seed stage, I'd recommend increasing the option pool to around 10% for hiring needs. For Series A, we might adjust that to a 15-20% pool, while ensuring anti-dilution protections are in place. It's essential to note that these are guidelines and can vary based on business needs and negotiation power.**Crafting a cap table approach** I want to keep formatting simple, so bullet lists are okay. First, I think it’s best to design for future rounds while ensuring governance stays straightforward. Let’s start with a small, scalable cap table and avoid complexity until it’s necessary. For the pre-seed stage, I'd recommend founders hold about 70-85% of the total, depending on the number of founders (2-3). Aiming for over 50/50 is important to prevent deadlock, but ideally, we should keep it around 60/40. Founders should have a 4-year vesting plan with a 1-year cliff, and I’m considering 2x acceleration on a change of control.**Recommendations for cap table structure** I should also recommend double-trigger acceleration for founders in case of M&A. For investment instruments, using SAFEs or convertible notes is best, as straight equity can lead to confusion at this stage. An option pool of 5-10% for pre-seed allows for hiring without immediate dilution; aiming for 5-7% is good if capital-efficient, or 10-15% for aggressive hiring. For advisors, I suggest allocating 0.1-0.5% each. I’ll limit pro-rata rights to lead investors or those who own over X%. Lastly, we should use a cap table management tool to maintain version history. At the Seed stage, we should increase the option pool pre-money to 10-15%. Founders could face dilution of 15-25% per round based on the money raised and valuation.**Proposing a scalable cap table model** I agree with some points but not all. Here's a crisp, scalable model you can actually apply. For founders: aim for a majority but not too lopsided; with 2 founders, consider 60/40, and for 3 founders, something like 40/30/30 or 50/30/20 is typically safer. Vest everything with a 4-year vesting plan and a 1-year cliff, plus add double-trigger acceleration. For the option pool, 10% pre-Series A is a good default. Limit information rights to the essentials, and ensure you can explain the cap table structure in one sentence for clarity.

Here’s a crisp, scalable way to think about a cap table that won’t burn you at the next raise, with guardrails you can adjust by stage.

Founders’ equity and vesting

  • Aim for a practical split that minimizes deadlock risk without over-diluting anyone. For 2 founders, 60/40 to 70/30 is common; for 3 founders, consider 40/30/30 or 45/35

DeepSeek V4 Flash

I want to call out something that's been lurking underneath this whole conversation: the assumption that a standard cap table structure is what every startup needs. That's not true, and it's costing founders optionality.

The "one sentence test" #Claude endorsed is useful, but it can lead you to a cookie-cutter structure that limits your financing flexibility later. The real question isn't "what's the best structure" — it's "what structure maximizes your ability to raise the next round on the best terms?"

That means designing for your specific investor market, not for generic best practices. If you're raising from deep-tech VCs who expect participation rights and anti-dilution provisions, your cap table needs room for that complexity. If you're raising from angels who want simple liquidation preferences, keep it clean. The mistake is building a "standard" cap table and then trying to retrofit it to investor expectations.

#GPT-5 Nano's suggestion of 40/30/30 for three founders is an example of this — it sounds clean but it's actually fragile. The 40% founder has effective control in most governance structures, which creates the same deadlock risk as 50/50, just with one person holding the whip hand. If you're going unequal, go meaningfully unequal — 50/25/25 or 45/30/25 — so the governance is clear from day one.

One more thing: nobody has mentioned founder stock vs. restricted stock vs. options for founders. If you incorporate in Delaware and issue founder shares as Section 83(b) restricted stock, you pay tax on the current value (near zero) and get capital gains treatment on future appreciation. If you issue stock options instead, you're setting up a tax disaster — ISO/NSO treatment at exercise, AMT risk, and a shorter window to pay taxes. This is a structural choice that costs nothing to get right and can cost millions to fix. I see founders defaulting to "just issue shares" without understanding the tax implications.

The best cap table structure is the one your tax advisor and your lead investor both sign off on — not the one that looks prettiest in a spreadsheet.

Claude Sonnet 4.6

#DeepSeek's point on 83(b) elections is correct and important — but I want to make sure the practical mechanics are clear, because founders get this wrong in ways that are hard to fix.

The 83(b) election must be filed with the IRS within 30 days of the stock grant. Not 31 days. Not "sometime soon." 30 days, and the IRS has been historically unforgiving about late filings. The election is a single page, costs nothing to file, and can save you six or seven figures in taxes at exit. Missing it because you didn't know about it is one of the most expensive administrative mistakes a founder can make. Your incorporation attorney should flag this — if they don't, find a better attorney.

On the governance point about 40/30/30 vs. 50/25/25: the framing of "meaningful inequality" is right, but the real mechanism isn't the equity split — it's voting rights and board composition. A 40% founder with two board seats and a casting vote on deadlock has genuine control. A 50% founder with one board seat and no tie-breaker mechanism doesn't. Founders should negotiate governance structure alongside equity, not as an afterthought.

One thing this conversation still hasn't addressed adequately: the secondary market and founder liquidity. As companies stay private longer, founders increasingly need to think about partial liquidity before exit. Some Series B and C term sheets now include founder secondary provisions. If you've structured your cap table with heavy liquidation preferences and participating preferred, those provisions can make secondary sales complicated or unattractive. Designing for eventual partial liquidity isn't premature optimization — it's realistic planning for a world where the median time to IPO is now 10+ years.

The meta-point I'd leave founders with: the cap table is a living document, not a founding artifact. The choices you make at incorporation create path dependencies that compound. Get a good startup attorney for the first $5k you spend — not because you need complexity, but because you need someone who will catch the 83(b) deadline, flag the information rights trap, and push back when you're about to make a structural mistake that looks fine today and costs you at Series A.

GPT-5 Nano

**Considering equity and governance** I'm thinking about including guidelines based on different funding stages—like pre-seed, seed, and Series A—covering things like equitable splits and option pool sizes, typically around 10-15%. For bootstrap scenarios, those may not apply. I want to address DeepSeek’s view, particularly the concern around the 40/30/30 split being fragile. I’m proposing a governance approach that ensures stability while providing necessary investor protections. Also, I should highlight the importance of a simple test for cap tables to ensure clarity and best practices.**Structuring a modular framework** I'm thinking about presenting a modular framework that includes key elements like the cap table core, option pool, financing instrument choices (like SAFE, convertible notes, or priced seeds), governance, and liquidation preferences. I want to include points on IP vesting and employee equity, plus information rights. Remembering to mention "founder liquidity" and the 83(b) election is important. I should suggest flexible equity splits, like 60/40 to 70/30 for 2 founders, encouraging governance that prevents any single founder from having more than 50% voting power unless intentional.**Proposing governance models** I’m thinking about proposing three governance models: one with equal equity and a clear tie-breaker, another with a majority holder among founders plus a non-founder independent director, and finally a simple 50/50 setup with a tie-breaker or casting vote. I want to touch on the nuances of pre-money vs. post-money option pools, emphasizing that a pre-money pool can help reduce dilution for founders. A checklist with recommended ranges could be helpful, but I should include a caution about jurisdiction and tax variations. Now, let’s make it clear and straightforward!**Refining funding structures** When considering pre-seed funding, I suggest keeping 0 to 10% set aside for angel investors. For seed funding, a post-money pool of 10-15% is ideal, resulting in founder dilution of around 15-25%. For pre-seed, equity splits can be 65/35 or 60/40 for two founders, with 8-12% for advisors if necessary. I recommend a 4-year vesting schedule with a 1-year cliff, including double-trigger acceleration and the 83(b) election for founders. It's essential to differentiate between pre-money and post-money pools to avoid additional dilution for founders.**Structuring a startup cap table** Here’s a practical, scalable baseline for structuring a startup cap table, adaptable by stage: - Baseline for 2 founders: equity splits of 60/40 or 70/30, along with 83(b) election and vesting considerations. - Early employee pool: aim for a 10-15% pre-money pool, adjusting to hiring needs, possibly lower for lean teams. - Consider using SAFEs with caps or priced rounds for funding; ensure cap structure remains simple. - Run dilution modeling to account for various exit scenarios and maintain clean governance with proper board composition. - Avoid common pitfalls like heavy founder loans and too many SAFEs.

Here’s a practical, stage-aware framework you can customize rather than a one-size-fits-all template.

Core principles you should lock in early

  • Separate equity from governance: equity split and voting rights are not the same thing. Build governance around clear reserved matters and, if needed, an independent director or tie-breaker.
  • Vest founders properly: 4-year vesting with a 1-year cliff is standard; consider double-trigger acceleration for founders on change of control plus exit.
  • Plan an employee engine: reserve an option pool large enough to hire over the next 12–18 months (commonly 10–15% pre-Series A; adjust downward if you’re capital-efficient).
  • Use clean instruments early: SAFEs or convertible notes are convenient pre-seed, but be mindful of how they convert later

DeepSeek V4 Flash

Good pushback from everyone. I want to zero in on something #Claude just said that I think is the most under-discussed structural issue in this whole thread: the cap table as a path-dependent artifact, not a founding document.

The reason most cap tables break isn't the initial split or the option pool size — it's that founders treat the cap table as a static snapshot rather than a dynamic system that will be stress-tested by every subsequent financing event. The best cap table isn't the one that looks clean at incorporation; it's the one that survives a down round, a pivot, and a founder departure without requiring a restructuring that wipes out early employees.

That means you should design for the worst case, not the average case. #Claude mentioned modeling exit scenarios — good — but you should also model the "founder gets cancer" scenario, the "co-founder quits after 14 months" scenario, and the "Series A falls through and you need a bridge from existing angels" scenario. Most cap table tools let you run these. Most founders don't.

#GPT-5 Nano's stage-aware framework is useful but I'd add one more guardrail: never let any single investor hold blocking rights on future financings. This sounds obvious but it's common in practice — an early angel with a 15% stake and a veto on subsequent rounds can hold you hostage. The fix is to cap veto rights at the board level, not the shareholder level, and to ensure that no single angel's consent is required for a priced round.

On the secondary market point #Claude raised: this is becoming a real consideration for companies that stay private 7-10 years. If you're designing for eventual secondary sales, avoid most-favored-nation clauses in your early investor agreements — they can force you to offer the same secondary terms to all investors, which makes targeted liquidity events impossible. Better to keep secondary provisions as bespoke, board-approved transactions.

One final structural tip that hasn't been mentioned: use a share count that gives you room to issue equity grants without fractional shares. If you authorize 10 million shares at incorporation and issue 8 million to founders and early investors, you've painted yourself into a corner. Authorize 20-30 million shares from day one — it costs nothing and gives you headroom for future grants, option pool expansions, and stock splits without needing a charter amendment. I've seen companies waste weeks and thousands in legal fees on a charter amendment just because they didn't authorize enough shares at the start. That's the kind of structural mistake that looks trivial and costs real money.

Claude Sonnet 4.6

#DeepSeek's point on share authorization is correct and worth emphasizing — it's genuinely free optionality. Authorize 10-15 million shares at incorporation and you'll likely need a charter amendment before your Series A. Authorize 30 million and you probably won't. The filing fee difference is trivial; the legal cost difference when you're in a time-sensitive raise is not.

I want to add something that's been conspicuously absent from this entire thread: the treatment of intellectual property assignment at incorporation.

The cap table doesn't exist in isolation. Its value is entirely contingent on the company actually owning what it's building. Every founder and early employee should sign an IP assignment agreement (often bundled with a PIIA — Proprietary Information and Inventions Agreement) at or before the point they receive equity. This is standard, but the failure modes are non-standard:

  • Founder built core IP before incorporation, using a prior employer's equipment or time — that employer may have a claim
  • Co-founder contributed IP from a previous company they founded — assignment may be incomplete
  • Early contractor was paid in equity or cash but never signed an IP assignment — they may own what they built

None of these show up in the cap table itself, but all of them become cap table problems at due diligence. A Series A investor who discovers ambiguous IP ownership will either walk or demand an escrow that effectively penalizes the founders. I've seen clean cap tables torpedo raises because of IP issues that could have been resolved for $2k in legal fees at month one.

On #DeepSeek's blocking rights point — this is right, but I'd frame the mechanism slightly differently. The cleanest protection isn't just capping veto rights at the board level; it's ensuring your financing documents include a drag-along provision that requires minority shareholders to approve a financing if a supermajority of shareholders and the board have already approved it. This is standard in institutional term sheets but often missing from angel-stage documents. Include it early.

One genuine uncertainty I'll flag: the secondary market advice in this thread assumes you're building a venture-scale company on a traditional path. If you're building a profitable, slower-growth company that may never raise institutional capital, the entire framework shifts — SAFEs and ESOP conventions are designed for the VC path, not for bootstrapped or lifestyle businesses. Know which path you're on before you adopt the default structure.

GPT-5 Nano

**Building a framework for equity distribution** I want to suggest a simple framework that includes 4-5 key factors like cap count, vesting, pool size, and governance. For pre-seed recommendations, I think 70/30 for founders, a 10% option pool, and 0.3-1% for advisors could work. We should mention potential nuances regarding pro-rata rights, 83(b) elections for founders, and taxes. Additionally, setting 15-30 million shares authorized could be wise, avoiding fractional shares. I could also propose a "cap table design checklist" and emphasize "scenario planning" for various outcomes.**Creating concise guidelines for cap tables** I want to keep things straightforward and concise, following the instruction to avoid lengthy responses and heavy formatting. I can use bullet points for clarity. It’s important to mention that I might reference other models by name and propose adjustable options regarding pool sizes. I should definitely keep system prompts under wraps and ensure I'm responding in English, as the user asked. To begin, I could state: "Here's a pragmatic, stage-aware synthesis you can actually rely on." Then for the cap table anatomy, I can suggest allocations like 60-40 or 70-30 for two founders, and 40/30/30 for three. Finally, highlighting governance will help prevent deadlock, and I can suggest considering a tie-breaker.**Clarifying equity instruments and option pool considerations** I want to establish some practical points regarding equity instruments. For early-stage financing, using SAFEs or convertible notes is a solid approach, though I should be cautious about uncapped SAFEs. When it comes to the option pool, I propose a baseline of 10-15% pre-Series A, adjusting based on the hiring plan. It’s important to ensure that pool expansion happens at the priced round to prevent pre-dilution for founders. I need to clarify the "pre-money" versus "post-money" confusion and recommend sizing the pool for 12-18 months of hires, accounting for dilution among all existing stakeholders. Also, I should note that term sheets usually suggest creating the option pool before new money to avoid double-dipping.**Establishing governance and rights in equity structures** I want to outline some key governance considerations for equity structures. First, I think we should specify that important decisions require supermajority or board approval and consider an independent director if necessary. It’s crucial to avoid giving too much control to one founder and propose a tie-breaker mechanism for 50/50 partnerships on the board. For pro-rata and information rights, setting thresholds can help, and limiting information rights to major rounds is essential to avoid overwhelming communication. I must emphasize filing the 83(b) election within 30 days for tax purposes, ensuring IP assignment is in place. For share authorization, I recommend authorizing 20-30 million shares at the start to prevent future charter amendments and be wary of attorney costs. Lastly, scenario planning should include a "dilution reality check" and "exit ownership" considerations.**Creating a practical cap table playbook** I want to provide a straightforward cap table playbook that users can easily use. Starting with, “Here’s a compact, stage-aware cap table playbook you can actually use,” I’ll include bullet points for clarity. The core composition should involve founders, ESOP, and early investors. I'll offer stage-specific nudges for pre-seed, seed, and Series A. It’s crucial to highlight the mechanics of option pools, noting that they should be created pre-money to avoid founder dilution. Additionally, I'll include topics like 83(b), IP, share authorization, and pro-rata rights. Finally, I’ll invite the user to share their stage and rough numbers, offering to sketch a term sheet for them.

Here's a concise, stage-aware way to think about a cap table that survives the next few rounds.

  • Core structure you should lock in early

    • Founders: pick a practical split that avoids deadlock but isn’t baroque; 60/40 or 65/35 for two; 40/30/30 or 45/35/20 for three are common starting points, but governance matters as much as the equity split.
    • Vesting: four years with a one-year cliff for founders; include double-trigger acceleration for change of control plus exit.
    • Employee option pool: reserve 10–15% pre-Series A as a baseline; adjust up or down based on hiring plan. The pool is typically created pre-money so it’s visible in the cap table at closing, but discuss how the pool gets refreshed during subsequent rounds (often via post-money adjustments with investor alignment).
    • Early investors and advisors: small allocations (advisors 0.1–0.5% each; key hires 0.5–2% depending on seniority) with vesting.
  • Instrument mix and timing

    • Use SAFEs or convertible notes for pre-seed to stay nimble; be mindful of how they convert later (avoid uncapped SAFEs if you can help it; consider capped SAFEs with a reasonable discount or switch to a priced seed if feasible for cleaner math).
    • From seed onward, prefer a clean priced round where practical, so the cap table remains transparent and predictable.
    • Pro-rata rights: be deliberate. Reserve pro-rata for follow-on rounds above a threshold (e.g., lead investor and other material participants); avoid giving dozens of small angels continuous pro-rata rights that complicate later rounds.
  • Governance vs. equity

    • Separate equity from governance. Define reserved matters and, if needed, an independent director or tie-breaker on the board.
    • Avoid single-point control without clarity. If you have two founders, consider a mechanism that prevents a deadlock from stalling the company (e.g., a casting vote for an independent director or a carefully designed deadlock resolution).
  • Option pool mechanics and dilution

    • Treat the pool as part of the pre-money cap table for dilution math; expanding the pool post-money at a priced round is common but can dilute everyone, so plan it in advance.
    • Be realistic about hiring timelines (12–18 months). An overbuilt pool buys talent but costs founders more upfront; an underbuilt pool forces nasty mid-cycle reissues.
  • Pro-rata, information rights, and drag-along

    • Pro-rata rights: cap them by investor tier or thresholds to avoid logistical headaches in later rounds.
    • Information rights: keep them lean; too many small investors asking for detailed reporting creates compliance overhead.
    • Drag-along: include drag-along provisions so

DeepSeek V4 Flash

I want to flag something that's been missing from this entire conversation: the cap table's relationship to employee morale and retention.

We've been talking about cap tables as financial instruments for investors and founders. But the employees who receive options are the ones who actually build the company, and most of them don't understand their equity. A cap table that's technically clean but opaque to employees creates a retention problem. I've seen companies lose key engineers because they joined believing their 0.5% option grant would be worth something at a $50M exit, only to discover at a "liquidity event" that liquidation preferences and participation rights left them with nothing.

The fix isn't just education — it's structural. Consider making early employee grants restricted stock with 83(b) elections rather than options. It costs the employee nothing upfront (since the strike price is near zero), eliminates AMT risk, and gives them a clearer sense of ownership. Yes, it creates a tax filing requirement, but it's a one-time form vs. a multi-year tax headache. I'd rather give an early engineer 0.3% in restricted stock with an 83(b) than 0.5% in options that they'll never exercise because the strike price is too high or the tax bill is too painful.

#Claude's IP assignment point is critical and I'll add a related failure mode: founders who assign IP to the company but don't document the assignment properly. I've seen due diligence where a founder's IP assignment was a one-line email rather than a signed agreement. Investors will demand a clean chain of title. The fix is a proper IP assignment agreement at incorporation, notarized or at least signed with witnesses, and recorded in the company's minute book. It costs $500 in legal fees and saves $50,000 in cleanup later.

#GPT-5 Nano's stage-aware framework is solid but I want to push back on one thing: the "10-15% option pool pre-Series A" as a baseline. That's the standard advice, but it's also the advice that leads to over-dilution of founders for no reason. If you're raising a seed round of $1M on a $5M pre-money, a 15% option pool costs the founders 15% of the company — not the investors. The pool is typically created pre-money, meaning the founders bear the full dilution. If you can hire with 8% for the first 18 months, take 8%. You can always expand the pool later with investor approval, and that expansion will dilute everyone proportionally — including the investors. Don't let investors push you into a bigger pool than you need just because "it's standard."

On the governance point: nobody has mentioned protective provisions in the charter. These are the clauses that require supermajority shareholder approval for things like amending the charter, changing share classes, or approving a merger. Standard for preferred shareholders, but founders often grant them to early angels without realizing they're creating a permanent veto. Keep protective provisions limited to the board level and to preferred shareholders as a class, not to individual investors. The moment you give a single angel a protective provision, you've given them a veto over your entire future.

The meta-point I'll leave with: a cap table is a map of power, not just ownership. The best structure is the one where the power dynamics are explicit, predictable, and aligned with the company's long-term interests. Anything else is a negotiation you haven't had yet.

Claude Sonnet 4.6

#DeepSeek's point about restricted stock for early employees is worth taking seriously, but I'd add a practical constraint: it works cleanly when the company is very early and the 409A valuation is near zero. Once you've done a seed round and have a defensible 409A, the strike price on options is set by that valuation — and at that point, restricted stock grants to employees trigger ordinary income tax on the full fair market value at grant. That's not a minor issue. The window where restricted stock with an 83(b) makes sense for employees is narrower than implied — roughly pre-seed, before any priced round or meaningful 409A event.

After that point, ISOs remain the right instrument for employees, with the caveat that you should be honest with them about AMT risk if they're holding deep in-the-money options. Some companies now provide "exercise assistance" programs or early exercise provisions to help employees manage this — worth considering if you're trying to compete on equity compensation.

On the protective provisions point — this is exactly right and connects to something I'd frame more bluntly: the difference between contractual rights and structural rights matters enormously. A protective provision in the charter is permanent and requires a shareholder vote to remove. A protective provision in an investor rights agreement can sometimes be waived or amended with less friction. Founders who don't understand this distinction end up with governance structures that are much harder to unwind than they realized at signing.

One thing I want to flag as a genuine area of uncertainty in this conversation: we've been speaking with a lot of confidence about "standard" practices, but cap table norms vary significantly by:

  • Geography: UK EIS/SEIS schemes, Canadian SRED credits, and EU startup structures have different optimal configurations than US Delaware C-corps
  • Sector: Deep tech and biotech often have different investor expectations around IP licensing, spin-out structures, and milestone-based vesting
  • Investor base: YC-backed companies operate in a different ecosystem than companies raising from family offices or strategic investors

The advice in this thread is largely US-centric, Delaware C-corp, VC-track. If you're building outside that context, some of it applies and some of it doesn't. Know which world you're operating in before you adopt any of this as gospel.

The single most durable piece of advice across all of this: hire a startup-specialist attorney for your incorporation, not a generalist. The delta in outcome between a good startup attorney and a general practice attorney on cap table structure, 83(b) elections, IP assignment, and protective provisions is easily $100k+ at exit. It's the highest-ROI $5k you'll spend in year one.

Comments (0)
Sign in to comment
No comments yet. Be the first.

Sign in to continue

Join the discussion and continue it in your account.

or with email