The VC term sheet clauses that change your deal

A VC term sheet is short because it records the bargain, not because the bargain is simple. The clauses that deserve your attention change one of three things: how much of the company you own, who gets paid in a sale, or who can stop a decision. Dense pages about information rights and registration can look alarming while a quiet phrase about the option pool or a separate series vote moves far more value and control.

Treat the document as a model of future outcomes. Run the ownership math at closing, the payout math at several sale prices, and the approval path for decisions you expect to make. Then negotiate the terms that produce an outcome you cannot live with. A friendly investor and a high headline valuation do not repair bad mechanics.

This is a US venture financing primer, not legal advice. Company counsel should translate the agreed term sheet into definitive documents and tell you how Delaware law, your charter, existing SAFEs or notes, and earlier investor rights affect your specific deal. You still need to understand the business bargain yourself.

Valuation means little until you calculate the ownership

Valuation tells you the price per share, but the definition of the capitalization used in that calculation determines what you actually sell. A term sheet usually states a pre-money valuation, the amount invested, and sometimes a post-money valuation. In the simplest case, post-money valuation equals pre-money valuation plus the new cash. A $2 million investment at an $8 million pre-money valuation produces a $10 million post-money valuation, so the new investor owns 20% immediately after closing.

That clean calculation assumes everyone agrees on what sits inside the pre-money capitalization. The phrase "fully diluted capitalization" can include issued common stock, issued preferred stock on an as-converted basis, granted options, the ungranted option reserve, warrants, and shares that SAFEs or notes will receive. The treatment of converting instruments matters because putting them in the pre-money side makes existing holders absorb their dilution. Post-money SAFEs were designed in part to make ownership sold by each SAFE easier to calculate, but a priced round can still contain conversion details that deserve an updated cap table from counsel.

Ask for a capitalization table that shows share counts, not percentages alone, immediately before and after the financing. It should identify every security, its conversion assumption, the new preferred shares, and the option reserve. Then ask for the same table on an as-converted and fully diluted basis. If the spreadsheet cannot tie every percentage to a share count and a document, the negotiation is happening on an unreliable model.

The arithmetic also exposes an easy mistake: comparing two offers by pre-money valuation alone. An offer at $12 million pre-money with a large pre-closing option pool increase can leave founders with less ownership than a $10 million offer that sizes the pool to the hiring plan. Compare the founder and employee percentage after all conversions, the pool increase, and the new money.

Price still matters. A small difference in dilution compounds across later rounds, and ownership affects voting as well as exit proceeds. But "highest valuation wins" is bad advice because valuation can hide a larger pool, harsher preference, or control terms that block the next financing.

The option pool can move the price without changing the headline

A pre-closing option pool increase shifts dilution to existing holders while preserving the new investor's negotiated percentage. Investors often ask for enough ungranted equity to cover hiring until the next round. That request is reasonable in principle. The argument is about the hiring plan, the pool's current size, and whether the increase happens before or after the investment.

Use the $2 million on an $8 million pre-money example. With no other changes, the investor owns 20% after closing. If the term sheet also requires a new pool equal to 15% of the company after closing and puts the entire increase into the pre-money capitalization, the old holders end at 65%, the investor still gets 20%, and the new pool gets 15%. The valuation headline did not change, but the old holders gave up another 15 percentage points.

Do not negotiate the pool as an abstract market number. Build a role-by-role hiring forecast through the next expected financing, assign realistic grant ranges, subtract the unused options already available, and add a modest buffer. If the investor assumes eighteen months of hiring while the cash plan supports twelve, fix the operating plan and the pool together.

Push back on a pool sized by a round percentage with no hiring model. Also resist language that lets the investor recalculate the pool after diligence without reopening valuation. A sensible fallback is a stated share number or stated post-closing percentage tied to an agreed hiring plan. This is one of the few places where a single spreadsheet row can quietly cost more than pages of legal language.

Liquidation preference decides who gets paid in an ordinary exit

Liquidation preference sets the payout order when the company is sold, merged, wound down, or otherwise enters a defined liquidation event. It matters most when the exit value is near or below the money invested, which is exactly where founders tend to stop modeling because the scenario feels discouraging. Model it anyway.

A 1x non-participating preference gives an investor a choice: take back the original investment before common stock receives proceeds, or convert to common and take the ownership percentage. Return to the investor who put in $2 million for 20%. At a $6 million sale, the investor takes the $2 million preference because 20% of the sale would be $1.2 million; common holders divide the remaining $4 million. At a $20 million sale, the investor converts and takes $4 million because that beats the $2 million preference.

Participating preferred changes the first result and many outcomes above it. The investor first receives the preference and then participates with common in the remaining proceeds. At a $20 million sale, a 1x participating investor takes $2 million first and then 20% of the remaining $18 million, for $5.6 million total. A participation cap limits that double recovery, but you must model where the cap applies and when conversion pays more.

A multiple magnifies the downside. With a 2x preference on the same $2 million investment, the investor takes $4 million before common receives anything. In a $6 million sale, founders and employees split only $2 million. Multiple preferences can stack across rounds, and the order can be senior, pari passu, or tiered. Senior means one series gets paid before another. Pari passu means the covered series share the available preference pool according to the document's formula.

Wilson Sonsini's guidance calls 1x non-participating preference standard for a healthy US venture deal and warns that participation or a multiple above 1x deserves special attention. That is sound advice, but "1x non-participating" is not enough to finish the review. Check whether dividends add to the preference, whether the preference uses the original purchase price or a growing amount, which transactions count as a liquidation, whether different series rank together, and whether investors can block an exit before the waterfall even applies.

Negotiate hard against participating preferred, a preference above 1x, cumulative dividends, and senior stacking that pushes common far out of the payout. If your bargaining position is limited, trade one economic protection for another instead of accepting all of them. A 1x non-participating preference with broad-based weighted average anti-dilution is materially cleaner than a package containing a 2x preference, participation, and full ratchet protection.

Board composition determines who manages the company

Board composition allocates management authority, including approval over budgets, executive hiring and firing, option grants, major contracts, financing proposals, and a sale process. Stock ownership and board control are related, but they are not the same. A founder can own the largest block and still lose a board vote. An investor can own a minority and hold one seat without controlling the board.

A common early priced-round structure has three seats: one chosen by common holders, one chosen by the new preferred investors, and one independent chosen by agreement. The labels do less work than the appointment and removal language. Ask who can nominate each director, who elects and removes that person, what happens when a seat is vacant, and whether the independent must receive approval from both sides.

The independent seat is often presented as neutral. In practice, it can become the deciding vote on replacing a CEO, accepting a financing, or selling the company. Do not leave the seat empty indefinitely and do not agree that one side may fill it alone after a deadline. Define a mutual selection process, a useful experience profile, and what happens if the parties cannot agree. The person matters more than a polished biography; references from founders who disagreed with that director tell you more than references from easy years.

Keep the early board small enough to work. An odd number reduces deadlock risk, but adding an independent merely to create an odd number can shift control before you have chosen the person. Observer rights also deserve boundaries. An observer usually does not vote, yet receives materials and sits in discussions. The company should be able to exclude observers for privilege, conflicts, competitive sensitivity, or duties owed to others.

Negotiate board composition when the term sheet arrives, not after counsel starts drafting. Removal rights, vacancy rules, committee control, and founder-seat conditions can turn "one investor seat" into a different deal. Pay particular attention to a founder seat that disappears automatically if the founder changes roles or holds less than a stated percentage. Sometimes that condition is fair; sometimes it lets a financing or termination erase the only founder voice in the room.

Protective provisions are vetoes, not board seats

Protective provisions give preferred stockholders a separate consent right over specified corporate actions, even when the board and a majority of all shares support the action. They protect the investment against structural changes. They should not require investor permission for routine operations.

Typical provisions cover amending the charter in a way that harms preferred rights, creating stock senior to or on parity with the preferred, changing the authorized number of shares, paying dividends, redeeming stock, changing board size, taking on debt above a threshold, or completing a sale or liquidation. Orrick's preferred stock term-sheet checklist places voting rights and protective provisions beside board composition because each appears in the deal, but they operate through different approval channels. Confusing them leaves founders blind to a second layer of control.

Read the consent threshold and the voting group before reading the action list. "Approval of the preferred" might mean a majority of all preferred voting together. It might give the new series its own vote. A separate series veto lets a small holder block a later round even when every other investor supports it. Ask whether the right belongs to a defined investor majority, all preferred as one class, or each series separately.

Thresholds need numbers and exceptions. A debt veto should sit above ordinary credit cards, equipment leases, and already approved budget items. A hiring veto should not cover normal employees. A sale veto often belongs in the package, but it must fit the drag-along provision and the board vote so one small group cannot freeze a reasonable acquisition forever.

Negotiate the scope, the dollar thresholds, and when the rights end. Rights can terminate at an IPO, when the series falls below a stated ownership level, or when a particular investor no longer holds enough shares. Avoid consent rights attached permanently to a class that may later contain a tiny residual holder. The legitimate purpose is protection against exceptional acts, not a permanent permission slip for running the company.

Pro rata rights reserve a seat in the next round

A pro rata right lets an investor buy enough securities in a later financing to maintain an ownership percentage. It does not give the investor free shares and it does not prevent dilution automatically. The investor must receive notice, exercise on time, and invest more money on the new round's terms.

Founders often dismiss pro rata as an investor-side detail. The cost appears in the next fundraise. If existing investors can take a large part of a small round, a new lead may not get the ownership it wants. You may then increase the round, reduce another investor's allocation, or ask insiders to waive rights under time pressure. The problem grows when side letters grant rights that the main cap table does not display.

Limit the right to major investors or a stated ownership threshold, require a prompt election, and let unused allocations return to the company. Check whether the right covers equity financings only or also SAFEs, notes, strategic issuances, equipment transactions, and employee grants. Customary exclusions keep ordinary compensation and commercial deals from triggering an investment process.

Super pro rata rights let an investor buy more than needed to maintain its percentage. They may help when an insider has committed to support the next round, but they can crowd out a new lead and give the holder a cheap option on your momentum. Do not grant them casually in a side letter. If you accept one, cap the extra allocation, give the company discretion over the remainder, and set an expiration.

The negotiation is less about whether serious investors receive participation rights and more about who qualifies, which issuances count, how much notice you owe, whether the right transfers, and when it ends. Put every such right into the fully diluted financing model so you know how much of the next round is already spoken for.

Anti-dilution protects price, not a fixed ownership percentage

Anti-dilution adjusts the preferred stock's conversion price after a later financing below the earlier round's price. It compensates the earlier investor with more common shares on conversion. It does not maintain the investor's exact percentage, and it does not protect founders or employees from the ordinary dilution caused by issuing new shares.

Broad-based weighted average protection accounts for both the lower price and the size of the down round. A common formula is:

CP2 = CP1 * (A + B) / (A + C)

CP1 is the old conversion price. A is the agreed fully diluted share count before the new issue. B is the consideration raised in the down round divided by CP1, and C is the number of new shares issued. Definitions vary, so counsel must confirm the charter language rather than copying the symbols from a blog post.

Suppose CP1 is $1, A is 10 million shares, and the company sells 2 million new shares at $0.50, raising $1 million. B equals 1 million shares and C equals 2 million. CP2 becomes about $0.9167. Each old preferred share then converts into about 1.091 common shares instead of one. The adjustment hurts common holders, but it reflects both the discount and the relatively limited size of the down round.

Full ratchet protection resets CP2 to the new $0.50 price regardless of how few shares the company sells. In this example, every old preferred share converts into two common shares. That can punish founders and employees far beyond the capital raised and make a rescue financing harder to assemble. Full ratchet is worth a hard no in an otherwise healthy financing. Narrow-based weighted average sits between the two because its smaller definition of A produces a larger adjustment.

Review the exclusions as carefully as the formula. Employee equity within an approved pool, stock splits, acquisition consideration, equipment financing, and securities issued under existing convertibles often receive exemptions. A pay-to-play term can require investors to participate in the down round to keep anti-dilution or other preferred rights. That can align support in a difficult financing, though its exact penalty needs careful drafting.

The NVCA model legal documents present alternatives and explanatory notes rather than a universal answer. Use them as a comparison point, not as proof that every bracketed option is acceptable. Broad-based weighted average protection with sensible exclusions is familiar. Full ratchet, a narrow denominator, or protection triggered by ordinary employee grants changes the economic bargain and deserves negotiation.

Some frightening clauses are mostly plumbing

Long clauses about information, registration, transfer, and documentation often cost less than one vague sentence about control. They still need review, but founders should spend negotiation time according to consequence rather than typography.

Information rights usually require periodic financial statements, budgets, and inspection access for major investors. Agree on reports the company can actually produce, reasonable delivery periods, confidentiality duties, and an exclusion for privileged or competitively sensitive material. A promise to deliver audited monthly financials shortly after month end is not harmless if the company has neither an audit nor a finance team.

Registration rights concern a future public offering and resale process. They occupy many pages in definitive documents because securities law mechanics are detailed. For an early company, they rarely change tomorrow's operating control. Counsel should compare them with the current NVCA forms and flag unusual demand rights, expenses, penalties, or survival, but founders rarely gain much by rewriting customary language line by line at term-sheet stage.

Rights of first refusal and co-sale rights restrict transfers of founder shares and let investors participate in certain founder sales. Drag-along provisions can force holders to support a company sale after specified approvals. The scary verb "drag" is not a reason to delete the provision; the approval formula is the issue. Require a board vote and sensible stockholder approvals, check whether preferred holders get a separate say, and make sure dragged holders receive the same form of consideration subject to their preference rights and do not give broader representations or liability than appropriate.

Term sheets also contain the clauses most likely to bind before closing: exclusivity or no-shop, confidentiality, expenses, access, and governing law. A term sheet may call most deal terms nonbinding while making these provisions binding. Negotiate the no-shop period to match a realistic diligence and document schedule, require the investor to move promptly, preserve the board's legal duties, and cap company-paid investor counsel fees. A routine economic clause in a nonbinding summary may wait for definitive drafting. A binding no-shop starts restricting the company when you sign.

Founder vesting or "revesting" looks administrative but may reset equity already earned. Tie any new vesting to the actual retention concern, credit time served, and negotiate acceleration for a qualifying termination around a sale. This clause belongs in the consequential pile because it affects whether a founder keeps shares after losing her role.

Negotiate the downside case, not the legal vocabulary

A good negotiation memo converts each disputed phrase into a financial or decision outcome. It also gives counsel clear instructions.

Compare offers on one normalized cap table rather than accepting each investor spreadsheet as its own reality. Use the same closing date, financing amount, conversion assumptions, hiring pool, and outstanding securities for every offer. Add the legal-fee cap and any debt that must be repaid from the round so the net cash figure is honest. Then place the board and consent maps beside the economics. An offer that leaves two more ownership points with the founders may still be worse if a separate series can block fundraising or the independent seat is not mutually chosen.

Record the requested change and a fallback before the negotiation call. For example, ask to remove participation, then decide whether capped participation is tolerable and at what cap. Ask for one class-wide preferred vote, then decide whether a series vote that expires below a stated holding threshold works. This preparation stops a live conversation from turning an undefined concession into agreed language. Before responding to the investor, make a five-line issues list like this and replace each example with the actual term-sheet language:

1. For valuation and the option pool, model the case where the pool grows before closing. Request a pool sized from the hiring plan; a fixed post-closing percentage is the fallback.
2. For liquidation preference, model a modest sale in which common receives little. Request 1x non-participating preferred; capped participation is the fallback.
3. For the board, model a deciding director who supports removing the CEO. Request mutual selection of the independent; founder approval of the first independent is the fallback.
4. For protective provisions, model a small series blocking a needed financing. Request one preferred-class vote; an ownership threshold and sunset are the fallback.
5. For anti-dilution, model a rescue round below the old share price. Request broad-based weighted average protection; sensible exclusions and pay-to-play are the fallback.

Run at least four exit values through a waterfall, including a sale below total invested capital, near total invested capital, around the preference conversion point, and a strong outcome. Run a flat round and a down round through the cap table. Write who approves the annual budget, a new debt facility, the next financing, a CEO change, and a sale. These exercises turn abstract rights into decisions.

Prioritize irreversible economics and control. I would usually negotiate valuation and pool together, remove preference participation or multiples, settle board composition, narrow separate series vetoes, and reject full ratchet before spending time on customary registration language. The order can change if one clause has an unusual trigger or if an existing investor already holds rights that the new term sheet expands.

Do not negotiate every line to prove sophistication. Investors notice when a founder spends an hour on a standard notice period and misses a senior 2x participating preference. Ask the lead which terms are firm, explain the concrete company problem with the disputed language, and propose wording or an economic trade. Keep a written issues list so concessions do not disappear between calls.

Get company counsel who regularly closes venture financings, and involve counsel before signing because the binding clauses begin then and the business terms become hard to reopen. Investor counsel does not represent the company. A mentor can help you judge behavior and market context, but she should not substitute for your lawyer's document review.

Sisters gives women founders a place to ask peers who have already handled fundraising, get candid feedback on a plan or deck, find advisors, and seek warm investor introductions. That kind of pattern recognition helps you decide where to press, while your cap table and counsel tell you what the words do in your company.

The signature decision should come down to a finite set of outputs: an agreed fully diluted ownership table, exit waterfalls, a board map, a consent map, and a list of binding pre-closing duties. If the investor will not let you model a term, define its trigger, or record the answer, the ambiguity is part of the offer. Price it accordingly or decline it.