How to Read a Term Sheet
The clauses that determine what you keep. Liquidation preference, anti-dilution, board control, protective provisions and drag-along - explained with worked examples.
Why this document matters more than the valuation
Founders negotiate hard on valuation and sign the rest. That is backwards.
A high valuation with a 2× participating liquidation preference can leave you with less than a lower valuation with clean 1× non-participating terms. The economics live in the clauses, and a term sheet is where they are set - because whatever is agreed here flows into the binding documents almost unchanged.
Read it properly, and get a lawyer who has seen a hundred of them.
Economic terms
Valuation and the option pool
The pre-money valuation, the amount invested, and the resulting ownership. Two things to check.
Pre-money or post-money. ₹5 crore at ₹20 crore pre-money gives the investor 20%. At ₹20 crore post-money it gives them 25%.
Where the option pool comes from. If the term sheet requires a 15% ESOP pool "on a post-money fully diluted basis", check whether it is created before or after the investment. Created pre-money, existing shareholders absorb the entire dilution. This is standard practice and it is negotiable - both the size and, occasionally, the allocation.
Liquidation preference
The most economically significant clause in the document. It determines the order and amount of payment in an exit.
1× non-participating - the market standard. Investors receive either their money back, or their pro-rata share as if converted to common, whichever is greater. Not both.
Participating preferred - investors receive their money back and their pro-rata share of the remainder. Sometimes called "double dipping".
Multiple preference - 2×, 3×. Investors receive a multiple of their investment before anyone else is paid.
Worked example. An investor puts ₹20 crore in for 20%, and the company sells for ₹100 crore.
| Structure | Investor receives | Everyone else |
|---|---|---|
| 1× non-participating | ₹20 Cr, or 20% = ₹20 Cr → ₹20 Cr | ₹80 Cr |
| 1× participating | ₹20 Cr + 20% of ₹80 Cr = ₹36 Cr | ₹64 Cr |
| 2× non-participating | ₹40 Cr, or ₹20 Cr → ₹40 Cr | ₹60 Cr |
| 2× participating | ₹40 Cr + 20% of ₹60 Cr = ₹52 Cr | ₹48 Cr |
Same valuation, same ownership percentage, ₹32 crore of difference to the founders and employees.
Now the same structures on a ₹40 crore exit:
| Structure | Investor receives | Everyone else |
|---|---|---|
| 1× non-participating | ₹20 Cr | ₹20 Cr |
| 1× participating | ₹20 Cr + 20% of ₹20 Cr = ₹24 Cr | ₹16 Cr |
| 2× participating | ₹40 Cr | ₹0 |
At a modest exit, a 2× participating preference can take everything. This is why the preference stack matters so much and why it should be modelled at several exit values, not one optimistic one.
Anti-dilution
Protects investors if a future round prices lower than theirs.
Broad-based weighted average - the standard. Adjusts the conversion price based on the size and price of the new round. Mild, and reasonable.
Narrow-based weighted average - same mechanism, smaller denominator, so a larger adjustment.
Full ratchet - the conversion price resets entirely to the new, lower price, regardless of how small the new round is. Punitive, and it can transfer a large share of the company on a small down round. Resist this.
Dividends
Usually non-cumulative and rarely paid in venture deals. Cumulative dividends accrue and are added to the preference amount at exit, which quietly increases what investors take. Check which.
Pay-to-play
If a future round happens, existing investors must participate pro-rata or lose some rights, typically converting to common. Arguably founder-friendly, since it encourages continued support.
Control terms
Board composition
Who sits on the board and who appoints them. A typical Series A board is five seats: two founders, one investor, one or two independents appointed by mutual agreement.
The question to ask is not "do investors have a seat" - of course they do - but who controls the board in a disagreement. Count the votes.
Protective provisions
A list of actions requiring investor consent regardless of board or shareholder majorities. Standard scope:
- Selling the company or a material asset
- Raising further capital or issuing new shares
- Changing the share capital or creating a senior class
- Taking on debt above a threshold
- Changing the board size
- Amending the constitutional documents
- Paying dividends
- Winding up
These are normal. The negotiation is about scope and thresholds - a debt consent threshold of ₹5 lakh will make routine operations painful; ₹2 crore probably will not. Push for thresholds that leave you able to run the business.
Founder vesting
Investors will require it, even on shares you have held for two years. Common outcome: some credit for time served, with the balance vesting over the next three to four years.
Negotiate double-trigger acceleration - vesting accelerates only if the company is acquired and you are terminated. Single trigger on acquisition alone is harder to get and less commonly accepted.
Drag-along
If a defined majority approves a sale, remaining shareholders must go along with it. Reasonable in principle - it prevents a small holder blocking an exit everyone wants. Check the threshold and whether founders are part of the approving majority.
Information rights
What you must report and how often. Agree to something you can actually sustain: monthly management accounts and a quarterly board pack is typical. Committing to weekly reporting you will not deliver starts the relationship badly.
Other terms
Right of first refusal and co-sale. If a founder sells shares, the company and investors get first refusal, and investors can join the sale pro-rata. Standard.
Registration rights. Relevant only at IPO. Rarely negotiated hard.
Exclusivity / no-shop. You cannot talk to other investors for a defined period, usually 30–60 days. This is binding. Keep it short - 30 days is reasonable, 90 is not - because it removes your leverage entirely for the duration.
Expenses. Whose legal costs the company pays, usually with a cap. Make sure there is a cap.
Conditions precedent. What must happen before the money arrives. Read these carefully; an open-ended condition is a way out.
What to focus on
If you can only push on a few things:
- Liquidation preference - 1× non-participating, nothing more
- Anti-dilution - broad-based weighted average, never full ratchet
- Board control - count the votes
- Protective provision thresholds - set high enough to operate
- Option pool size - sized to your real hiring plan
- No-shop duration - 30 days
- Double-trigger acceleration on founder vesting
And model the preference stack at three exit values: a disappointing one, a decent one, and a good one. The disappointing one is where the terms actually bite, and it is the scenario nobody wants to model.
The one-line summary
The valuation is the headline; the term sheet is the substance. Liquidation preference determines what you keep in every outcome except the best one, anti-dilution determines what a down round costs you, and protective provisions determine what you can do without asking. Model a modest exit before you sign - that is the scenario the clauses were written for.
Frequently asked questions
- What is a liquidation preference?
- It determines who gets paid first in an exit and how much. A 1× non-participating preference means investors get their money back before common shareholders receive anything, or convert to common if that pays more. Multiples above 1×, or participating preferred, take substantially more.
- What are protective provisions?
- A list of actions the company cannot take without investor consent - typically raising further capital, selling the company, changing share capital, taking on significant debt, or altering the board. Standard in any priced round; the negotiation is about scope and thresholds.
- Is a term sheet legally binding?
- Mostly not. Term sheets are generally non-binding on the commercial terms, with exceptions usually for confidentiality, exclusivity and expenses. But the terms agreed here flow into the binding documents almost unchanged, so this is where the negotiation actually happens.
