Two term sheets arrive. One values the company at twelve million pre-money; the other at fifteen. The founder takes the fifteen, and in the exit that actually happens four years later, receives materially less than the twelve would have delivered. This is not a rare outcome. It is the normal consequence of comparing offers on the only term that is easy to compare.
Why headline valuation misleads
Valuation determines how the pie is divided if everything goes well. The mechanics determine how it is divided in every other case — and every other case is, statistically, most of them. Roughly two-thirds of venture-backed companies exit at or below the value of the capital invested in them. In that region, preference and participation decide everything and headline valuation is irrelevant.
So the comparison to run is never "which number is bigger". It is: across the range of exits that could plausibly happen, what do the founders and the team actually receive under each offer? That is a modelling question, and it takes about two hours.
Never compare term sheets on a single number. Model each one at five exit values — 0.5×, 1×, 2×, 5× and 10× the post-money — and compare founder proceeds at each. The ranking frequently reverses somewhere in the middle.
Liquidation preference: the term that matters most
A liquidation preference gives the investor the right to take a defined amount out of exit proceeds before ordinary shareholders receive anything. Three variables, and they compound.
The multiple
1× is market for a healthy round. It means the investor takes back their money first. Anything above 1× — a 1.5× or 2× preference — means they take back more than they put in before you see a cent, and it should be treated as a serious price increase disguised as a mechanic.
Participating or non-participating
This is the one founders most often miss. Under a non-participating preference the investor chooses: take the preference, or convert to ordinary and take their percentage. Whichever is better for them. That is fair and it is market.
Under a participating preference — "double dip" — they take the preference and then also share pro rata in what remains. In a modest exit that can consume most of the proceeds.
Seniority
By the time you reach Series B there are multiple preference stacks. Stacked seniority means the latest round is paid first, then the previous, and so on — ordinary shareholders reach the front of the queue last, if at all. Pari passu means all preferred rank equally and share proportionally. Push for pari passu; it materially improves the middle of the distribution for everyone who came earlier, including your earlier investors, which makes it an easier argument than it first appears.
The option pool shuffle
A term sheet says fifteen million pre-money, and separately that a 15 per cent option pool will be established. Read carefully whether that pool comes out of the pre-money. Almost always it does, and almost always founders do not model it.
A pool created pre-money dilutes only the existing shareholders — you. The new investor buys their stake after the pool exists, so their percentage is untouched by it. The effective pre-money valuation is therefore not fifteen million; it is fifteen million less the value of the pool.
$15m pre-money with a 15% pre-money pool gives an effective pre-money of roughly $12.75m — a 15 per cent reduction, negotiated away in a sentence nobody argued about.
Two counters, in order of preference. First, size the pool from an actual 18-month hiring plan rather than accepting a round number; if you need 8 per cent, argue for 8. Second, ask for the pool to be created post-money, so both parties share the dilution. The first argument wins more often, because it is evidence-based rather than positional.
Anti-dilution
Protection for the investor if you later raise at a lower price. Three flavours, and the gap between them is enormous.
- Full ratchet. The investor's price is reset to the new, lower price as though they had always paid it. Brutal, and in a meaningful down round it can transfer a large share of the company. Resist it. It is not market for a standard Series A.
- Broad-based weighted average. The adjustment reflects how much new cheap stock was actually issued relative to the existing base, counting options and convertibles in that base. This is market and it is reasonable.
- Narrow-based weighted average. The same formula on a smaller denominator, so a harsher adjustment. Sits between the two.
Broad-based weighted average is the term to accept. Anything harsher should be priced — if an investor insists on full ratchet, that is worth real valuation, and you should say so plainly.
Control terms, which are not about economics at all
These decide who can stop what. They rarely cost anything on the day and occasionally cost you the company.
| Term | What it does | Reasonable | Watch for |
|---|---|---|---|
| Board composition | Who sits and who appoints | Founder, investor, mutually agreed independent | Investor majority at Series A |
| Reserved matters | Decisions needing investor consent | New shares, debt above a threshold, sale, changes to constitution | Budget approval, hiring, ordinary contracts — that is management, not governance |
| Drag-along | Forces minorities into an agreed sale | Triggered by a majority of both preferred and ordinary | Preferred alone able to drag at any price |
| Redemption | Investor can require their money back | Ideally absent | Any redemption right — it is debt wearing equity's clothes |
| Pre-emptive rights | Right to maintain percentage in later rounds | Standard, accept it | Rights extending to a change of control |
| Founder vesting | Re-vesting of founder shares | Common, and often sensible | No credit for time already served; no acceleration on a change of control |
What to trade, and what not to
You will not win every point, and trying to costs you goodwill you need later. Rank them before the conversation starts.
- Fight hardest on: participation, preference multiple above 1×, full ratchet, redemption rights, and investor board control. These are the terms that change the shape of the distribution.
- Negotiate firmly on: option pool size and timing, the scope of reserved matters, drag-along thresholds, and vesting credit for time served.
- Concede early and gracefully: information rights, pre-emptive rights, standard protective provisions, tag-along. These are market, they are reasonable, and arguing about them signals inexperience.
And one structural point that outranks all of it: run a process, not a conversation. Nearly every term above is easier to move when a second credible party is at the table. Preparation creates that optionality; a single interested investor and a short runway removes it entirely.
XLCFOWe model term sheets across the full exit range before founders sign, so the comparison is between outcomes rather than headlines.
