Not all terms matter equally. These are the ones where a 'bad' answer compounds the most across future rounds:
1. Valuation (pre-money / post-money). Obvious one. But pre-money valuation × dilution is what determines founder ownership, not the headline number. A $20M pre-money with a 20% post-money option pool top-up dilutes founders the same as a $16M pre-money with no top-up.
2. Option pool top-up. The investor wants the option pool (usually 10-15% post-money) to come from pre-money — diluting existing holders (founders) but not the new investor. This is often negotiable. Push for a smaller pool based on a real hiring plan, or for the pool to be diluted across both pre- and post-money.
3. Liquidation preference. 1x non-participating is standard. 1x participating or >1x preferences flip the economics of small/medium exits and can leave founders with $0 on a $30M sale.
4. Anti-dilution. Weighted-average is standard. Full-ratchet is aggressive and rarely justified — push back hard.
5. Board composition. At seed: 1 founder seat, 1 investor seat, 1 mutually-agreed independent. Anything that gives investors a board majority pre-Series A is aggressive.
6. Protective provisions. Investor consent for: sale of the company, new share class, debt above threshold, change of business model. Standard set is reasonable; expanded sets (consent for hiring senior execs, opening new offices, etc.) are over-reach.
Terms that don't move the needle much: drag-along, tag-along, ROFR, information rights, pro-rata. These have market norms; deviate only if you have specific reason.