Term sheet red flags in 2026: what to negotiate
Most seed term sheets in 2026 are clean. The few that aren't can cost you millions at exit. Here are the seven clauses to flag.
Term sheet red flags in 2026: what to negotiate before you sign
Most seed term sheets in 2026 are boringly clean: 1x non-participating preferred, broad-based weighted-average anti-dilution, no funny board structure. The seven clauses below are the ones that aren't. Each looks negotiable on the page and costs real dollars at exit. Push back on the wording before you sign, not after the cap table is locked.
Most seed term sheets in 2026 don't have term sheet red flags. The 2026 dataset says quietly that the standard deal is genuinely founder-friendly: 95% of Q2 2024 priced rounds had a 1x liquidation preference, 94% had non-participating preferred, and broad-based weighted-average anti-dilution showed up in 95–100% of deals. The danger is the 5–8% of sheets that aren't standard, where one clause buried on page three quietly transfers seven figures from founders to investors at exit. This is the list of seven, what's actually standard, what's negotiable, and the language to push back with.
The 7 term sheet red flags to flag in 2026
- Participating preferred liquidation preference. Investor takes their money back AND shares pro-rata in what's left. On a $50M exit after a $10M round, that's roughly $4M shifted from common to preferred. Push to 1x non-participating; it's in 94% of deals.
- Liquidation preference above 1x. A 1.5x or 2x preference means the investor must clear 1.5x or 2x their check before founders see anything. Carta reported 1x-or-higher preferences in only 8% of all Q1 2024 rounds; above-1x is not a seed-stage norm.
- Full-ratchet anti-dilution. Repriced all investor shares to the lowest future price, regardless of round size. Catastrophic in a down round. The market standard is broad-based weighted-average; full-ratchet is a hard no.
- Investor board control before Series A. A 2-1 board (one founder, two investors) at seed locks you out of every major decision. Standard at seed is founder-majority or a 3-person board with one independent.
- Redemption rights. Investor can force you to buy back their shares after 5–7 years if there's no exit. Adds existential pressure exactly when you don't need it. Strike entirely at seed.
- Pre-money option pool shuffle above what you need. A 15% pool refresh on the pre-money line dilutes only founders, not the new investor. If the ask is bigger than your 18-month hiring plan, push it post-money or shrink it.
- Founders excluded from pro-rata, drag-along below majority of common. Two paper cuts: founders losing pro-rata in their own company, and drag-alongs that let a minority of preferred force a sale. Both are negotiable and routinely struck.
The 2026 baseline: what "clean" actually looks like
Before flagging anything, know the floor. A clean 2026 seed term sheet has: 1x non-participating preferred, broad-based weighted-average anti-dilution, founder-majority board, no redemption, standard drag-along (majority of preferred AND majority of common), and pro-rata rights for major investors only.
The NVCA Model Legal Documents, updated in 2024, are the reference text every venture lawyer drafts from. If a clause in your term sheet diverges from NVCA defaults, ask why. Sometimes the answer is fair (a lead negotiating real downside protection in a competitive deal); sometimes the answer is "we always ask," and you can push back without consequence.
Use the NVCA fallbacks as your negotiation script. When an investor's lawyer pushes a non-standard clause, your lawyer's first response is "NVCA model has it this way, what's the reason for the change?" Half the time, the clause reverts.
Participating vs non-participating preferred: the most expensive clause you'll skim past
Participating preferred is the single most expensive red flag at exit. It's also the one founders most often miss because the term sheet just says "participating" in one line.
Here's the math on a $10M raise at $40M post, exiting at $50M:
| Structure | Investor take | Common take |
|---|---|---|
| 1x non-participating preferred | $12.5M (better of 1x or 25% of $50M) | $37.5M |
| 1x participating preferred | $10M preference + 25% of $40M = $20M | $30M |
| 2x participating preferred | $20M preference + 25% of $30M = $27.5M | $22.5M |
That's $7.5M moved from founders and employees to one investor on a moderate exit, from one clause. Push back language: "We're modeling on NVCA terms which are 1x non-participating. Participating would be a material deviation from the Cooley benchmark of 94% non-participating at our stage. Can we align on non-participating?"
When participating is acceptable: almost never at seed in a normal round. Wilson Sonsini found participating preferences in 20% of down rounds in 1H 2025, up from 8% in 2024, so if you're raising a bridge at a flat-or-down price, expect to see it. Outside that, treat it as a red flag.
Anti-dilution: broad-based weighted-average is the only acceptable answer
Anti-dilution kicks in when you raise a future round at a lower price. There are three flavors, and only one is standard.
| Type | What happens in a down round | Prevalence at seed |
|---|---|---|
| Broad-based weighted-average | Investor shares reprice partially, accounting for total cap including option pool. Modest founder dilution. | 95–100% of deals |
| Narrow-based weighted-average | Same logic, smaller denominator, so a bigger reprice. More founder dilution. | Rare |
| Full-ratchet | Investor shares reprice all the way to the new (lower) price. Founder gets crushed. | Outlier; mostly bridge or distressed |
The pushback line is simple: "Broad-based weighted-average is the universal standard, confirmed by Wilson Sonsini in 95–100% of recent deals. Full-ratchet would be a deal-breaker for us." Investors who pushed full-ratchet at seed almost always concede.
Board control before Series A is a quiet killer
A seed board with two investors and one founder hands operational control to people who own 20–30% of the company. Standard at seed is a 3-person board with founder-majority, or a 3-person board of one founder, one investor, one mutually-agreed independent.
The non-obvious cost: an investor-controlled board can fire you, block a sale, force a fundraise, or change strategy without your consent. At Series A you're going to give up a board seat anyway; doing it at seed compresses the founder-control window to zero.
Pushback: propose a 3-person board (you, your co-founder, the lead investor), or a 3-person board with one mutually-agreed independent. Either is market.
Option pool shuffle: the trick that dilutes only you
Most term sheets ask for a 10–20% option pool refresh as part of the round. The trick is where it goes on the cap table.
If the pool is pre-money, only the existing shareholders (you) get diluted. If it's post-money, the new investor takes their share too. On a $10M round at $40M post with a 15% pool refresh, the pre-money version costs you about 1.5% more equity than the post-money version. That's roughly $750k of value transferred at the same exit.
Two pushbacks that work:
- "Size the pool to our actual 18-month hiring plan, not a percentage." Build the plan with your lawyer; usually 8–12% is what you actually need at seed.
- "Move the unused portion of the pool to post-money." Splits the difference; gives the investor coverage for hires they want, doesn't dilute you for hypothetical ones.
Redemption, drag-along, and founder pro-rata: the quiet three
Three smaller clauses worth flagging line-by-line.
- Redemption rights: investor can force the company to buy back their shares after a fixed period (typically 5–7 years). At seed, strike entirely; it has no place in a primary seed round and exists mostly to add pressure for a sale you may not want.
- Drag-along threshold: the percentage of preferred that can force common shareholders to sell. Standard is majority of preferred AND majority of common. If the sheet says majority of preferred only, push back; that's the clause that lets a minority of investors force-sell the company over founder objection.
- Founder pro-rata: the right to participate in future rounds to maintain your ownership. Most sheets give it to the lead and major investors and quietly omit founders. Ask for it explicitly. It's almost always granted when asked and almost never offered by default.
When a "red flag" clause is actually fine
Sometimes accepting a non-standard term is the right call. The framework is simple: if the valuation premium covers the expected cost of the clause, take it.
A 1.5x liquidation preference on a $20M check at a $30M pre-money sounds bad. But if the market price for your round is $20M pre-money, you're getting $10M of extra pre-money valuation in exchange for the investor clearing $30M before you participate. On a $200M exit, that costs founders roughly $10M and gives them $10M back via lower dilution. Wash.
The arithmetic only works above a certain exit size. Below that exit, you lose; above it, you gain. Have your lawyer model the breakeven exit value for any non-standard term before you sign, and only accept the trade if your honest expected exit is comfortably above the breakeven.
Why this matters: the seven-figure clause buried on page three
A 1x participating preferred on a $10M seed round costs founders and employees roughly $4M at a $50M exit, vs the non-participating standard that 94% of 2024 rounds used. One sentence on page three.
If you're raising more than one term sheet, redline both against the NVCA model and against each other; the cleaner sheet is often worth more than a $1–2M valuation bump. If you're only raising one, that's the moment to spend $3–5k on a real venture lawyer for a 90-minute review. Founder-friendly leads send clean sheets. The ones that don't are signalling something about how they'll behave for the next ten years.
FAQ
What are red flags in a term sheet? Participating preferred liquidation preferences, full-ratchet anti-dilution, redemption rights at the investor's discretion, board control flipping to investors before Series A, large pre-money option pool refreshes you didn't budget for, founder pro-rata stripped out, and drag-along thresholds below a majority of common. Each one looks small on the page and is expensive at exit.
What term sheet clauses should founders negotiate? Liquidation preference (push to 1x non-participating), anti-dilution (push to broad-based weighted-average), option pool size and timing (post-money, not pre-money, if you can), board composition (founder-majority through Series A), and pro-rata rights for the founder, not just the lead. Valuation gets the airtime, but these five clauses move more dollars over the life of the company.
Is a 1x liquidation preference standard? Yes. In Q2 2024, 95% of priced deals featured a 1x liquidation preference and 94% featured non-participating preferred. Anything above 1x at seed is a hard red flag, not a negotiation.
What is a participating preferred? A participating preferred stock gets its money back first (the liquidation preference), then also shares pro-rata in whatever's left alongside common. So on a $50M exit with $10M raised at 1x participating, the investor takes $10M off the top and then their full equity share of the remaining $40M. Non-participating forces them to pick one or the other.
Related on the hub
- The VC fundraising process in 2026: inside the firm — Related vc process guide.
- The VC due diligence checklist for a seed round in 2026 — Related vc process guide.
- VC term sheet explained: what to fight for in 2026 — Related vc process guide.
Run this playbook inside Causo.
Match to the best-fit partner at 1,000+ funds, draft a hyper-specific email, and send from your email — in one place.