The headline number is two different markets

Seed benchmarks have become hard to use, because the middle of the market and the top of it are moving apart. Carta's latest seed data puts the 95th percentile seed valuation at $200.4 million in Q2 2026, compared with $72.2 million in Q2 2025, a rise of 177%. Separately, Carta's median for software seed rounds over the prior six months sits at $24.3 million on a median raise of $4.1 million.

Two caveats matter. The figures come from different Carta posts with different time windows, so they should not be read as one distribution. And the median covers software companies using Carta cap tables, with bridges and extensions excluded. Carta itself notes that medians are a guideline only, since each deal differs.

The practical lesson for a founder is that a record at the top says very little about where your round will price. If you anchor your expectations on the 95th percentile, you will be disappointed. If you anchor on the median and your company has unusual traction, you may leave value on the table. Use the benchmark to frame the conversation, then let your own numbers carry the argument.

Price is only one line on the term sheet

Founders tend to negotiate valuation hard and the rest of the document lightly. The HSBC Innovation Banking Venture Capital Term Sheet Guide 2026 is a useful corrective. It analyses 711 term sheets (643 from UK-headquartered companies) supplied by 29 law firms, representing £11.2 billion in aggregate investment value, and it finds that investor competition is shifting terms as well as prices. Participating preferences, for example, became less common as competition pushed investors toward more founder-friendly terms. The guide also associates higher use of participating preferences in some regions with lower competitive tension and less capital.

The guide's author, Glen Waters, makes a point that is easy to skip past: seed is the main chance to set favourable terms for later rounds. Whatever you concede at seed tends to become the template that Series A investors expect to see, and the reverse is also true.

A sensible order of priority when you receive a term sheet:

  • Economics that compound. Liquidation preference multiple and whether it is participating. A 1x non-participating preference is the cleanest structure. Anything beyond that deserves a clear explanation of what you get in return.
  • Dilution you cannot see. The option pool size and whether it is created before the investment (so it dilutes you, not the new investor). Ask to see the pool built from a hiring plan rather than accepting a round number.
  • Control. Board composition, protective provisions and what requires investor consent. Early control terms are hard to claw back.
  • Future rounds. Pro rata rights, information rights and any side letters that other investors may expect to match.

None of this is legal advice, and the HSBC page itself says founders should consult a qualified lawyer before negotiating. But the structure above gives you a way to compare two offers on something other than the pre-money figure. A lower valuation with clean terms can be worth more to the founders than a higher one with a participating preference and an oversized pool.

Build leverage before you send the first email

Competitive tension is the one variable the HSBC data ties directly to better terms, and it is the variable founders control most. A few habits create it:

  1. Run a compressed process. Contact investors in the same two-week window so that conversations reach the term sheet stage together. A single slow-moving lead reduces your options.
  2. Segment your list. Separate investors who lead from those who follow, and approach likely leads first. Warm introductions from founders in their portfolio still outperform cold outreach for most funds.
  3. Know your alternative. Be able to state what you will do if no one leads: a smaller round, a bridge from existing holders, or a longer runway plan. Investors can tell when a founder has no fallback.
  4. Keep the story consistent. The deck, the data room and the answers you give in partner meetings should match. Contradictions are what slow diligence down.

Diligence is a deadline, not a formality

Most term sheets are non-binding on price until diligence is finished, and terms can be revisited if something surprising turns up. That makes data room readiness part of your negotiating position. Before you go to market, assemble:

  • A clean cap table that reconciles with your legal documents, including every SAFE, convertible note and option grant.
  • Executed founder IP assignments and contractor agreements.
  • Monthly financials and a simple metrics file that ties to what the deck claims.
  • Customer contracts or pilot agreements, with any unusual terms flagged.
  • A short list of known risks with your own explanation, so that investors hear them from you first.

Founders who are prepared shorten the gap between term sheet and close, which reduces the chance of a re-trade and keeps other interested investors from drifting away.

What to watch next

Carta's quarterly data for Q3 will show whether the top-end surge is spreading down the distribution or staying concentrated in a few hot categories. The HSBC guide is UK-weighted and tied to its sample of law-firm term sheets, so US founders should treat its term trends as directional rather than definitive. If you are raising now, the strongest position is the unglamorous one: a realistic price anchor, a term sheet you have read line by line, and a data room that is ready before anyone asks for it.

This article is information and commentary, not legal or investment advice.

Sources