Every pitch deck guide on this site ends the same way: a founder asks what they should be raising. The honest answer is that the number in the headlines is not the number in the data.
We keep a library of 492 decks from funded startups. 417 of them state a round we can read, spanning 2003 to 2023. Here is what those 417 rounds actually look like.
The median round is much smaller than the one you read about
| Stage | Decks | Median round | Range |
|---|---|---|---|
| Pre-Seed | 42 | $150K | $30K – $50M |
| Seed | 197 | $750K | $20K – $66M |
| Series A | 72 | $9.15M | $500K – $180M |
| Series B | 31 | $30M | $8.5M – $123M |
| Series D | 8 | $105M | $17M – $250M |
The medians are unremarkable. The ranges are the story. A seed round in this set runs from $20K to $66M — the same word, covering a 3,000x spread. When a founder says "we are raising a seed round," the word carries almost no information about the amount.
That matters for how a deck is built. A deck raising $300K and a deck raising $6M are answering different questions, and copying the structure of the second while raising the first is how a deck ends up over-promising on a slide it did not need.
Most of these companies sell to businesses
Of the 492 companies, 249 sell B2B and 174 B2C, with 50 B2B2C and 28 C2C. Only 8 sell to government. On business model, SaaS dominates at 241, followed by apps (82), marketplaces (61) and e-commerce (52).
This is a library of decks that got funded, not a survey of startups, so the skew is a statement about what investors funded rather than what founders built. The read for a founder outside those defaults — hardware, services, anything selling to a ministry — is that there are fewer templates to copy, and the deck has to do more explaining.
Where the seed number moves
Splitting seed rounds by sector, only six sectors have enough decks (eight or more) to be worth reporting:
| Sector | Seed decks | Median seed round |
|---|---|---|
| Fintech | 10 | $3.75M |
| Retail | 9 | $1.6M |
| Art | 8 | $788K |
| Software | 11 | $650K |
| Marketing | 13 | $500K |
| Analytics | 9 | $325K |
Fintech seeds at roughly five times software. That is not a quality gap. It is what it costs to reach a first transaction when a licence, a sponsor bank and a compliance stack sit between the product and the first dollar. Sectors that need infrastructure before revenue raise more at seed because they have to.
What to do with this
Three things a founder can act on:
- Name your number against your sector, not the headlines. The median seed here is $750K. If you are pitching $5M at seed, the deck has to carry a reason that is visible on the slide.
- Stage words are not amounts. Say the amount and what it buys. "Seed" tells an investor nothing about your plan.
- If you are outside SaaS/B2B, expect to explain more. The decks that worked in this set were mostly in a shape investors already recognised. Yours may not be.
Method: 492 decks in the 2080 library; 417 with a round we could parse; stages and sectors as tagged on each deck. Medians, not means, because a handful of very large rounds sit inside every stage. Sector cuts reported only where at least eight decks exist.