
Who funds AI, and at what stage
Every live product in the catalog by the funding stage of the company behind it: how that differs across 29 markets, what it correlates with, and how often companies at each stage stop existing.
16 September 2026 · 9 min read · Data frozen 16 September 2026
Funding announcements describe what was raised. This report describes where the money actually sits: every live product in the catalog by the stage of the company behind it, how that distribution differs across twenty-nine markets, what it correlates with, and how often companies at each stage stop existing.
The short version
- All 6,441 live products carry a funding stage. Series A leads at 1,569, ahead of growth (1,295) and bootstrapped (1,272).
- Named B and C rounds are almost absent — 60 products between them. Companies in this catalog go from Series A to a growth label without passing through a labelled middle.
- Stage correlates strongly with market quality. Seed and Series A companies sit in markets averaging 28–33 saturation; growth and bootstrapped companies sit at 46.
- As a rate, public products fail most often (12.5%), then seed (11.1%). Series A produces the most dead tools by count but not by rate.
Coverage
Funding stage is recorded for every live product in the catalog, which makes this the most completely populated dimension we publish. It reflects the last stage we observed, not necessarily the company’s position today; a company that raised last month may still be recorded at its prior stage. “Growth” is a bucket for companies past the labelled venture ladder but not public, and “bootstrapped” means no external round on record rather than confirmed profitability.
One artefact to carry through the report: our pre-seed population is dominated by companies added to the catalog recently, most of them founded in 2024 or later. Wherever pre-seed appears below, it is describing a young cohort rather than a stage-level property.
Finding one: the missing middle
All 6,441 live products. Falcoscan catalog, 16 September 2026.
Series A holds 1,569 products and Series B holds 30. That is not a story about companies failing to raise a B; it is mostly a labelling effect. Companies that raise beyond a first priced round tend to be described — by themselves and by everyone else — as growth-stage rather than by round letter, and our 1,295 growth-stage products absorb most of what would otherwise be B, C and D.
The useful reading is the shape rather than the letters: roughly 1,250 early-stage products, 1,570 at first priced round, 1,400 past it, 940 inside public companies, and 1,270 that never took the ladder at all.
Finding two: stage varies enormously by market
| Market | Products | Early | Series A | Late | Public | Bootstrapped |
|---|---|---|---|---|---|---|
| Healthcare | 137 | 47 | 65 | 3 | 22 | 0 |
| Legal | 83 | 42 | 21 | 9 | 8 | 3 |
| 3D & AR/VR | 68 | 28 | 14 | 2 | 21 | 3 |
| Gaming | 57 | 23 | 10 | 1 | 10 | 13 |
| Real Estate | 82 | 31 | 30 | 5 | 15 | 1 |
| Research | 87 | 31 | 27 | 5 | 12 | 11 |
| Avatars | 66 | 25 | 21 | 1 | 4 | 15 |
| Music | 61 | 28 | 9 | 6 | 7 | 11 |
| Human Resources | 96 | 25 | 56 | 9 | 5 | 1 |
| Automation | 125 | 45 | 37 | 15 | 17 | 10 |
| Search & SEO | 68 | 5 | 8 | 4 | 5 | 46 |
| AI Models | 313 | 32 | 75 | 93 | 82 | 30 |
| Agents | 344 | 119 | 87 | 50 | 33 | 55 |
| Data | 309 | 78 | 96 | 62 | 42 | 31 |
| Security | 257 | 31 | 88 | 61 | 54 | 22 |
| Finance | 281 | 48 | 93 | 88 | 47 | 5 |
| Ecommerce | 304 | 57 | 88 | 104 | 34 | 20 |
| Support | 230 | 49 | 56 | 56 | 52 | 17 |
| Voice | 272 | 33 | 75 | 73 | 28 | 63 |
| Learning | 258 | 63 | 38 | 52 | 60 | 45 |
| Video | 296 | 53 | 54 | 84 | 41 | 64 |
| Sales | 276 | 55 | 97 | 56 | 39 | 29 |
| Productivity | 303 | 57 | 74 | 45 | 32 | 95 |
| Coding | 422 | 68 | 82 | 115 | 60 | 96 |
| Design | 357 | 41 | 53 | 71 | 49 | 143 |
| Marketing | 280 | 23 | 83 | 96 | 51 | 27 |
| Social | 266 | 31 | 38 | 60 | 27 | 110 |
| Image | 392 | 36 | 57 | 91 | 70 | 138 |
| Writing | 351 | 50 | 37 | 81 | 15 | 168 |
Live products per market, grouped by stage. Early combines pre-seed and seed; late combines Series B, C, D+ and growth. Falcoscan catalog, 16 September 2026.
The contrast that matters is between markets where capital is a prerequisite and markets where it is optional. Healthcare has zero bootstrapped products out of 137. Search & SEO is 67.6% bootstrapped. Between those two sits the whole spectrum of how expensive it is to reach a first customer.
Percentage of live products per market recorded at bootstrapped stage. Falcoscan catalog, 16 September 2026.
Finding three: funded companies are in better markets
| Stage | Products | Opportunity | Saturation | Gap |
|---|---|---|---|---|
| Pre-seed | 292 | 69.7 | 30.1 | 39.6 |
| Seed | 962 | 68.3 | 28.3 | 40 |
| Series A | 1,569 | 70.8 | 32.9 | 37.9 |
| Series B | 30 | 73.7 | 26.6 | 47.1 |
| Series C | 30 | 68.9 | 31.1 | 37.8 |
| Series D+ | 43 | 66.1 | 35.8 | 30.3 |
| Growth | 1,295 | 56.3 | 45.9 | 10.4 |
| Public | 942 | 62.9 | 41.9 | 21 |
| Bootstrapped | 1,272 | 55.5 | 46.4 | 9.1 |
Mean scores across live products at each stage. Falcoscan catalog, 16 September 2026.
Seed companies sit in markets averaging 68.3 opportunity against 28.3 saturation — a 40-point gap. Bootstrapped companies sit at 55.5 against 46.4, a 9-point gap. Growth-stage companies are barely better at 56.3 against 45.9.
Two mechanisms are tangled here and it is worth separating them. Early-stage companies score well partly because capital buys entry to expensive markets, and partly because our early-stage population is younger, and younger companies are in newer markets that have not had time to fill. The growth-stage figure is the cleaner observation: those companies entered markets years ago that have since crowded around them.
Finding four: where companies actually stop
Share of tracked products at each stage that no longer resolve. Stages with fewer than 100 tracked products excluded. Falcoscan liveness sweep, 16 September 2026.
Public leads at 12.5%, which is not a story about public companies failing — it is large organisations retiring standalone products. Seed follows at 11.1%, bootstrapped at 8.8%, and Series A at 8.2%.
That ordering is worth holding against the count-based version in our mortality report, where Series A produces the most dead products of any stage. Both are true. Series A is the largest venture-funded population in the catalog, so it supplies the most of everything; per company, it is among the safer stages to be at. Counts tell you where the bodies are. Rates tell you where the risk is. Reports that mix them up produce the familiar and wrong conclusion that raising a Series A is dangerous.
What this changes
For founders. The stage distribution of your market is a better guide to whether you need to raise than any general advice about it. If your market is 40% bootstrapped, capital is not the constraint and raising mostly buys speed. If it is 0%, the first customer is behind a door that only money opens.
For investors. The opportunity-saturation gap by stage is the sharpest number here. Growth-stage AI companies are, on average, in markets barely distinguishable from the ones bootstrappers occupy. Whatever was true about their market when they entered it, it has closed since.
Photo: Pavel Danilyuk / Pexels. Colour-graded for Falcoscan.
Citing these numbers
Every figure here is from the Falcoscan catalog as it stood on 16 September 2026, and is frozen at that reading. Later changes to the catalog will not alter this page, so a number you quote today will still say the same thing when someone checks it.
Falcoscan, “Who funds AI, and at what stage”, 16 September 2026. https://falcoscan.com/reports/who-funds-ai-and-at-what-stage