B2B software used to be the default venture category. It is now the sector competing hardest for attention, because roughly half of all venture capital in 2025 went to AI and five companies alone took a fifth of the entire market. What still gets funded here is well defined: software with a wedge into a workflow, a reason it cannot be rebuilt in a weekend, and an account of what happens when a general model reaches your category.
Reviewed Full-year 2025
These are market-wide figures rather than B2B software totals, and they are the relevant ones because they describe what this sector is now raising against. Roughly half of 2025 funding went to AI, and OpenAI, Scale AI, Anthropic, Project Prometheus and xAI took $84bn of it between them. The good news underneath is that early-stage capital grew: $37bn in the fourth quarter, up 36%. Enterprise software has not stopped being funded, it has stopped being the default.
A fund tagged "enterprise" sits in one of four columns. They diverge most on what evidence they want at seed, and on how much revenue concentration they will tolerate before it reads as a services business.
| Horizontal SaaS | Vertical software | Data & infrastructure | Software-enabled services | |
|---|---|---|---|---|
| Typical entry stage | Pre-seed to Series A | Seed to Series A | Seed to Series A | Seed |
| Cheque shape | Ordinary venture | Ordinary venture | Venture, sometimes with usage-based milestones | Venture, with margin scrutiny |
| Wants to see | Fast self-serve adoption or a repeatable sales motion | Depth in one industry and a wedge others cannot copy | Usage growth and a technical moat | Gross margin trending towards software |
| Time to revenue | 6-12 months | 6-18 months | 12-18 months | Immediate |
| Biggest risk they underwrite | That a general AI product absorbs the category | A total market too small to return the fund | A cloud vendor bundling it for free | That headcount scales with revenue |
| Who else must be in the round | Nobody in particular | An operator angel from that industry | Angels with developer reach | Nobody in particular |
Enterprise software investors we hold in the Causo catalogue, grouped by the cheque they actually write. Open any of them to see the partners, the stage and the recent deals. This is not every enterprise software investor in the market, and no catalogue is.
Firms whose entire mandate is software sold to businesses.
Firms writing the first institutional cheque, several built around operator networks.
Firms that lead once there is a repeatable sales motion and durable net retention.
Balance-sheet investors whose parent is a plausible channel, platform or acquirer.
The gates that are specific to this sector, and that a generalist fundraising guide will not tell you about.
One worked example of what reading those sources produces, from the Causo catalogue with the identity removed.
A seed fund built around an operator network rather than a thesis, investing in business software with a stated preference for founders selling into functions its members have run. The value proposition is distribution: introductions to buyers inside the network.
B2B software specialists include Emergence Capital, Point Nine Capital, Scale Venture Partners, Notion Capital, Boldstart Ventures, Dawn Capital and Meritech Capital Partners. Corporate arms including Salesforce Ventures, Dell Technologies Capital and Intel Capital invest strategically. Which of them fits depends mostly on whether your product is horizontal, vertical or infrastructure.
Yes, but it is no longer the default category. Global venture funding reached $425bn in 2025 across more than 24,000 companies, up 30%, while roughly half of that went to AI-related companies and five AI companies alone took $84bn. Early-stage capital grew, with $37bn deployed in the fourth quarter alone, up 36% year on year.
You need a credible answer to what AI does to your category, which is not the same thing. Investors are underwriting the risk that a general model absorbs your product. Companies with proprietary data, deep workflow integration or a regulatory position defend that well. Companies whose only differentiation is a better interface generally do not.
Net revenue retention and gross retention first, then expansion by cohort, cost of acquisition against payback period, and the length and repeatability of the sales cycle. At seed, before those exist, they look for a wedge into a specific workflow and evidence that early customers rely on it rather than trial it.
There is no single threshold and any number quoted as universal is wrong. What consistently matters more is the shape: growth rate, retention, and whether the sales motion repeats without the founder in every deal. A company with less revenue and a proven motion is more fundable than one with more revenue and no pattern.
Horizontal investors back software sold across industries and expect fast adoption and a very large market. Vertical investors back software for one industry, expect deep domain expertise, and will accept a smaller market if the product is hard to displace. They ask different questions and misreading which one you are talking to is the most common way these meetings go wrong.