If you're running an AI professional services firm of any description (e.g. training, consulting, implementation etc.), sooner or later someone in the room – a co-founder, a board advisor, a well-meaning friend at a dinner party – will raise the question of investment. And when that moment comes, you'll quickly discover that "investment" is not a single thing. It's a category with at least a dozen subcategories, each with its own logic, its own incentives, and its own way of reshaping your company in ways you may not have anticipated.

Paul Graham mapped this territory back in 2005 in his widely-read essay "How to fund a startup." It was a useful guide for its time. But that was 21 years ago, and the definition of a "tech startup" has expanded considerably since then. The AI enablement firm – part consultancy, part technology provider, part training business – didn't really exist in his framework. It barely exists in most investors' frameworks today.

That gap is precisely the problem.

Most founders understand, in the abstract, that different investors want different things. What they often don't appreciate is how much the type of capital you raise determines the trajectory of your company: not just its growth rate, but its shape, its culture, and the decisions you're allowed to make three years from now.

For AI Enablement founders, the stakes of getting this wrong are particularly high. That's because most investors don't quite know how to classify these businesses. Are they services firms? (Good profitability, limited scalability, low upfront capital requirements.) Are they tech-first platforms? (Moderate upfront investment, much higher ceiling on scale.) Are they recurring revenue businesses? And if so, what multiple should apply?

No one has settled on a consensus answer yet. I suspect we won't for another two or three years. In the meantime, founders end up in conversations with investors who are pattern-matching to the wrong archetype entirely. At best wasting time, at worst signing a term sheet that quietly re-routes the company toward an outcome that works for the fund and not for the founder.

To avoid that outcome, you need to understand the investor landscape properly.

Before we get into individual investor types, one distinction cuts across all of them.

Dilutive capital means the investor receives a stake in your company – they own a percentage of it – in exchange for their money. Depending on the terms, they may also install a director on your board and have formal say in major decisions.

Non-dilutive capital means the investor provides funds without taking equity. You pay interest; you keep ownership. (There are some nuances here with venture debt, which we'll cover in Part 2.)

This piece covers dilutive capital: venture capital, private equity, angel investors and family offices, and strategic investors. Non-dilutive options – revenue-based financing, venture debt, grants, and others – will follow next week.

The basic model

Venture capital funds are designed around a specific and somewhat counterintuitive math. A typical early-stage fund might manage, say, $100m–$500m. It deploys that capital across 20–40 companies, fully expecting that the majority will return little or nothing. The math only works if a handful of companies grow large enough, think $1bn+ valuations, to generate returns that cover all the losses and still deliver a 3x net return to the fund's own investors over a 10-year horizon.

That single fact explains almost every behaviour you'll observe from venture investors.

VC Math: When a $50M Exit Isn't Enough

On paper, VC is often the most attractive-looking capital. There's no interest to pay. Valuations at early stages can be generous: it's not unusual for a team with a working prototype and a compelling pitch to get valued at $10m–$20m pre-money, even before meaningful revenue. You raise $2m, give up 15%, and get back to work.

But "cheap" in finance means something specific: it means the expected cost to you if things go well. If your company doesn't reach venture-scale outcomes, that "cheap" money becomes very expensive indeed, because the fund will spend years pushing you toward a larger exit that may not be the right outcome for the business you've actually built.

A $50m exit is life-changing for founders and employees. For a $500m fund that needs to return $1.5bn to its own investors over 10 years, it barely moves the needle.

The ratchet effect

There's a structural dynamic that catches many founders off guard: once you raise venture money, you typically keep raising venture money. Each round sets a higher valuation benchmark. Each subsequent investor needs to believe the company will grow into that valuation. The pressure to scale (headcount, revenue, market footprint) compounds with each cheque. The more you raise, the higher the bar a potential acquirer has to clear for the deal to make sense for everyone on the cap table.

This is fine if you're building a category-defining platform business. It is less fine if you're building a highly profitable services and training firm that could generate strong returns at $5m–$20m revenue and would be a natural acquisition target for a large consultancy or adjacent tech player. Raising the wrong kind of capital for your business model can get you very stuck.

What does this mean specifically for AI Enablement founders?

Historically, VCs avoided professional services businesses on the grounds that they weren't scalable: growth required adding people, and people are slow to hire and expensive to carry. The margin profile was good, but the ceiling felt low.

AI Enablement firms change some of that calculus. The demand is genuinely unprecedented. Implementation and training work that would have taken a team of 15 three years to build is now being deployed in months. And some founders are genuinely building towards a productised offering: a platform, a library of IP, a repeatable methodology, that does have VC-grade scale characteristics.

But here's the thing: the push toward a "scalable product" often comes from the investor's mandate, not from market demand. To a fund with a hammer, most things look like a nail. Just because your customers are buying bespoke AI implementation projects, they are not necessarily waiting to buy a $99/month SaaS tool from you. In fact, pitching them the SaaS tool might cost you the implementation contract.

There are a few famous exceptions that proved it could work. Palantir spent years doing heavy lifting with services before their product could truly change the game for their consulting customers. Faculty.ai also found success by mixing consulting with research as they scaled up (check out our deep-dive on their $1bn acquisition by Accenture here).

But these cases are rare for a reason. Most AI services companies that take venture capital end up spending years struggling, trying to force themselves into a "software-only" mold that just doesn't fit how they actually operate.

The older, more conservative cousin

Private equity manages roughly three times as much capital as venture, globally. And that's before accounting for the large venture-backed companies that stay on VC books at paper valuations. PE has been around longer, and individual funds have specialised considerably: growth equity funds, buyout funds, turnaround specialists, sector-focused funds etc.

The key differences from venture, in practical terms:

PE funds really don't like to lose money. That means they come in later (typically once a business is generating at least $1m–$2m in EBITDA, a threshold that has been creeping upward as fund sizes grow), they take protective deal terms, and they tend to engineer the business toward a specific exit: usually a strategic acquisition or a secondary sale to a larger fund  within 4–7 years.

Is this the right fit for AI Enablement firms?

Honestly, very few AI Enablement firms have reached the scale where PE becomes relevant. But for those that have, or are approaching it, the fit questions are worth examining.

PE works well when commercial operations are streamlined and predictable. If your revenue is lumpy (a big training contract one quarter, a dry spell the next) carrying debt through a PE-backed structure creates real cash-flow risk. A couple of volatile quarters can put you in a difficult position with your lender.

If your business isn't growing fast enough for the fund's return model, expect to be merged with similar firms in a "roll-up" strategy. The fund aggregates several smaller players, creates a combined entity at sufficient scale, and sells the whole thing. You may end up running a division of something larger than you intended to build.

Success cases exist in adjacent categories: Skillsoft's acquisition by Bain Capital, PA Consulting Group's acquisition by Carlyle Group. Both were mature businesses with long histories of profitable, repeatable operations. It’s going to be a while until AI firms get to that point. We might see some “tourists” lurking around. But it is more likely that PE funds will focus on injecting “AI” into more traditional businesses first, before backing pure play AI Enablement businesses.

The most varied category, by a wide margin

Angel investors and family offices sit in a broad and heterogeneous category. At one end: individual investors, sometimes wealthy professionals, sometimes tech executives, writing $10k–$100k cheques with light terms and minimal governance expectations. At the other end: large multi-family offices managing hundreds of millions, participating in majority transactions alongside PE funds.

Neither end behaves like the other, which is both the appeal and the challenge of this category.

The access problem

Institutional investors ( VCs, PE funds) have websites, partners, and reasonably predictable processes. Angels and family offices mostly operate through introductions. Finding the right one for your specific business requires either a strong network or a lot of patience (often both). For founders spending that time, it is worth periodically asking whether the search is taking more attention than the business itself.

The EIS/SEIS factor in the UK

If you're raising from private individuals in the UK, this matters. The Enterprise Investment Scheme (EIS) and Seed Enterprise Investment Scheme (SEIS) allow qualifying investors to claim meaningful tax relief, which makes the effective cost of a failed investment substantially lower. For many private investors, EIS/SEIS eligibility is close to a prerequisite.

Here's the tension: training businesses tend to qualify relatively easily. They have a clear, repeatable service with identifiable IP. Consultancies often don't. HMRC has historically viewed consulting firms, particularly where the founders are the primary value generators, as "lifestyle businesses" (trust me, I LOL at this together with you) rather than high-growth, scalable companies. That distinction can make AI consulting and implementation firms difficult to qualify, and therefore harder to finance through UK-based angels.

What about family offices?

Large family offices are less constrained by EIS/SEIS considerations. They're also, I suspect, more likely to have the domain understanding to evaluate an AI Enablement firm on its actual merits. The challenge is finding the specific individuals with both the sector knowledge and an active mandate to deploy in this area. They exist. They're just not easy to find without a warm introduction.

The non-financial investor

Strategic investors are operating companies, often larger players in your sector, sometimes customers, that invest in businesses where they see a commercial or competitive rationale. Unlike the other categories, they're not primarily motivated by financial returns. They're investing to get closer to a technology, a team, or a market position.

That has two meaningful implications. First, a credible strategic investor can be a significant signal to the market and a potential path to acquisition. Second, their presence on your cap table can quietly complicate your commercial relationships with their competitors.

If you're primarily serving companies in a specific industry vertical, and a large player in that vertical wants to invest, the question isn't just "do we like the terms?" It's "how will this affect our ability to sell to everyone else in that vertical?"

What to watch in the term sheet

Strategic investors are typically less sophisticated as investors than financial investors. That's not a criticism, it's a structural reality. They don't spend their days reviewing term sheets. Which means the terms they propose may have provisions that would be unusual in a financial transaction, and which could create problems later.

Pay particular attention to: information rights (what can they see about your business, and does that overlap with commercially sensitive client data?), anti-dilution provisions (which could create friction when you raise from financial investors later), and any exclusivity or right-of-first-refusal clauses on an eventual acquisition.

If the strategic investor has a portfolio of other companies or a strong network of financial investors, that's worth a great deal. If they're largely isolated from the fundraising ecosystem, be realistic about how much they can help you with subsequent rounds.

The irony of the investment world is that a field entirely about money has managed to make money non-fungible. A cheque from a VC and a cheque from a strategic investor are not the same thing, even if the numbers are identical. They come with different expectations, different pressures, and different definitions of what "success" looks like in five years.

For AI Enablement founders specifically, the challenge is that the industry sits in an unclassified zone. Investors are trying to fit it into frameworks built for businesses that aren't quite like yours. Some will see a services firm and pass. Some will see a tech company and push you toward a product pivot you never planned. A few, the genuinely useful ones, will see it for what it is: a category that's still being defined, with real commercial momentum and a range of plausible exit paths.

Finding that last group is harder. But understanding the logic and constraints of every investor type is the prerequisite to even recognising them when you do.

Next week, we'll cover non-dilutive options: revenue-based financing, venture debt, grants, and what each of them means for an AI Enablement business.

— Daria