Last week I covered the four types of potential equity investors for AI professional services companies (e.g. training, consulting, implementation), and why none is a clean fit right now (Catch up on it here).

But the window to build in this market is open today, not in three years. So what other acceleration options are on the table?

Quite a bit, as it turns out. The options below are all non-dilutive, meaning no equity changes hands, no new cap table entries, no investor on your board telling you what to do. That sounds straightforwardly good, and in many ways it is. But each comes with its own mechanics, qualifications and gotchas. Let’s walk through them.

Bank loans are the oldest and most widely used form of business financing. Working in tech, it's easy to forget that most of the world runs on them. 

Here is the basic structure: you borrow a fixed amount, repay principal plus interest over a set term, and your cap table stays untouched. Interest rates run 6-10%+, depending on the underlying rate environment. This is by far the cheapest form of commercial capital outside of grants. 

The catch: most banks require a personal guarantee from directors. If the business can't service the debt, the bank can come after your house. That's a real risk.

The bigger catch: banks almost certainly won't lend to you unless your numbers look like an established SME, not an early-stage growth business. Their assessment framework is the same whether you’re a local hardware shop or a fast-growing AI implementation firm. The core question is not "how fast is this growing?" or "how large could this become?" It is simply: "can this company reliably repay a loan?"

To get past that filter, you'll need to show consistent, predictable EBITDA; a debt service coverage ratio of at least 1.3 (i.e. your operating income comfortably covers principal and interest); ideally three or more years of operating history; and a diversified client base. If you clear those hurdles, expect to borrow an amount equivalent to 0.5–1.5x EBITDA. Not transformational, but real capital on cheap terms.

If you're running an early-stage AI Enablement business: Unless you are past the £2–3m revenue mark with a clean (and long!) track record, bank debt probably isn’t available to you yet. Build the relationship now anyway: open a business account, meet the relationship manager, understand what they need to see. You want to be ready when the numbers support it.

Venture debt fills the gap left by traditional banks. Still debt, still repaid with interest, but the lenders are comfortable with high-growth profiles.

The structure differs in a few ways. You typically pay only interest for an initial period before principal repayments begin. Loan sizes can reach 40% of annual revenue. And, crucially, lenders take a small equity stake via warrants (usually 0.5–2%). Interest runs 9-13%. So this is materially more expensive than bank debt, and not perfectly clean from a dilution standpoint, though the impact is far smaller than a typical equity round.

Venture debt funds prefer companies with VC backing. A VC backer signals quality, provides due diligence the lender can piggyback on, provides board oversight, and offers a credible refinancing option if things go sideways. Without that safety net, a bootstrapped company would only qualify for venture debt if it had £3-5m in revenue, 30-60%+ year-on-year growth, and gross margins in the 60-80% range. In short, software-like numbers that only the best-in-class service businesses will reach.

For service firms, lenders will want strong recurring or at least re-occurring revenue, a diversified enterprise client base, and repeatable inbound demand. If you can show all three, venture debt can provide meaningful capital – particularly when you need to scale delivery to meet demand or have an acquisition target in sight.

If you're running an AI Enablement service business: Venture debt is probably a £5m+ revenue conversation. But if you're building toward that number with strong retainer revenue, it's worth understanding the mechanics now. Lenders in this market move fast when they like what they see.

These get lumped together but serve different purposes. 

Revenue-based financing works like this: a lender provides capital and you repay a fixed percentage of monthly revenue, typically 10–15%, until you've repaid 1.3-1.6x the original amount. So on a £300k facility, you'd repay somewhere between £390k and £480k in total. Repayment flexes with revenue, limiting your downside in slow months. Approval is usually fast. And there is no dilution.

The true cost, however, is high. That 1.3-1.6x repayment cap translates to an effective annual rate well above 20-30% depending on how quickly you repay. And the model  works best when revenue is predictable. The lender's core question is "can I share a fixed percentage of this company's revenue every month and reliably get paid back?" The answer is stronger from a subscription training business than from a consulting or implementation firm with lumpy, project-based revenue.

Invoice financing is a more natural fit for project-driven services. Rather than borrowing against the whole revenue base, you borrow against specific unpaid invoices. The mechanics are straightforward: you issue a £100k invoice to a corporate client with 60-day payment terms. The lender advances £85k immediately. The client pays in 60 days, the lender keeps £2-4k as their fee, and you receive the balance. You've effectively converted a 60-day receivable into same-week cash.

The lender’s main risk exposure is to your customer, not you. Enterprise clients with strong balance sheets are ideal; SME clients make lenders sceptical. And invoice financing, by definition, only works once you have invoices to finance – it can't fund growth ahead of revenue. And if you rely on it constantly, the fees compound into a real cost.

If you're running an AI Enablement service business: Invoice financing is probably the first product you should explore. It's the fastest path to unlocking cash tied up in receivables. Revenue-based financing makes sense once you've built a meaningful recurring revenue base.

If you're running an AI Enablement services business:

The path isn't linear and it rarely involves just one product. Most founders who've navigated this well have combined two or three instruments at different stages – not because any single one was perfect, but because the landscape rewards those who understand all of it.

That's not as glamorous as closing a Series A. But it's how you stay in control while the market is still wide open.

— Daria