Float CFO on the credit risk behind rising AI revenue
A massive surge in revenue can make any company — but especially an AI company — look like a strong candidate for financing.
At Stockholm, Sweden-based Float, where co-founder, CFO and COO Jannis Koehn helps provide funding to European tech companies, the question nowadays is what happens to the business six months later. He has seen some European companies grow quickly before customer churn catches up with them.
That pattern has made Koehn more cautious about lending to AI companies, he told CFO.com in a recent interview. He’s more optimistic about established SaaS businesses using AI to improve their products, but for a lender, the test is whether customers keep paying once the excitement and initial launch of a new product fades.
Permission granted by Permission granted by Jannis Koehn
Jannis Koehn
Co-founder, CFO, COO, Float
First CFO Position: 2015
Notable previous employers:
- Looklet
- Boston Consulting Group
This interview has been edited for brevity and clarity.
ADAM ZAKI: Your company lends directly to businesses, but where does the money you lend actually come from?
JANNIS KOEHN: We work with credit funds and banks. Before starting this company, I didn’t know much about private debt funds. They were our first source of funding. Our main funder today is a U.K. credit fund, and we’re in discussions about adding a second facility, most likely with a European bank.
We also found another source of capital when we ran into a growth constraint. A senior lender might provide an advance rate of around 80% to 85%, which means we need to fund the remaining portion ourselves. We had customer demand and senior funding available, but raising the equity to support more lending was very difficult at the time, especially for a fintech with a capital-intensive business.
A credit fund wasn’t going to provide a mezzanine investment of around €1 million; that was too small for them. So we worked with a Swiss bank to issue private debt tokens on-chain to fund that junior portion. I believe we were the first, or perhaps the second, in Europe to make institutional-grade private debt tradable on-chain.
It gave us access to capital we couldn’t otherwise reach and allowed us to grow roughly six to seven times since then. It was a fascinating project because it solved a real problem for the business.
When a company comes to you for funding, how do you decide whether your financing makes sense for them or whether they should raise equity or go to a bank?
I’m a founder myself, and our goal is to help European founders build their businesses. If we think a prospective customer would be better served by equity or a bank loan, we’ll say so. We may be able to fund them, but that doesn’t mean our financing is the right choice.
Over the past few years, we’ve spoken with thousands of tech companies, mostly SaaS businesses. We’ve funded more than 150 companies and deployed more than €100 million across 17 European countries. Our sweet spot is usually a company with around €1 million to €20 million in annual recurring revenue, though we’ll look at companies slightly below that range.
We look at the balance between growth and cash burn. Most of our customers are burning cash, and that’s fine. It’s one reason banks may be unwilling to finance them. We’re comfortable taking additional risk because we believe we understand SaaS and other tech businesses. But the burn has to be moderate relative to the company’s growth.
If a company is burning an amount equal to its monthly recurring revenue every month, I might tell the founder they have a fantastic business, but it’s an equity case right now. Come back in six or 12 months, when the burn has come down, and you want to avoid dilution from another equity round.
You’ve spoken with thousands of founders. Are you seeing any recurring mistakes in the businesses seeking funding?
One pattern we’ve seen recently is revenue shooting up very quickly at some companies, often AI companies. Then, six or 12 months later, churn rises sharply. Revenue might still be growing, but at a much slower pace and with a lot more effort behind it.
“A loan agreement could run 20, 30 or 40 pages. Ours is about six or seven pages because we don’t think all that complexity is necessary.”
Jannis Koehn
Co-founder, CFO and COO, Float
A new product can hit a nerve. Customers have AI budgets, so they’ll try it. The question is whether the product provides lasting value or whether people are just experimenting. There’s also the risk that a large language model provider adds a feature that replaces a product someone built on top of its model.
We funded one or two AI companies early on and then became more cautious. That doesn’t mean we won’t fund AI companies. As with any lending decision, we need enough of a track record to assess customer retention, net revenue retention and whether the growth will last.
Does this rise of AI change how you view the SaaS companies you fund?
A few months ago, there was a narrative that SaaS was dead and AI would replace it. We wrote a memo fairly early on saying we thought that was largely nonsense.
SaaS companies already have an established business model. AI can make them more competitive because their development costs go down, or their developers become more productive. Tech companies also tend to adopt new technologies quickly, which gives them an advantage.
Virtually all our customers today have a strong AI element in their products. I would say most of them are stronger businesses than they used to be.
In the U.S., some alternative financing products have reputation and regulation problems, particularly around products like merchant cash advances, invoice factoring and lines of credit with unclear terms. How do you make your financing transparent to customers?
We want to offer financing that is simple, flexible and transparent. A customer gets a credit facility and can draw as much or as little as they need, whenever they need it. If they need €200,000, they don’t have to draw the full €1 million and pay for money they aren’t using. If they need another €100,000 later, they can draw that then.
The pricing is pay-as-you-go. There’s one charge on the money the customer actually draws. We don’t pass on structuring, setup or legal fees. There’s no availability fee, commitment charge or exit fee. Before a customer draws money, we show them every cash flow, both graphically and in writing, so they know what they’re getting into.
We also try to make the agreement understandable to someone with business experience. A loan agreement could run 20, 30 or 40 pages. Ours is about six or seven pages because we don’t think all that complexity is necessary.
I’ve come to believe the scarcest resource a startup has is its founder, and the founder’s scarcest resource is energy. If something gives you energy, you’ll find the time for it. If it drains you, everything takes longer. We want the financing process to save founders both time and energy.
I had lunch today with a Spanish customer who was visiting Stockholm. He told me he likes the freedom the product gives him: He can grow on his own terms without adding anyone to his cap table.
You’ve worked as a consultant, CFO, COO and CEO. Now you’re a founder and CFO, too. Which role has taught you the most about running finance?
Being a founder and CFO at the same time has taught me the most, especially at a fintech that deals with credit. You have the responsibilities any company has. You need to manage liquidity, report to investors and understand whether the business is doing well and whether it’s profitable.
Then there’s the credit business, which I find challenging and enjoy. How do you structure a financing facility? How do you manage the capital behind the loans you make? On top of that comes the emotional investment and commitment of building your own company. Those are the three layers: the operational work of a startup, the credit decisions and being a founder.
In a B2B lending fintech, you also deal with much larger capital swings. Our revenue is only about 5% to 10% of the capital we move, so the amounts involved are 10 to 20 times larger than what I dealt with at my previous company, where I was CFO and CEO. Much of that capital is lent out rather than earned as revenue.
Large companies are starting to give CFOs formal operational responsibilities. You’ve held both finance and operational roles. Why do you think those responsibilities are coming together?
It reduces silos that can work against each other. The CFO can easily become the person who always says no or steps on the brakes. Then someone else in the company is the person who always says yes and pushes the accelerator. That tension can be productive, but it can also lead to entrenched fights. Eventually the CEO has to step in, and someone gets overruled.
I think it’s valuable for finance people to have real operational experience and incentives tied to how the business performs. It’s just as valuable for salespeople to think about the bottom line.
We try to put that into practice here. I don’t want our salespeople pushing for more loans while our credit team pushes for fewer. If the underwriters’ only goal were to minimize credit losses, we wouldn’t make a single loan. We’d have zero losses, but we also wouldn’t have a business.
