Most founders raise money the way most people buy a car: they work out the total they need, then find someone to hand them the total. One instrument, one cheque, one price.
That is almost always the most expensive way to do it, because businesses are not one thing. A business is a bundle of costs with wildly different risk profiles, and the capital markets price risk. Fund a predictable cost with equity and you have paid a venture return for a bond-shaped risk. You will not feel it this year. You will feel it at exit, when you find out how much of the company you sold to finance your float.
This playbook is the fix. It takes about two hours with your own numbers, and it is the same exercise behind structures like Félix’s recent round, where roughly 57% of a $200 million raise came as credit rather than equity.
The one idea
Every dollar you need falls into one of two buckets.
- Uncertain-return money. You cannot say when or whether it pays back. New products, new markets, research, brand, a senior hire in a function you have never had. If it fails there is nothing to repossess.
- Known-return money. It has a measurable, repeating payback you can evidence from history. Inventory you have sold before. Receivables from customers who have paid before. Customer acquisition in a channel with proven cohorts. Float and working capital.
Equity is the right instrument for the first bucket and a terrible one for the second. Equity costs you a permanent share of all future profit; it is priced for the possibility of total loss. Credit is the right instrument for the second bucket and a dangerous one for the first: it costs a rate and then it ends, but it must be repaid whether or not the thing worked.
The entire playbook is: sort your need into the two buckets, then price each bucket with the instrument built for it.
Step 1 — Split the ask (30 minutes)
Take your funding number and break it into line items. For each line, answer three questions in writing:
- Has this exact spend produced a measurable return before, at least twice? Not something similar. This.
- Can I evidence the payback period with data a stranger could audit? Cohort tables, an aged receivables report, inventory turns.
- If it fails, does an asset survive? Stock, an invoice, a contracted subscription.
Three yeses means known-return: finance it with debt. Any no means uncertain-return: that is equity’s job. Be strict. “We’re confident it’ll work” is not a yes. The test is evidence, not conviction.
A worked example. A company decides it needs $2 million:
- $700k — paid acquisition in two channels running 14 months at a 9-month payback. Three yeses. Debt.
- $450k — inventory for a SKU sold profitably for two years. Three yeses. Debt.
- $600k — six engineers to build a new product line. No. Equity.
- $250k — entering a new country. No. Equity.
The instinct was a $2 million equity round. The correct structure is $850k of equity and $1.15 million of credit. At a $10 million pre-money, that is the difference between selling roughly 17% of the company and selling roughly 8%.
Step 2 — Price both instruments in the same units (20 minutes)
Founders compare an interest rate to a valuation and conclude nothing, because the units differ. Convert both to dollars.
Cost of the debt. Rate plus fees, times the amount, times the years you will hold it. $1.15m at 14% all-in for two years is roughly $322,000. Then it is gone.
Cost of the equity. The percentage sold, times what the company is plausibly worth when you exit. Selling an extra 9% of a business that later sells for $60 million costs $5.4 million. Selling it in a business that fails costs nothing — which is precisely why equity is the right instrument for things that might fail.
Write both numbers on one line. The comparison is now honest, and it usually settles the argument fast.
The trap: “Non-dilutive” is a marketing word, not a description of risk. Credit does not take your ownership; it takes your optionality. It sits ahead of you in a liquidation, it usually carries covenants, and a borrowing base sized on your metrics shrinks exactly when your metrics do — which is exactly when you need it. Debt converts a bad quarter into a solvency event faster than equity ever will. Use it for the predictable half. Never use it to buy time you have not earned.
Step 3 — Pick the instrument (30 minutes)
The known-return bucket has more options than most founders realise. Roughly cheapest to dearest:
- A bank line of credit or overdraft. Cheapest money available. Requires trading history and usually profitability or hard collateral. Ask first, always — people skip this because it is unglamorous.
- Invoice finance / factoring. Advances against issued invoices. Priced on your customers’ credit, not yours — useful if you are small but sell to large, reliable buyers.
- Inventory or asset-backed lending. Secured against stock or equipment. Fine for physical businesses with real turns.
- Revenue-based financing. Repaid as a percentage of monthly revenue. Flexible, and expensive — a “1.35x over 12 months” deal is not 35%; because you repay continuously, the effective annual rate is roughly double. Always convert to an APR before comparing.
- Venture debt. Usually only available alongside or after an equity round, often with warrants and covenants tied to your cash runway. Read the material adverse change clause properly.
- Growth/customer-value facilities. Capital lent against proven cohort economics to fund acquisition. Structurally the closest thing to “debt for growth,” and the underwriting is only as good as your cohort data.
Trade credit deserves its own line: negotiating 60-day terms with a supplier is financing, it is free, and nobody sends a term sheet for it.
Step 4 — Diligence the lender (40 minutes)
Ask these before signing anything. Written answers, not a call.
- “What is the all-in APR, including origination, monitoring, unused-line and exit fees?” Insist on one number.
- “What are the covenants, and what happens the first time I breach one?” Cure periods and remedies matter more than the rate.
- “How is the borrowing base calculated, and how quickly does it reset if my metrics fall 20%?” This is the question that separates a facility from a trap.
- “Is there a material adverse change clause, and who decides what qualifies?”
- “Where do you sit relative to my existing lenders and shareholders?”
- “What does this do to my next equity round?” Some structures make a priced round harder. Your future lead will read this document.
- “Can I speak to two founders you funded who then had a bad quarter?” The best question on the list. Anyone is pleasant while you are paying.
The five failure modes
- Financing hope. Using debt for the uncertain bucket because equity felt too dilutive. The spend fails, the repayment does not care, and you raise a distressed round anyway — at a worse price, with a lender ahead of you.
- Comparing the wrong numbers. A 1.3x factor rate looks smaller than 20% and is usually far larger. Convert to APR or you are guessing.
- Stacking. Three small facilities from three providers, each cheap, together unrepayable. Lenders now check for this. Some borrowers still do not.
- Ignoring the covenant cliff. The facility is sized against metrics you assume will hold. Model it at 70% of plan before you sign, not after.
- Skipping the boring option. Founders will run a six-week revenue-based-financing process without once asking their existing bank, or their supplier for 45 extra days. Do the free things first.
The 20-minute version
If you do nothing else: take your funding number, draw a line down the page, and put every line item on the side where it belongs — evidence on the left, hope on the right. Total both columns. Then ask whether the round you are about to raise matches that split.
If it does not, you are about to sell ownership to pay for something a lender would have funded at a rate with an end date. That is the single most expensive routine mistake in company building, and it takes two hours to avoid.