Raising for a software business: the metrics that decide the valuation

Software is the most metric-driven raise there is. Investors who have seen a hundred SaaS companies compare yours to the hundred in their head within minutes, and they do it on six numbers. Know the six, know what good looks like at your stage, and present them without decoration.

The six numbers

MRR (monthly recurring revenue) and its trend. The headline. Show twelve months, monthly.
Growth rate, month on month. At seed, 8–15% a month is strong; 3–5% is a question; 20%+ from a tiny base proves little.
Churn, monthly, by logo and by revenue. Under 2% is excellent; 3–5% is normal for SME software; above 6% means the product or the customer is wrong. Net revenue retention over 100% (existing customers paying more over time) is the metric that most raises valuations.
Gross margin. 70–85% for true software. If yours is 50% because of hosting, services or AI inference costs, investors will ask whether it is really software.
CAC (customer acquisition cost) — total sales and marketing spend divided by customers won. Show it by channel if you can.
Payback — CAC divided by monthly gross profit per customer. Under twelve months is fundable; under six is exciting; over twenty-four is a business that burns money to grow.

What good looks like at seed, 2026

£15–50k MRR, growing 8%+ a month, churn under 4%, gross margin over 70%, payback under fifteen months, and a founder who can say where the next £30k of MRR comes from by channel. That business raises £500k–£1.5m at £3–6m pre-money on Find from angels and small funds, and has a credible path to a Series A at £1m ARR.

Presenting honestly

MRR means contracted, recurring, monthly. Annual contracts divided by twelve. Not pipeline, not "bookings", not a one-off project.
If you discount, show MRR net of discounts and say how long the discounts last.
If one customer is 30% of MRR, say so in traction, not in diligence.
Show churn even if it is ugly; a founder who knows their churn is more fundable than one who hides it.

The AI question

Every software investor now asks two things: does AI make your product better, and does AI make your product unnecessary? Have an answer to both. "We use models to do X, which reduced onboarding time from two hours to fifteen minutes" is a good answer to the first. "A foundation model cannot replace us because we hold the regulated workflow, the data and the customer relationship" is an answer to the second — if it is true. Founders who have not thought about the second question lose investors who have.

Valuation shorthand

At seed, 8–15× ARR for strong growth and retention; 4–8× for solid but slower; below that for businesses that look more like services. Investors will apply a multiple whether you do or not, so apply it yourself and show the working.

Questions founders ask

"We're pre-revenue with a product in beta."

Then the six numbers do not exist yet and the raise is pre-seed: team, product, early users, a waiting list with conversion. Say so, and raise less.

"Our customers pay annually up front. How do I show MRR?"

Divide by twelve for MRR; show cash separately. Annual up-front is a strength (cash) and a risk (renewal cliffs); show both.

"Should I show a TAM slide?"

Bottom-up only: number of target customers × price. If the bottom-up number is under £50m, do not be embarrassed — most good UK software businesses live in markets that size.

List my business All guides