Seed-stage startup: SaaS ROI & cash runway model

What actually decides this one

Start with month one, because it sets the terms for everything that follows. Four engineers at a hundred and thirty-five thousand each, sixty hours of contractor time and a baseline cloud bill all arrive whether or not a single customer signs. Twelve thousand of MRR at a seventy-two percent margin does not come close to covering them — the gross profit this business earns in its first month is worth less than a week of payroll, and about a sixth of the fixed base it has to clear. That gap is the entire scenario. Revenue is a rounding error against the cost structure and stays one for the better part of three years. Founders at this stage often describe themselves as running a business. They are running a countdown, and the business is the thing they are trying to finish building before it reaches zero.

Churn near four and a half percent a month is the input most likely to be waved off as normal for the stage, and it is the one doing the most quiet damage. At that rate the average account lasts a little under two years, and more than two in five of any cohort is gone inside twelve months. Acquisition has to refill that hole before it adds anything on top. At nine hundred dollars of CAC against a hundred and eighty of ARPA, a new account needs just under seven months of its own gross profit to pay for itself — tolerable in isolation, and much less comfortable beside a customer lifetime measured in months rather than years. The unit economics clear the usual three-to-one bar, narrowly, and that is genuinely the most that can be said for them.

Here is the part that surprises people, and it is worth watching happen rather than taking on trust. Halve the churn — take it to roughly two and a half percent, leave every other control untouched — and the runway barely moves. A month, give or take. The cash trough is set by fixed costs that the retention rate cannot reach, and by the time better retention has compounded into materially more revenue, the money has already gone. What halving churn does do is rewrite everything after the trough: break-even arrives more than a year earlier and the fifth year ends with a real cash balance instead of a hole. Retention is the right work. It is simply not emergency work, and treating it as the answer to a runway question is how a team spends its last two quarters on the wrong problem.

The control that genuinely moves the runway is headcount, which is the one nobody wants to touch. Going from four engineers to three buys roughly eight months — more than any retention or growth change reachable from this starting point. Going to two makes the model solvent across the full five years without another change anywhere. Cutting the contractor hours to zero buys a few more weeks and costs almost nothing politically, which is why it usually happens first and why it is rarely enough on its own. None of that is a recommendation about who to let go. It is a statement about where the sensitivity lives: if runway is the binding constraint, the cost base is the only control that responds inside the window where the decision still matters.

The re-investment rate is set at forty-five percent and does almost nothing here. The model only pushes operating profit into paid acquisition, and on these inputs there is no operating profit to push until late in year four. Move the control anywhere between zero and ninety and the runway, the trough month and the break-even month come back identical — the only thing that changes is the shape of the final year, and a higher rate actually finishes with less cash on hand, because the cash was spent. It is a clean demonstration of a general point that costs seed companies a lot of planning time: a growth lever that runs on profitability is not available to a company that has none.

Two things would change the outcome outright. Organic growth would have to climb from nine percent a month to nearly thirteen and hold there — a forty percent improvement in the rate, sustained across five years, which is a very different request from the one it sounds like. Or the opening cash would have to roughly double, at which point the trough clears zero by a hair and nothing else about the company has to change at all. That second option is what most seed rounds are actually buying: not fuel for growth, but enough months for the retention work described above to finish compounding. Price the raise against the trough month, not against the plan.

What to move, and what happens when you do

Engineering headcount

The fastest-acting control on the page. Each engineer removed from the four is worth several months of runway, and the model turns solvent well before the team reaches zero. Nothing else responds this quickly.

Monthly revenue churn

Halving it barely touches the trough and completely changes years three through five. Use it to decide what kind of company this becomes, not to decide whether it survives the next four quarters.

Opening cash

Roughly double it and the cash trough clears zero with nothing else altered. This is the control that tells you what the round actually needs to be, as opposed to what the deck says it is for.

Re-investment rate

Inert on this profile. There is no operating profit to re-invest until the fourth year, so the slider moves the ending ARR and slightly reduces the ending cash, and changes neither the runway nor the break-even month.

Organic MRR growth

Would need to reach nearly thirteen percent a month to keep the company solvent on the current cost base. Worth setting once, to see how far the growth answer really is from the cost answer.

Seed-stage startup: five years at the default inputs