Skip to content

Unit economics

How to answer investor questions about unit economics

Updated

Unit-economics questions are about one customer. Can that customer be worth more than it costs to win and to serve, on a timeline that does not require you to subsidize them forever? Partners are not asking for a decorative ratio. They are asking whether the model still works when the assumptions move.

Define the words before they do

People in the room will use the same labels for different stacks of costs. Say yours out loud. Gross margin is revenue minus the cost of delivering the product. Say what you included: hosting, payments, support, implementation. Contribution after variable cost to serve is what is left once the costs that arrive with each new customer are in. CAC is fully loaded. The founder's time is not zero, and a discount to close the logo is part of the cost. Payback is how many periods of gross profit it takes to earn that CAC back. Retention changes the answer. A customer who churns before payback never returns the cost, even if a spreadsheet still shows a healthy month 36.

Do not borrow a target ratio and paste it on the slide. Compute the one you have, say what is inside it, and say what would make it fail.

Founder-sold is not the channel

If every customer came from your network, say so. Do not drop that cost into a model of hired sellers or paid acquisition. The Series A checklist is where this excuse expires: the round is supposed to fund a motion that is not you. At pre-seed you can be honest that the channel is you, as long as the deck does not pretend otherwise.

Cohorts, services, and the marketplace version

A rising customer count can hide a worsening recent group. Show the cohort you have, including a short one, rather than a cumulative total. If onboarding fees or custom work sit inside the number you are calling recurring, split them out before someone else does. That split is one of the SaaS mistakes that shows up the moment a model is opened.

On a marketplace, the take rate is not your margin. You can show a healthy fee and still be subsidizing one side, or charging more than the supplier keeps. The public Nexus sample is a board memo that proposes raising take rate from 15% to 28%. The dossier scores 86/100, and the unit-economics axis is 95, higher worse. The objection is not that monetization is illegitimate. It is that the higher fee can exceed what suppliers keep, so repeat orders leave the platform. The marketplace guide walks the surrounding mistakes. Those sample figures are the memo's proposal and the dossier's score, not a benchmark for your category.

When you do not have the number yet

State the assumption, the threshold that kills it, and the experiment that will produce the number. We will optimize acquisition cost after the round is the polite dodge the last-minute checklist tells you to delete. The Fragility Index treats unit economics as one of five axes, not as the whole company. Read how that score is built so you do not treat a single crater as a grade of the founder.

Paste the model narrative into the free demo audit and see which assumption the questions attack. Then rehearse the spoken version: acknowledge the weak number, pivot to the mechanism that changes it, and prove it with the cohort or the test you actually have.

Questions founders ask

What if payback takes a long time?
Say the number of periods, say why, and say what has to change: price, retention, or a cheaper channel. A long payback is a plan only if the path to a shorter one is specific. A customer who leaves before payback never pays it back, no matter what a later month in the model says.
Do I need a lifetime-value ratio at pre-seed?
Do not paste a ratio you borrowed from a blog. If you have no cohorts, you do not have that number. State the assumption, the threshold that would kill it, and the experiment that will produce the figure. Inventing a lifetime value is worse than not having one.
Which costs belong in gross margin?
The costs that show up because you delivered the product to that customer: hosting, payment fees, support that scales with accounts, implementation you cannot skip. If you are unsure, list them on the slide. Moving a cost into operating expenses so the margin looks higher is how the slide falls apart in the room.
Is a marketplace take rate the same as margin?
No. The take rate is the fee you charge. Margin is what remains after the costs of running the transaction. A high fee can also be more than the supplier keeps, which is a different problem from your own gross margin.