Business model slides: simple math beats aspirational TAM
A business model slide should prove one customer’s repeatable economics before it claims a giant market.
The slide said $1.4 trillion. The founder couldn't explain one customer.
A founder is mid-pitch. The business model slide is up. It shows a $1.4 trillion TAM, a sliver labeled SAM, a smaller sliver labeled SOM, and a sentence about capturing one percent of a massive market. The partner across the table asks a small question: "Walk me through what happens economically when one customer signs up. What do they pay, and what does it cost you to serve them?"
The founder freezes. The deck has a market the size of a continent and no answer for a single customer. The pitch does not recover, because the partner just learned the thing that matters most: the founder has been thinking about the market and has not yet thought about the business.
This is the most common failure of the business model slide. It answers a question nobody serious is asking ("is the market big?") and skips the question every investor is actually holding ("does a customer make you money, and does that get better as you grow?"). A big TAM with no unit economics is a slide that says "we have not figured out the business yet" in expensive-looking font.
Why the TAM-heavy slide fails
Aspirational TAM fails for a reason that has nothing to do with the number being wrong. It fails because it is not falsifiable and not yours. Anyone in your category can put up the same $1.4 trillion. It tells the investor nothing about whether you specifically have a business. It is a claim about the world, not a claim about your company.
Unit economics are the opposite. They are small, specific, and yours. The price one customer pays, the cost to serve that customer, the margin left over, the time it takes to earn back what you spent to acquire them. These numbers are checkable, they are defensible, and they are the actual content of "do you have a business." A founder who can say "one customer pays us $X, costs us $Y to serve, leaves us $Z in gross margin, and we earn back acquisition cost in N months" has shown more than any market chart can.
Investors triangulate the big market on their own. They have seen the category, they have their own TAM instinct, and a believable bottom-up build matters more to them than a top-down number you sourced from a consulting report. What they cannot triangulate without you is your unit. So the slide should spend its space proving the unit, not decorating the market.
The framework: prove economics at the smallest useful unit
Every business has a smallest unit that has its own economics. Find it, and the slide writes itself. The unit is the thing you can describe paying you money and costing you something to serve, in isolation, before you multiply by the market.
To define the unit, answer five questions in order. These five are the entire business model slide. Everything else is decoration.
- Unit of value. What is the one thing the customer buys? A seat, a transaction, a workflow run, an installed location, a delivered project. Name the noun.
- Buyer. Who actually pays and who holds the budget? The user and the buyer are sometimes different people. The slide should name the one who signs.
- Pricing axis. What does the price scale with? Seats, usage, GMV, outcomes, fixed subscription. The axis is your bet on how value grows for the customer.
- Gross margin. What does it cost you to deliver one unit, and what is left? For software this is hosting and support. For AI it is inference. For services it is people. State the cost honestly.
- Payback. How long until the gross margin from one customer earns back what you spent to acquire them? This is the number that tells an investor whether growth funds itself or burns capital.
If you can answer those five with real numbers, you have a business model slide. If you can answer them with honest estimates, you have a business model slide for an early round. If you cannot answer them at all, the problem is not the slide. The problem is that you have not decided how the business works yet, and the slide is doing you a favor by making that visible.
Writing the slide before your data is mature
Most founders raising pre-seed and seed do not have a year of cohort data. That is fine. Investors at that stage are not auditing your numbers. They are auditing your thinking. The slide before data maturity is allowed to use estimates, but it must do two things: label them as estimates, and show the logic that produced them.
The move is to write the slide as a hypothesis with its math exposed. "We charge $500 per seat per month. A typical customer starts at 4 seats, so $2,000 in monthly revenue. Hosting and support run roughly 20 percent, leaving 80 percent gross margin. We estimate acquisition cost around $6,000 from early outbound, which pays back in under four months at current margin." Every number there is a guess. But the structure is rigorous, and the investor can poke at any single input and watch the conclusion move. That is what a business model slide is for: not certainty, but a model someone can stress-test in the room.
Never invent the numbers and present them as fact. If you have three customers and one price point, say so. A labeled estimate with visible logic beats a confident fabrication every time, because the moment an investor catches one invented number, they discount all of them.
The mini-model: the artifact behind the slide
Before you design the slide, build the model the slide summarizes. It fits in ten rows of a spreadsheet and it is the thing you actually defend in the meeting. The slide is the headline. This is the body.
| Line | Input | Example (SaaS) | Notes |
|---|---|---|---|
| Unit of value | one thing sold | seat | name the noun |
| Price per unit | what one unit costs the buyer | $500 / seat / mo | your pricing axis |
| Units per customer | typical starting size | 4 seats | land size, not expansion |
| Revenue per customer | price × units | $2,000 / mo | the top line of one customer |
| Cost to serve | variable cost per customer | $400 / mo | hosting, support, infra |
| Gross margin % | (rev − cost) / rev | 80% | the honest number |
| Gross margin $ | revenue − cost | $1,600 / mo | what one customer contributes |
| CAC | cost to acquire one customer | $6,000 | blended, labeled as estimate |
| Payback (months) | CAC / monthly gross margin $ | 3.75 | self-funding test |
| Expansion path | how the unit count grows | seats → teams → org | the second act |
Fill this in for your own business and the slide becomes a transcription job. The mini-model is also what you hand an investor who asks for detail after the meeting, which turns a pricing question into a document instead of a scramble.
The slide layout
The slide itself should hold four blocks and nothing else. Resist the urge to put the market chart here. The market lives on a different slide.
+-------------------------------------------------------------+ | BUSINESS MODEL | +-------------------------------------------------------------+ | | | HOW WE MAKE MONEY ONE CUSTOMER, ECONOMICALLY | | We charge [buyer] per Revenue: $2,000 / mo | | [unit] on a [pricing axis] Cost to serve: $400 / mo | | basis. Gross margin: 80% | | CAC: $6,000 | | Payback: ~4 months | | | +-------------------------------------------------------------+ | WHY THIS COMPOUNDS STATUS OF THESE NUMBERS | | Customers expand from 3 live customers, 1 price. | | [unit] to [larger unit]. CAC and payback estimated | | Net revenue grows without from early outbound. Margin | | proportional cost. measured, not modeled. | | | +-------------------------------------------------------------+
Top left states the model in one sentence. Top right is the unit economics, the four or five numbers that are the real content. Bottom left is the expansion story, how one unit becomes more units over time. Bottom right is the honesty block: what is measured, what is estimated, how many customers stand behind it. That last block is what makes a sophisticated investor trust the other three.
Four worked examples
The unit changes by business type. Here is how the same five questions resolve across four common shapes.
SaaS. Unit is a seat or a workspace. Buyer is a team lead or department head. Pricing axis is seats or tiers. Gross margin is high, 75 to 85 percent, because the variable cost is hosting and support. Payback should sit under 12 months at seed, ideally under 6 for early outbound. The expansion story is seats to teams to whole-org, which is why net revenue retention is the number that follows this slide.
Marketplace. Unit is a transaction. Buyer is whichever side you monetize, often both. Pricing axis is take rate on GMV. Gross margin looks high per transaction but the real economics live in liquidity: how often a listed item or available provider actually transacts. The honest version of this slide shows GMV per active user and take rate, not just the percentage. Payback is measured against the cost of acquiring the constrained side of the market.
AI usage-based. Unit is a workflow run, a generation, a processed document. Buyer is the operator who owns that workflow. Pricing axis is usage, sometimes with a platform fee. The number investors will hammer is gross margin, because inference cost is real and variable in a way SaaS hosting is not. If your margin is 55 percent because every run hits a model API, say 55 percent. A founder who pretends AI has 85 percent SaaS margins gets caught and loses the room. Show the cost per run and the path to lowering it as volume grows.
Services-enabled software. Unit is a delivered outcome or a managed account. Buyer is the budget holder for that outcome. Pricing axis is often a fixed fee plus usage, or a retainer. Gross margin is the hardest part because people deliver the value, so margin starts lower, 40 to 60 percent. The slide that wins here shows margin improving as software absorbs work that humans do today. The model has to show the trajectory, not just the snapshot, because the whole thesis is that the services layer shrinks over time.
Where this connects to the rest of the raise
The business model slide does not end in the deck. It ends in a follow-up. When a partner pushes on your margin or your payback assumption, that pushback is the most useful signal in the meeting, because it tells you exactly which number to firm up before the next conversation. The mistake founders make is letting that question evaporate the moment the meeting ends. Three weeks later a different investor asks the same thing and the founder has the same shaky answer, because nobody turned the objection into work.
Treat every pricing and economics question as a task with the investor attached. "Northpeak pushed on inference margin, wants to see cost per run trending down. Send the updated mini-model before the partner meeting." That is the loop that turns a weak slide into a strong one over a few weeks of conversations.
RoundOS keeps that loop intact. Investor questions about pricing and economics stay connected to the follow-up task they should produce, grounded in the meeting note where the question was actually raised, so the next version of your business model slide answers the objection the last investor handed you instead of guessing what the next one will ask.
Prove the unit before the market.
Replace the TAM-first slide with the ten-row mini-model and let the simple customer math carry the story.