Economics, markets & strategyPart III

The three scalability tests

A business is scalable only if the market is big enough, the unit economics work, AND it can deliver at scale.

Most "is this scalable?" conversations conflate three different questions. They need to be unbundled because the answer can be yes on two and no on one, and the business still doesn't scale.

Test 1: Is the market big enough?

The TAM question. Even a perfect business with 100% market share has a ceiling. Selling enterprise software in a country with 50 enterprises has a ceiling. Selling pet food in the US doesn't.

Practically: estimate the realistic top size of your specific addressable market. Not the global TAM your investor deck claims. The actual addressable opportunity for your specific product, in the specific segments you can realistically sell to. If the answer is "if we win everything, we're still small," scaling isn't possible no matter how well you execute. You have a lifestyle business, which is fine. It just isn't scalable.

Test 2: Do the unit economics work?

Each transaction has to make money, after all the costs of acquiring and serving that customer. The numbers vary by industry but the principle doesn't:

• LTV (lifetime value of a customer) needs to be meaningfully higher than CAC (cost to acquire them). The ratio matters; 3x LTV/CAC is a common rule of thumb.
• Gross margin per unit needs to be high enough to cover overhead at reasonable scale
• The contribution margin can't depend on subsidies or magic future leverage

If you're losing money per unit, scaling means losing more money faster. WeWork-style. The unit economics are the thing the spreadsheet has to show, and they have to be honest. Future-state unit economics don't count.

Test 3: Can we actually deliver at scale?

Some businesses look great at 100 units and fall apart at 100,000. Operational complexity grows non-linearly. Quality drops. Hiring breaks. Customer experience degrades.

Things that often break at scale:

• Customer service (a 24-hour response time at 100 customers becomes a week at 10,000)
• Quality control (what the founder did personally can't be done by 50 people)
• Supplier relationships (the hand-picked supplier can't handle the new volume)
• Culture (the original team's instincts can't be transmitted to 500 hires by osmosis)

All three tests have to be yes. Many businesses pass two and fail the third, but believe they can fix it later. They usually can't, because by the time the third one matters, the company has hired and committed in ways that lock in the failure mode.

A useful rule: if you can't honestly answer yes to all three tests, you don't have a scalable business yet. You might have something valuable, but it's not the same thing.

Examples in the wild