# Describe how you’d conduct a market sizing exercise for a new product or market segment.
> How to answer a market sizing question as a PMM, with a bottom-up method, a top-down check, a worked leak-sensor example, and the numbers not to claim.

Source: https://mokk.ai/pmm-questions/market-research/market-sizing (updated 2026-09-22)
Markdown: https://mokk.ai/pmm-questions/market-research/market-sizing.md

Define the buyer and the unit first, then build bottom-up by multiplying the number of buyers by the share that would plausibly buy and what each would pay per year. Cross-check against a top-down figure or a comparable product. Give a range with a likely case, name the weakest assumption, and say how you would test it and which decision the number supports.

## What a market sizing prompt is scoring

Nobody expects a correct number. What gets scored is how you define the market, whether your assumptions are visible, and whether you know what the number is for. Analytical rigor carries most of the score, with market expertise showing up in the assumptions you choose.

Ask one clarifying question before you start. The answer changes depending on whether the number decides a go or no-go, sets a revenue target, or sizes a sales team. A sizing for a go decision can tolerate a wide range. A sizing that sets quota can’t.

## Define the buyer before multiplying anything

Most sizing mistakes happen in the first thirty seconds, when the market is left vague. “Smart home devices” is too loose to count. “Owners of small apartment buildings who manage their own properties” can be counted.

- **Who buys.** The person or organization that pays, which may differ from the user.
- **What they buy.** The unit, such as a device, a seat, or a yearly subscription.
- **Where and when.** Geography and time frame, usually one country and one year.
- **Total, serviceable, obtainable.** Everyone who could buy, those this product and channel can reach, and a realistic share within a few years.

## Bottom-up sizing first, then a top-down check

1. Count the buyers from a source you could name, such as census or registry data.
2. Narrow to those with the problem and the budget.
3. Estimate what each pays per year, including recurring revenue.
4. Multiply, then build low, likely, and high cases.
5. Compare with a top-down figure, such as category spend or a similar product’s revenue, and explain any large gap.
6. Name the least certain input and how you would test it.

Talk while you calculate. Silence during arithmetic reads as struggle, while narrating each step lets the interviewer correct an assumption early. Round hard. Forty thousand is far easier to multiply out loud than thirty-eight thousand six hundred, and nobody will mark you down for it.

## A sizing walk-through for a leak-sensor subscription

**Strong answer (invented example)** "Let me define it first. The product is a water-leak sensor kit with a monthly monitoring plan, sold to owners of small apartment buildings in the US who manage their own properties. I’ll count buildings with five to twenty units. I’d pull the real figure from housing survey data, and for now I’ll use about five hundred thousand buildings. Maybe half are self-managed, so two hundred fifty thousand. Each building needs roughly one sensor per unit plus the boiler room, call it twelve sensors. At forty dollars per sensor up front and fifteen dollars a month per building, the first year comes to about six hundred sixty dollars per building. Across the serviceable base that’s around a hundred sixty-five million dollars in year one. For a top-down check, I’d look at what small landlords lose to water damage claims, because the product only sells if it costs less than one bad leak. My range would be a hundred to two hundred fifty million. The weakest input is the self-managed share, so I’d test it with a quick survey through landlord associations before anyone sets a target."

The top-down check in that answer does double duty. Tying the market to the cost of the problem shows market expertise, and it sets up the pricing follow-up before it arrives. For more practice with open-ended prompts like this one, see the [case study interview guide](https://mokk.ai/guides/pmm-case-study-interview).

## Numbers a sizing answer shouldn’t claim

- **That serviceable market equals revenue.** A market of a hundred sixty million says nothing about what the company will sell next year.
- **One percent of a huge number.** Taking a slice of an industry figure skips every assumption that matters, and interviewers have heard it many times.
- **Precision the inputs can’t support.** Saying 163.4 million implies certainty you don’t have. Round, and give a range.
- **Buyers without budget.** Counting users who can’t approve spend inflates the market, especially in schools, hospitals, and large companies.

## Market sizing follow-ups

Follow-ups after a market sizing answer
| Follow-up question | The skill in question | Handling it |
| --- | --- | --- |
| Which assumption worries you most? | Self-awareness about uncertainty | Name one, explain why, and say what happens to the total if it were half as large |
| How would the number change for large property firms? | Whether you can resegment | Walk through the new buyer count, the longer sale, and the higher price per account |
| How does this compare with the total market? | Understanding of total versus serviceable | Explain what you excluded and why |
| What would you do with the result? | Link to decisions | Say whether it clears the bar for investment, and what it implies for channel or pricing |
| What if a published report says the market is ten times bigger? | Handling conflicting evidence | Compare definitions first, since reports often count adjacent products |

## More questions in this category

- [Building buyer personas](https://mokk.ai/pmm-questions/market-research/buyer-personas)
- [Analyzing a funded competitor](https://mokk.ai/pmm-questions/market-research/analyze-well-funded-new-entrant)
- [Useful sales battlecards](https://mokk.ai/pmm-questions/market-research/battlecards-sales-will-use)
- [Misleading comparison page](https://mokk.ai/pmm-questions/market-research/competitor-misrepresents-comparison-page)
- [All market research questions](https://mokk.ai/pmm-questions/market-research)

## Frequently asked questions

**Should I memorize population and industry figures?** A few rough anchors help, such as the US population or the number of households. Beyond that, a stated assumption works fine. Interviewers care more that you say where you would verify the number.

**Can I use paper or a calculator during a sizing question?** Usually, yes, and it is fine to ask. On a video call, narrate the arithmetic even while you write it down, so the interviewer can follow along.

**What if my estimate turns out wildly off?** If the logic is sound and the assumptions are visible, a wrong total costs little. What loses points is a number nobody can trace. If the interviewer supplies a real figure, adjust the input it corrects and keep going.

## Practice

**Answer it out loud.** A Mokk session asks this question, follows up on your answer, and scores it on five competencies.