Market Sizing for a New Category
Market sizing for a new category begins with a definition that two people can apply the same way. Set the customer, product, geography, time period, revenue boundary, substitutes, and exclusions before calculating a number.
Why new-category sizing is difficult
A new category has no settled boundary. Suppliers may use different names, customers may buy a bundle, and public data may describe the underlying activity rather than the proposed market. The first job is to make the category observable.
Write a one-sentence inclusion rule and a one-sentence exclusion rule. Then test them against real companies, products, transactions, and customer budgets. If analysts classify the same example differently, the market definition is not ready.
What should the market definition contain?
A usable definition names the buyer, the paid-for offering, the use case, the geography, the period, and the measurement basis. It should distinguish revenue, bookings, users, capacity, shipments, or spend. These measures may all be useful but they are not interchangeable.
Record adjacent categories and substitutes separately. A narrow service can appear small because buyers purchase it inside a broader contract. That does not justify adding the entire contract to the total. It calls for a transparent allocation rule and a sensitivity range.
Which sources and methods should be combined?
Use more than one evidence route and make each assumption visible. A top-down estimate can anchor the broader economic context. A bottom-up model can connect customers, price, adoption, capacity, or supplier revenue. Interviews and surveys can test behaviour that public data cannot show.
Public sources such as the [U.S. Census Economic Census](https://www.census.gov/programs-surveys/economic-census.html) and [OECD data](https://data.oecd.org/) may help define industries and activity, but the analyst must document what is directly measured and what is inferred. Do not present an estimate as an observed statistic.
How should a bottom-up model be built?
Build the model from observable units. Define the number of potential customers, addressable share, adoption or use, price or spend, and timing. Keep each input separate so the reader can challenge one assumption without rebuilding the whole model.
Use ranges where evidence is uncertain. Explain the reason for the low, base, and high cases. A model that shows its uncertainty is more useful than a precise number whose inputs cannot be checked.
| Method | Best fit | Strength | Risk |
|---|---|---|---|
| Top-down | Broad context | Fast boundary anchor | Category mismatch |
| Bottom-up | Defined customer and price | Traceable assumptions | Input uncertainty |
| Supplier triangulation | Fragmented provider market | Revenue and capability view | Reporting bias |
| Primary research | Unobserved behaviour | Customer reality | Sample and response bias |
How should estimates be validated?
Validation is a challenge process, not a search for a confirming number. Compare methods, interview participants with different incentives, test classification with edge cases, and reconcile differences between supplier, customer, and public-data views.
Keep a source register with date, geography, definition, unit, transformation, and limitation. Recalculate the estimate when the definition changes. Never silently replace an old boundary with a new one while leaving the headline unchanged.
How should the result be presented to a decision-maker?
Lead with the definition, range, method, and decision use. Show what is included, excluded, observed, inferred, and still uncertain. A short bridge from the market total to the serviceable segment is more valuable than an unsupported growth curve.
Include a scenario table, an assumptions page, a source register, and a list of questions that would change the decision. If the buyer needs a market entry decision, connect the size to customers, competition, economics, and access rather than stopping at total addressable market.
What does not matter as much as buyers think?
A larger market number is not automatically a better strategy. Inflating the boundary by adding adjacent spend can make the opportunity look attractive while hiding the reachable customer and delivery model.
Do not over-polish the decimal places. Spend the effort on definitions, sources, sensitivity, and the decision the estimate must support.
How to turn this into a research brief
Turn the question in this guide into a brief with a fixed boundary. For market sizing for a new category, name the audience, decision, geography, time period, evidence standard, and output the team needs. State what is outside scope so a broader market label cannot quietly change the assignment.
The brief should let another analyst reproduce the route from question to conclusion. Keep a source register, an assumptions log, a list of unresolved questions, and a clear review point. That discipline makes the final work easier to use and easier to challenge. Record the decision rule and the date when the evidence should be refreshed.
- Define the decision: write the action the work must support.
- Set the boundary: specify buyer, offering, geography, period, and exclusions.
- Map the evidence: separate observed data, expert input, inference, and assumption.
- Choose the method: match desk research, interviews, surveys, modelling, or testing to the question.
- Set quality gates: decide what must be verified before a conclusion is accepted.
- Design the output: show the comparison, scenario, decision rule, and next action.
What should a strong brief leave unanswered?
A useful brief does not hide uncertainty behind a polished headline. It makes clear which parts are known, which are estimated, which depend on the buyer’s operating model, and which need primary research. Readers should be able to see what would change the recommendation.
Before commissioning the work, check that the team can answer these questions: who will use the result, what decision is pending, what evidence is acceptable, what alternatives must be compared, which risks are material, and what action follows. If the answer to one is missing, narrow the assignment rather than padding the report.
- Decision owner: who can act on the finding?
- Evidence boundary: what counts as verified?
- Alternative view: which credible option could disprove the first answer?
- Operational test: what must work in practice?
- Uncertainty: which assumption most affects the result?
- Next step: what happens after the report is read?
FAQ
What is the first step in new-category sizing?
Write the inclusion and exclusion rules and test them against real offerings and customer purchases.
Should a market estimate be a single number?
Usually present a range with a base case, assumptions, sensitivity, and a clear measurement unit.
Can public data prove a new market size?
Public data can provide context, but the analyst must explain the mapping, transformation, and inference used for the proposed category.
How do you avoid double counting?
Define the revenue or spend boundary, classify bundles consistently, and keep adjacent categories outside the total unless an explicit allocation rule applies.
What makes a sizing model credible?
A clear definition, traceable sources, visible assumptions, independent challenges, and a direct link to the decision it supports.
Sources and related research
Use the following public references to frame the question. They are starting points for evidence and governance, not substitutes for a study specific to the buyer’s scope.
Continue with Market Research, Market Sizing Methodology Buyer Guide, Custom Market Research Brief How To Write.
Request a focused brief through Talk to an analyst.