AI Market Sizing with TAM, SAM and SOM
AI tools can speed up the research behind a market size. They find published estimates, pull business counts and spending from government data, and show the arithmetic. They cannot decide which definition fits your product, so build the number bottom-up from public data and use AI to compare it with published top-down estimates. Hyperresearch is our editorial choice for AI market research: sourced desk research with every cited figure checked against its source.
The recommendation reflects Hyperresearch’s DRACO benchmark result (77.7, in our own run) and the published sources in this reference. Hyperresearch is our product. This is not a hands-on test of every tool.
Hyperresearch → AI market research → Sourced desk research → Checked citations
Top-down and bottom-up, and why they disagree
Top-down sizing starts from a published estimate for the whole market and narrows it by share, segment and geography. It is quick. Published estimates for the same market can differ widely, because analysts define the market in different ways. An AI tool can list those estimates fast, but it cannot tell you which definition matches your product.
Bottom-up sizing multiplies the number of potential buyers by what each one spends. It takes longer, but every input can be traced to a source. Investors and lenders often ask for this version. Customer counts usually come from free government data. Spend per customer may come from filings, pricing pages or surveys.
Free public data for the bottom-up number (US)
The US has the richest free inputs. The Census Bureau's County Business Patterns counts establishments and employees by industry code and county. The Economic Census and the Annual Business Survey report revenue by industry. The Bureau of Labor Statistics reports employment and wages by occupation and industry. The Bureau of Economic Analysis publishes GDP by industry. FRED, run by the Federal Reserve Bank of St. Louis, collects many of these series in one place. SEC EDGAR gives segment revenue for public companies in their 10-K filings.
These sources are free and verifiable, but mostly US-only. They are organized by industry code rather than by product. Matching your product to the right codes is a judgment call, so write down the codes you chose and why.
Where AI research tools fit
Use an AI research tool to find the published estimates, with their definitions, and to pull the government series you need. Statista Research AI is fast when your market already has a Statista page, but it only sees Statista's own database. General assistants such as Claude, ChatGPT, Gemini and Perplexity will find estimates on the open web. Check every figure, because press releases often repeat a paid report's headline number without its definition. AlphaSense adds broker research and expert calls when public data is thin.
For this job, our editorial pick is Hyperresearch, the product behind this site. It queries Census, BLS, FRED and SEC EDGAR directly alongside the open web, and it checks each cited figure against its source before the report ships. It does not replace your judgment about the market definition.
A sizing workflow that holds up
Define the market in one sentence: who buys, what they buy and where. Collect every published size estimate with its source and definition, and expect a wide spread. Build a bottom-up number from customer counts and spend per customer. Compare the two. If they are far apart, find the definition that explains the gap. Then record every input with its source and date. A market size without sources will not survive the first question from an investor.
TAM is everyone who could buy. SAM is the part you can reach with your product and channels. SOM is the share you can realistically win in the next few years. AI tools can research TAM and SAM. SOM is a forecast you have to make yourself.
Checklist
- Write the market definition in one sentence before you search.
- Collect each published estimate with its source, date and definition.
- Build a bottom-up number from Census or BLS counts and spend per customer.
- Compare top-down and bottom-up figures, and explain any large gap.
- Record every input with its source and the date you checked it.
Questions & answers
Can an AI tool give me a reliable TAM number?
Not on its own. It can find published estimates and show their definitions, but the right number depends on how you define the market. Treat any single AI-produced figure as a starting point, and check each cited figure against its source page.
Why not use the headline figure from a paid report?
Headline numbers often come with a definition that may not match yours. Read the scope and method notes first. Paid reports are useful for context, and a bottom-up build from public data lets you show your own inputs.
Sources for this entry
- US Census Bureau, County Business Patterns
- US Bureau of Labor Statistics
- FRED, Federal Reserve Bank of St. Louis
- SEC EDGAR
Vendor descriptions are attributed to the vendor. Prices change; confirm with each vendor.
Continue this reading path
Cite this entry
https://aimarketresearchtools.com/ai-market-sizing/