Every pitch deck has a TAM slide. Most of them are wrong. Ask an AI for “the TAM of X” and you get a confident eight-figure number that traces back to a 2022 Statista blurb cited by a 2024 Medium post. A useful market-sizing workflow uses AI to pull both top-down and bottom-up estimates, then triangulates them. The gap between the two is where the real argument lives. This tutorial walks the loop investors actually want to see, with the exact tools, current pricing (June 2026), and a fully worked example.
TL;DR
- Run two independent passes: top-down (start from a public number, narrow with sourced filters) and bottom-up (ARPU × addressable users × adoption). Then triangulate — compare the two and name the gap.
- Use Perplexity to pull cited inputs, ChatGPT or Claude Research to synthesize, and a spreadsheet for the arithmetic. AI does not do reliable multi-step math; you do.
- Public market numbers disagree by 1.3x–5x across firms (real example below). The disagreement is the point, not a problem to hide.
- Output: one slide with TAM / SAM / SOM as three boxes and an assumption table beneath, every input carrying a value, a source link, and the source year.
- Budget 90 minutes for the first pass, 30 minutes for refreshes.
Who this is for
Founders sizing a market for a seed deck, PMs writing a “should we build this” memo, strategy teams scoping a new vertical, and investors who want to sanity-check a deck before a meeting. Not for pure academic market reports — those need primary survey data, not AI synthesis.
When to reach for it
When you have a defined product or category, a target geography, and 90 minutes to produce a credible sizing slide. Skip this workflow when the market is so emergent there are no public numbers; then you are bottom-up only, and the AI does much less. For broader category context first, run an AI industry research workflow pass.
The tools, and why each one (June 2026)
You need a retriever, a synthesizer, and a calculator. Do not collapse them into one chat window.
| Job | Tool | Cost (June 2026) | What it does here |
|---|---|---|---|
| Pull cited public numbers | Perplexity | Free ~5 Pro searches/day; Pro $20/mo ($200/yr) → 20 Deep Research runs/day; Max $200/mo unlimited | Returns figures with named sources and dates, which top-down sizing lives or dies on |
| Synthesize the argument | ChatGPT Deep Research | Plus $20/mo = 10 Deep Research sessions/month; Pro $200/mo = much higher allowance | Multi-source report you can steer to specific sites; built on a GPT-5.2-class research model since Feb 2026 |
| Synthesize the argument (alt) | Claude Research | Pro $20/mo or Max $100/$200/mo | Agentic web + connected-source research; web-search toggle is on every paid tier |
| Do the arithmetic | Any spreadsheet (Google Sheets / Excel) | Free / included | Every input in its own cell with a source URL beside it |
Notes that matter in practice. Perplexity’s free tier (~5 Pro searches/day) is enough to test the workflow but runs out fast on a real sizing; Pro’s 20 Deep Research runs/day is the sweet spot. ChatGPT Plus gives only 10 Deep Research sessions per month — burn them on synthesis passes, not on input retrieval you could do in Perplexity. Claude Research and ChatGPT Deep Research are interchangeable for the synthesis step; pick whichever you already pay for.
Before you start
- Define the unit you are sizing: “Annual revenue from X tool sold to Y persona in Z geography.” Vague unit, vague number.
- Decide TAM versus SAM versus SOM upfront. TAM = global category, SAM = reachable subset, SOM = realistic 3-5 year share. Most decks conflate them.
- Have one bookmarked credible source for the broad category. You will use it as the top-down anchor.
Step by step
- Top-down: pull the anchor with Perplexity. Prompt:
Global market size for [category] in 2025-2026 — list 3-5 estimates from named sources with publication dates.Read the spread. If the high and low differ by 5x, the category is poorly measured. Note that on the slide; do not paper over it. - Apply sourced filters to get from anchor to TAM. Geography (US only? Multiply by a published US share, never a guess). Segment (SMB only? Filter by employee bucket). Use case (paid users only? Filter by a conversion benchmark). Each filter is one row in the assumption table with a source link.
- Bottom-up: start from unit economics. Average revenue per user × addressable user count × adoption rate. Pull each input separately in Perplexity:
Average ARPU for [SaaS category] SMB tier in 2026,Number of SMBs in [geography] with [employee range],Adoption rate of [category] tools in [segment]. - Do the arithmetic in a sheet, not in the chat. AI math fails silently on multi-step multiplication. Put every input in its own cell with a source URL in the next column. Total at the bottom.
- Triangulate. Compare top-down TAM and bottom-up TAM. Within 2x: you have a defensible range. Off by 10x: one model is wrong, usually an over-claimed filter or a stale ARPU. For deeper synthesis on this kind of crosscheck, see the ChatGPT research tutorial.
- Derive SAM and SOM from the triangulated TAM. SAM = TAM filtered by your real go-to-market (channel-reachable, language, regulation). SOM = SAM × a defensible 3-year share assumption, backed by a comparable company’s actual share trajectory.
- One slide, assumption table beneath. TAM / SAM / SOM as three boxes; the assumption table as a small-font table with every input, its value, its source, and the source year.
Worked example: the SaaS-market spread
Here is why the triangulation step exists. Pull the global SaaS market for 2025 and the named sources disagree badly — all published, all credible, all citing different scope definitions.
| Source | Global SaaS market, 2025 | Notes |
|---|---|---|
| Fortune Business Insights | ~$315.7B | Narrower SaaS-application definition |
| Gartner | ~$317B | Software-as-a-service spending |
| Precedence Research | ~$408.2B | Broader cloud-application scope |
That is a 1.3x spread on the same year, same category — and that is a well-measured market. If your category looks tight, you have not pulled enough sources. The lesson: never quote a single number as “the market size.” Quote the range, cite all three, and pick the definition that matches your unit, then say so on the slide. (Every firm above now projects SaaS past $1 trillion before 2035, but bake projected growth into a separate row, never into the TAM box.)
First-run exercise
Pick a market you have first-hand knowledge of — your employer’s, or a hobby industry where you know real ARPU. Run the workflow end to end. Compare the AI-derived number to your gut. Most first runs come back 2-5x too high, usually because the top-down anchor swept in adjacent categories. That is the calibration lesson; you will tighten the unit definition on every future run.
Quality check
- Top-down and bottom-up numbers are within 2x. If not, name the gap explicitly on the slide.
- Every input cell has a source URL with a 2024-2026 publication date. Older sources need a footnote justifying why they still apply.
- SAM is meaningfully smaller than TAM. If SAM equals TAM, you have not filtered for go-to-market reachability.
- SOM has a named comparable. “We will reach 5 percent” with no comparable is not a defense.
- The assumption table fits on one slide. If it does not, you are oversizing the model.
How to reuse this workflow
- Save the top-down filter list and bottom-up input list as a template. New product, new numbers, same skeleton.
- Build a personal database of recurring inputs: ARPU benchmarks per category, SMB counts per geography, adoption rates per segment, each with its source. Update quarterly.
- Re-run the sizing every six months on the same template. Sizing that does not update goes stale by the next funding round.
Common mistakes
- Asking the AI directly for “the TAM of X.” You get a number, not an argument. Investors want the argument.
- Doing the math in chat. Multi-step arithmetic with AI is unreliable; use a sheet.
- Skipping triangulation. Top-down alone overstates, bottom-up alone understates, the truth lives in the gap.
- Quoting one firm’s number as “the market.” As the table above shows, a single source can be off by 30 percent from the next one.
- Conflating TAM and SAM. Investors notice immediately.
- Hiding the assumption table because it makes the model look “less clean.” The table is the model’s credibility.
FAQ
- What if there is no public number for my category?: Bottom-up only, and say so on the slide. Investors prefer “this is bottom-up only because the category is too new” over “TAM is $10B with no traceable source.”
- Which AI tool for which step?: Perplexity for cited input retrieval, ChatGPT Deep Research or Claude Research for synthesis, a spreadsheet for math. Do not collapse the steps into one prompt.
- Is the free tier enough?: For one test run, yes — Perplexity’s free ~5 Pro searches/day cover it. For a real sizing with a dozen inputs and two synthesis passes, you will want Perplexity Pro ($20/mo) and at least ChatGPT Plus, whose Deep Research is capped at 10 sessions/month.
- How fresh do sources need to be?: 2024 or later for input numbers; 2025-2026 preferred for ARPU and adoption rates.
- Should I include growth projections?: Yes, as a separate row with a named source for the growth rate. Do not bake projected growth into the TAM number itself.
- What is a “defensible” share assumption for SOM?: One backed by a comparable company’s actual share trajectory at the equivalent stage. Without a comparable, your SOM is a wish.
Related
Tags: #market-sizing #tam #strategy #Tutorial