Market Sizing: TAM, SAM and SOM, With Worked Examples
You are not just picking a chart type. You are deciding whether the number your audience remembers is the entire market you would need to dominate, or the slice you can actually win, and getting that backwards is why market sizing slides get taken apart in the room. Below: what TAM, SAM and SOM actually measure, when to trust a published figure and when to build your own, a real fight between an investor and a finance professor over how far a top-down estimate can be trusted, the mechanical build, and the one-line route that renders it as an editable native slide.
Quick answer: what is market sizing?
Frame the opportunity as TAM, SAM and SOM, narrowing from the total market down to what you can realistically win. Size it two ways, top-down from a published figure and bottom-up from your own unit economics, then show both so the audience sees the cross-check rather than a single guess.
Market sizing is the process of estimating how big a business opportunity is and how much of it a specific company can capture, usually expressed as three nested figures. TAM, the total addressable market, is the entire revenue available if every potential buyer purchased. SAM, the serviceable addressable market, is the slice of TAM reachable given a company's product, segment and geography. SOM, the serviceable obtainable market, is the share of SAM realistically winnable within a defined period given competition and capacity. Each figure sits inside the one above it, and a credible market sizing slide states the source and filter behind every step rather than presenting the three numbers alone.
When a market sizing slide is the right exhibit
Marc Andreessen put the underlying reason bluntly in 2007: "In a great market, a market with lots of real potential customers, the market pulls product out of the startup." A market sizing slide exists to make the audience believe, with evidence, that the market in question is a great one. That belief is worth an entire slide in some situations and the wrong tool entirely in others.
- It fits whenever you must justify the scale of an ask. Fundraising decks, business cases, market-entry recommendations and budget requests all open with a number, because every later recommendation is judged against it.
- It fits when the number has to survive being challenged live, by an investor, a credit committee or a steering committee that has seen an inflated TAM before and will ask what narrowed it down.
- It is the wrong exhibit if the real question is share of an existing, named market rather than the size of the whole opportunity. Use a competitive landscape slide when the argument is about position among rivals, not the size of the prize.
- It is the wrong exhibit if the axes are two chosen criteria used to force a choice among options rather than to measure demand. That is a 2x2 positioning matrix doing a different job with a similar geometry.
- It is the wrong exhibit if the underlying problem is a root cause or a structural breakdown rather than an opportunity size. An issue tree answers what is broken; market sizing answers what it is worth to fix it.
Key takeaways
- Know your inputs before you pick a method. A bottom-up estimate needs your own customer count, price and purchase frequency; a top-down estimate needs a published figure credible enough to name on the slide. Missing either one rules out the method that needs it.
- Top-down is fast because someone else already did the research; bottom-up is slow because every input is yours to source or estimate. Budget the slower method for any number a sophisticated buyer will actually try to challenge.
- Triangulate rather than trust one number. Two independent estimates that land close together are far harder to argue with than a single point estimate, and a gap between them is a flawed assumption worth finding first.
- Match the method to how established the category is. Top-down and bottom-up both assume the market already exists in roughly its current shape; a genuinely new category needs a defended comparison, not a slice of what already exists.
- Treat every figure as perishable. A published market report is stale within a year and your own unit economics move every quarter, so budget for the re-cut rather than presenting a number you built once and never touched again.
Four ways to size the market, compared
| Method | Inputs needed | Speed to build | Defensibility | Best fit | Re-cut cost |
|---|---|---|---|---|---|
| Top-down TAM, SAM, SOM | A published market figure, plus segment, geography and product filters | Fast, often a single sitting | Weak alone, only as credible as the named source | Early conversations and board updates where a credible third-party figure already exists | Low, swap in a newer source figure |
| Bottom-up TAM, SAM, SOM | Your own customer count, price and purchase frequency | Slow, every input has to be sourced or estimated | High, every number is yours to defend | Fundraising and diligence, anywhere a sophisticated buyer will pressure-test the figure | Medium, needs updated unit economics each cycle |
| Triangulated, both methods reconciled | Both of the above, run in parallel and compared | Medium, two models instead of one | Highest, a reconciled range beats a single number | Any exhibit that has to survive being challenged live, which is most of them | Medium, both models move together |
| Analogous or expansion sizing | A defended comparison to an adjacent market, or a stated behavior-change assumption | Slow, the argument needs defending before the numbers do | Contested by design, wins only if the comparison holds | A genuinely new category with no direct precedent, where the existing market figure understates the opportunity | High, the argument itself needs revisiting, not just the inputs |
How do you size a market top-down?
Top-down starts from a number you did not calculate yourself. Find a published estimate of the whole category, ideally from a named research firm or a public filing, and narrow it with the filters that actually apply to you: the segment you serve, the geography you sell into, and the product category you compete in. Each filter should be a stated percentage or a named exclusion, not a silent adjustment, because the moment a filter goes unstated the number stops being top-down and starts being a guess wearing a citation.
The method is fast precisely because the expensive research already happened inside someone else's report. That speed is also the whole risk: a top-down TAM is only as credible as the report you are citing and the filters you applied to it, and a reader who does not trust either one does not trust the number. Name the source on the slide, not in a footnote nobody reads, and keep the year current, since a three-year-old market report narrowed with current-year percentages is quietly wrong twice.
The trade-off: top-down is the fastest way to put a number on a slide, and the easiest one for a skeptical reader to knock down by asking where the figure came from and why you get to keep the share of it you claim.
How do you size a market bottom-up?
Bottom-up starts from your own arithmetic instead of someone else's report: the number of target customers times the average price times the purchase frequency. Define the customer count first, and define it narrowly. Companies with 50 to 500 employees in a named country is a customer count you can defend; growing businesses everywhere is not. Multiply by a price you actually charge or credibly plan to charge, and by how often the purchase repeats in a year, and you have a serviceable addressable market built entirely from inputs you can stand behind in the room.
The cost of that defensibility is data. Bottom-up needs a real count of the target segment, which usually means a business data provider, a market map, or your own pipeline, and it needs an honest price and frequency rather than the optimistic version. Skipping the count and asserting a customer number from intuition is the fastest way to turn a bottom-up slide back into a guess with extra steps.
The trade-off: bottom-up is the harder number to fake and the harder number to build, and it rewards exactly the audience most likely to ask you to defend it, which is why fundraising and diligence decks lean on it even when it takes longer.
What is triangulation, and why does it beat a single estimate?
Triangulation means running both methods and comparing the results rather than presenting whichever number is more flattering. If a top-down estimate and a bottom-up estimate land within a reasonable range of each other, confidence in both goes up. If they diverge by an order of magnitude, you have found a flawed assumption before anyone in the room did, and there is almost always one to find: a filter applied to the wrong base, a price assumed at list rather than at what customers actually pay, a published figure that already excludes the segment you serve.
Getting this wrong is not an academic concern. CB Insights's 2026 analysis of 431 venture-backed companies that shut down since 2023 found poor product-market fit cited as a contributing cause in 43 percent of failures, and that two-thirds of those product-market-fit failures were early-stage companies that never found a market at all. A single-method market sizing slide cannot catch that failure mode before it happens; a triangulated one, built from two independent directions that are supposed to agree, at least gives the room a chance to ask the right question before the money is spent.
The trade-off: triangulation costs you the speed of picking one method and moving on, and it can surface an uncomfortable result, two numbers that refuse to agree, that a single-method slide would have hidden. That discomfort is the point.
How do you size a market that does not exist yet?
Top-down and bottom-up both assume the category already exists in roughly its current shape. That assumption breaks the moment a product could expand the market rather than just take a slice of it, and the clearest public example of the argument is a fight between a finance professor and an early investor over how big Uber's opportunity actually was.
In June 2014, NYU Stern finance professor Aswath Damodaran published a top-down valuation of Uber built from the existing taxi and limousine market: roughly $100 billion globally, built up from Japan at $20 to 25 billion, the UK at $14 billion, the US at $11 billion, and a judgment call adding another $50 billion for the rest of the world. Assuming Uber could win a 10 percent share of that market and keep 20 percent of gross receipts, his base case valued the company at $5.9 billion, well below the $17 billion implied by its funding round at the time.
Bill Gurley, a general partner at Benchmark and an early Uber investor, published a rebuttal the following month. His central objection: "In choosing to use the historical size of the taxi and limousine market, Damodaran is making an implicit assumption that the future will look quite like the past." In other words, the estimate assumed Uber would have zero effect on how much car-for-hire transportation people actually bought. Gurley argued the existing $100 billion market could itself grow three to six times over on faster pickups, broader coverage and lower prices alone, and layered on a second, larger opportunity: capturing even 2.5 to 12.5 percent of the roughly $6 trillion people spend globally on owning and maintaining their own cars, if ride-hailing became convenient enough to replace some of that ownership. Combined, his range ran from $450 billion to $1.3 trillion, an opportunity Damodaran's model could not see because it started from a market that assumed cars-for-hire would keep behaving the way taxis always had.
Neither side was dishonest. They disagreed about whether the historical market boundary still applied, which is exactly the judgment a genuinely new category forces on you. If your product only takes share from an existing behavior, size it top-down or bottom-up. If it could change how often people buy at all, say so on the slide, name the comparison you are drawing, and expect the number to be argued with.
The trade-off: this method produces the biggest numbers and the least agreement. It is right exactly as often as the analogy holds, and wrong in a way that is obvious in hindsight and genuinely uncertain in the room.
How do you actually build the TAM, SAM, SOM slide?
PowerPoint has no market sizing chart type, so you assemble it from shapes. The build is mechanical once the numbers are triangulated; the risk is losing the narrowing logic while making it look designed.
Define the market boundary in one sentence: which product, which buyers, which geography, which year. A fuzzy boundary makes every number below it meaningless, because SAM and SOM are just that boundary applied twice more.
Size TAM, SAM and SOM two ways, top-down and bottom-up, and record both sets of inputs so the cross-check is visible on request, not just the final figure.
Choose a layout that fits the story, not the one that makes the opportunity look biggest. Rings, a funnel, or bars each carry the narrowing differently, and the choice belongs to what you are trying to prove.
Keep every figure in the same unit and the same year. Converting between millions and billions, or between last year's TAM and this year's SOM, is the fastest way to make a triangulated estimate look sloppy.
Show the assumption beside each figure: the source for TAM, the filter for SAM, the share and period for SOM. This is the line between a number people trust and one they discount.
Write the action title last. Replace the topic label with a full-sentence so-what, leading with the obtainable share and the reason it is worth pursuing.
Concentric circles
Best for
Showing that SAM sits inside TAM and SOM sits inside SAM.
Watch out for
Scale each radius to the square root of its value so circle area, not radius, represents the number.
Narrowing funnel
Best for
Explaining the segment, geography, product and capacity filters that narrow TAM down to SOM.
Watch out for
Label every filter, or the narrowing looks decorative rather than analytical.
Horizontal bars
Best for
Comparing exact figures in a dense exhibit or beside other charts.
Watch out for
Keep one scale, one unit and one year, and size bar lengths to the underlying values.
Illustrative example, not a sourced estimate
Say you sell scheduling software to mid-size home-services companies, plumbing, HVAC and electrical, in the United States. Bottom-up: roughly 90,000 US companies in that size band, an average contract value of $6,000 a year, and a steady renewal rate gives a SAM near $540 million; if 8 percent of that segment would plausibly switch vendors within two years, SOM lands near $43 million. Top-down: say the published field-service-management software category is sized at $3.8 billion; your segment and geography filters take that down to roughly $610 million, close enough to the bottom-up SAM that both numbers earn a place on the slide. Every figure here is invented to show the arithmetic. Your own TAM, SAM and SOM must come from a named source or your own data, never from a plausible-sounding round number.
Gotcha
Three tells that a market sizing slide was rushed: mixing units or years between TAM, SAM and SOM; leading with TAM as the headline number instead of the SOM you can actually win; and a single point estimate with no source, which reads as a guess no matter how confidently it is presented.
Can Oria build the market sizing slide for you?
Yes, and it is a narrow, specific thing it is good at. Oria is built for corporate documents and consulting-style presentations: board packs, steering committee decks, diligence exhibits, dense frameworks held to an enforced corporate template. It reaches you two ways, a PowerPoint add-in running in the task pane on Windows, macOS and PowerPoint for the web, and a connector for Claude and ChatGPT over MCP, so the slide can be built from the chat you are already in without opening PowerPoint first. Either way the output is native PowerPoint objects on your own master, so every ring, label and assumption note stays editable after it lands: swap a figure, rename a segment, or recolor the emphasis without rebuilding the layout.
One-line market sizing prompt for Oria
Size and triangulate the market in Claude first, then hand the clean TAM, SAM and SOM figures to Oria to render.
Size top-down and bottom-up, then triangulate, in Claude
The same route covers the neighboring exhibits: see the Claude skill for slide design for the storyline side of the workflow, and for a numeric bridge instead of a funnel, see how to build a waterfall chart slide.
The trade-off: Oria is built for the corporate environment and loses on highly visual work. Founder fundraising decks, launch and campaign decks, student presentations and marketing one-pagers want visual impact more than defensible content, and a design-led tool such as Gamma, Canva, Pitch or Beautiful.ai will beat it there and look better doing it. It also will not tell you whether your TAM is honest; that judgment call is still yours.
Which tool should you use to size and show the market?
It depends what the number has to survive after you draw it.
| If you need | Use | Why |
|---|---|---|
| To run the arithmetic and argue about assumptions before anything is designed | A spreadsheet, or Claude | Sizing is a text and math problem first. Settle the inputs, then render once. |
| A published category figure you do not have the time or data to build yourself | A paid industry research subscription | Buying a credible top-down source, from a provider such as Statista or IBISWorld, is often cheaper than defending a number nobody outside your team can trace. |
| A pitch-deck version that has to look designed for a fundraising audience | Canva or Pitch | Where the job is visual impact rather than a defensible build, a design-led tool will beat a corporate template and look better doing it. |
| The exhibit inside your firm's template, as native editable shapes on a board pack | Oria | Built for board packs and steering committee decks: the rings, labels and assumption notes come out as real PowerPoint objects on your master. |
| A living estimate that updates as your pipeline or pricing changes | A spreadsheet model tied to your CRM, not a static slide | A market sizing slide is a snapshot. Keep the model where it can actually be re-cut. |
| A defensible bottom-up count of a named customer segment | A business data provider or your own CRM export | The customer count is the one input in a bottom-up estimate you cannot estimate your way around. |
Common mistakes to avoid
For the wider habits that make AI-built slides look board-ready rather than generic, the consultant's guide to Claude covers the end-to-end method from analysis to deck. For the framework slide that usually sits next to this one in a business case, see the SWOT analysis slide.
Frequently asked questions
What is market sizing?
Market sizing is the process of estimating how big a business opportunity is, usually broken into TAM, SAM and SOM: the total addressable market, the slice of it a specific company can serve, and the share it can realistically win in a given period. It is built two ways, top-down from a published figure or bottom-up from unit economics, and the strongest estimates run both and compare them rather than trusting either alone.
What is the difference between TAM, SAM, and SOM?
TAM is the total addressable market: the full revenue available if every potential buyer in the category purchased. SAM is the serviceable addressable market: the slice of TAM reachable given your product, segment and geography. SOM is the serviceable obtainable market: the share of SAM realistically winnable in a defined period given competition and capacity. Each figure nests inside the one above it.
Should I size the market top-down or bottom-up?
Use both and compare them. Top-down starts from a published figure and narrows it with filters; it is fast but only as credible as the source. Bottom-up builds from your own customer count, price and frequency; it is slower but far harder to challenge because every input is yours to defend. When the two land close together, trust rises; when they diverge, you have found an assumption worth fixing.
What is triangulation in market sizing?
Triangulation means sizing the same market two independent ways and comparing the results rather than presenting a single number. If a top-down estimate and a bottom-up estimate land within a reasonable range of each other, confidence in both rises. If they diverge sharply, the gap points at a flawed assumption worth investigating before an investor or a steering committee finds it first.
How do you calculate market size?
Pick a method and a boundary first: which product, which buyers, which geography, which year. For top-down, take a published estimate of the category and multiply by the share your segment and geography represent. For bottom-up, multiply the number of target customers by the average price and the purchase frequency. Run both, and treat agreement between them as your real confidence signal, not either number alone.
What is a good TAM for a startup?
There is no single number that makes a TAM good; a large TAM with no credible narrowing to SAM and SOM is a bigger red flag than a modest one built honestly. A public 2014 dispute over Uber's addressable market saw two credible analysts land more than ten times apart using the same public facts, so the size of the TAM matters less than whether the filters getting from TAM to SOM are visible and defensible.
What is the fastest way to build a market sizing slide?
Describe the structure in one line and let Oria render it. Give the TAM, SAM and SOM figures and the assumption behind each, and Oria builds a fully editable native PowerPoint slide in your template, including the nested rings or funnel layout and the supporting notes, without manual shape alignment or connector work.