Market opportunity analysis: find the one guess that decides go or no-go

A market opportunity analysis is a short, evidence-backed answer to one question: should we pursue this market, yes or no? Run it in five steps: name the customer and the problem, size the market from the bottom up, check who you would be competing with, ask why the timing is right, and then test your single shakiest assumption by talking to buyers and people who know the industry before you write "go."

That last step is the easiest one to skip. A spreadsheet can make any market look large; only a conversation tells you whether the one number holding it up is true.

1. Start from the decision, because an analysis without one never ends

Write the decision at the top of the page before you collect anything: "Do we spend the next two quarters building X for Y?" A market opportunity analysis that is not attached to a decision tends to grow into a 40-slide deck that everyone admires and nobody acts on.

Marty Cagan's opportunity assessment at SVPG is a good skeleton. He lists ten questions, covering the problem, the target market, the size of the opportunity, the alternatives, your differentiator, why now, go-to-market, how you will measure success and make money, the factors critical to success, and finally a go or no-go recommendation. His point is that the exercise either saves the company from a poor opportunity or makes clear what a good one will require.

George Day's Harvard Business Review article compresses the same idea into its title: is it real, can we win, is it worth doing? Keep those three questions in your head while you fill in the details.

Short version: one decision, one owner, one date by which you will say go or no-go.

2. Define the customer as someone you could phone this week

"Practice managers at independent veterinary clinics who reorder supplies from a spreadsheet" is a customer definition you can use, because you can picture finding one and asking about last Tuesday. "Small businesses" gives you nobody to call.

Define three things in a sentence each: who has the problem, what it costs them today (in time, money or risk), and who signs the check if they are different people. If you cannot name the buyer, read how the B2B buying process splits these roles before you go further.

The SBA's market research guide lists demand, market size, economic indicators, location, saturation and pricing as the factors to weigh. Demand comes first on that list for a reason: if nobody wants it, the other five are trivia.

3. Size the market bottom up, from a count you can source

Bottom-up sizing means counting the buyers, multiplying by what each would pay, and then cutting the result down to the share you could realistically win. It beats taking 1% of a large industry figure because every input is a number you can check or test. The TAM, SAM and SOM guide and market size research cover the mechanics in depth; here is the short version as a worked example.

Here is a made-up example. Dana runs a small software team and wants to build inventory reordering software for veterinary clinics. Her hunch comes from a friend, a practice manager in Ohio whose reorders live in a spreadsheet tab called REORDER_final2.

The starting count is real. The Bureau of Labor Statistics Quarterly Census of Employment and Wages reports a 2024 annual average of 35,677 private establishments in veterinary services (NAICS code 541940), with 467,667 employees. That is about 13 employees per establishment. "Private" is ownership code 5 in the QCEW ownership titles, and the BLS CSV slice documentation defines the establishment count as the annual average of the quarterly counts.

Everything after that is Dana's assumption, and she marks each one as a guess:

Check the arithmetic: 35,677 x 0.6 = 21,406 (rounded down), 21,406 x 0.5 = 10,703, 10,703 x $1,800 = $19,265,400, and 5% of 10,703 is 535 clinics, which at $1,800 each is $963,000 a year.

Dana's go bar is $1 million in yearly revenue within three years. Her estimate lands just under it, which is a common and useful result: the answer depends on one or two guesses.

Short version: the sourced count is the floor of your analysis. The guesses stacked on top of it are where the real work is.

4. Find the assumption that flips the answer, then stop polishing the rest

Change each guess and see which one moves the decision. In Dana's model, two do.

If clinics would pay $250 a month ($3,000 a year) instead of $150, the same 535 clinics bring in $1,605,000 a year, comfortably over the bar. If only 25% of independent clinics track inventory by hand, the pool drops to about 5,350 clinics, 5% of that is about 268, and revenue falls to $482,400 a year, well under the bar.

On paper the independent share moves the result just as much, but Dana can check it against chain ownership lists, and chains might buy too. The hand-tracking share is both shaky and hard to look up, so it is the biggest assumption, with price close behind. Everything else in the model can stay rough.

This is the step that turns a market opportunity analysis into a decision. The question has shrunk to one sentence: do half of independent clinics really reorder by hand, and what would they pay to stop?

5. Check competition by asking what buyers use today

List the alternatives the buyer already has, including the boring ones. On Dana's first list, each still to be checked, are practice management software with an inventory module, ordering tools from supply distributors, and the spreadsheet itself. The spreadsheet is free and never asks for a renewal, which makes it a stronger rival than it looks. Competitor analysis covers this in more depth.

The SBA guide suggests identifying competitors by product line and by segment, assessing their strengths and weaknesses, judging your window of opportunity, spotting barriers to entry, and naming indirect competitors. For a small team, the barrier question matters most: if a distributor gives inventory tools away to keep clinics ordering from it, your price assumption is in trouble.

If you are entering a market you do not know at all, market entry research is the better guide; this step assumes you mainly need to know who you would displace.

6. Answer "why now" with a change you can point to

Cagan's sixth question is "why now," which he labels the market window. A good answer names something that changed: a regulation, a price drop in a key input, a shift in who owns the businesses, a new channel. "The market is growing" names no change at all; most markets are growing in somebody's slide deck.

If you cannot name the change, the opportunity may be real but already taken, or real but not yet ready. Either way, write that down and lower your score on timing. Market trend research has five checks for deciding whether a trend you have read about is real.

7. Test the biggest assumption with buyers and industry insiders

Talk to two kinds of people. Buyers tell you what they do today and what it costs them. People who know the industry from the inside (a former distributor sales rep, a finance lead at a clinic group, a consultant who sets up practice software) tell you how common that behavior is across hundreds of clinics, which no single buyer can know.

For Dana, a reasonable test is 12 practice managers at independent clinics and 3 industry insiders. Ask each practice manager to walk through the last time they reordered supplies, step by step, and ask what they would cut from their budget to pay for a fix. Ask the insiders what share of independent clinics they think still reorders by hand, and why.

Write the kill rule before the first call: if fewer than 4 of the 12 practice managers reorder by hand and name a real cost, the answer is no-go. A kill rule written afterwards has a way of moving to wherever the results landed. (It is remarkable how often the bar turns out to be exactly one interview below what you got.)

Use customer interview questions that start from the last real event, and keep compliments out of the evidence column. Market validation covers the next step if the answer is go: asking for a paid pilot.

This analysis does not fit every situation. If you already sell to these buyers and have usage data, start from your own numbers instead. If the market is brand new and nobody has the problem yet, sizing from a count is premature, and you should run problem interviews first.

8. Put it on one page and score it

Here is a one-page scorecard you can copy. Score each row 1 (weak), 2 (unclear) or 3 (strong), and write the evidence beside the score. A row with no evidence scores 1, however confident you feel.

A total of 25 or more out of 30 with a 3 on the assumption test is a reasonable go. Anything with a 1 on the assumption test is a "test further," whatever the total says, because the rest of the page rests on it.

For Dana's veterinary example, the insiders worth reaching include CFOs in veterinary services and CIOs in veterinary services; for help structuring the assessment itself, a product management consultant can review the page. If finding those people is the slow part, Instant Expert lets you invite specific practitioners and buyers to a short call or a written or voice answer, and you pay only if someone books or answers.

Start today by writing the decision and the one sourced count at the top of the page. The rest of the analysis is an argument about the guesses in between.