← Guides

How to Research Your Market in 24 Hours

A working method for going from "I know nothing about this market" to a structured, cited view of its size, competitors, buyers, and risks — in one day. Hour-by-hour, with the sources to use and the traps to avoid.

Published 2 August 2026 · 12 min read · GetResearchReports

Why 24 hours is the right constraint

Most market research fails in one of two directions. It either never happens — the decision gets made on instinct because a proper study felt like a six-week project — or it sprawls, consuming a month and arriving after the window it was meant to inform has closed.

A one-day constraint fixes both. It is long enough to reach the point of diminishing returns on public information, and short enough that you will actually do it. The uncomfortable truth about most commercial questions is that the eighth hour of research changes your decision far more than the eightieth. After roughly a day, you stop learning new things about a market and start collecting confirmations of what you already found.

What a day genuinely buys you: an order-of-magnitude estimate of market size, a map of who is already serving the market and how they price, a working theory of who buys and why, and — most valuable — a clear list of the things you still do not know. What it does not buy you: statistically valid survey data, proprietary transaction data, or a figure precise enough to defend line-by-line to an investment committee. Know which of those you need before you start.

Before you start: write the decision down

The single highest-leverage thing you can do takes five minutes. Write, in one sentence, the decision this research is meant to inform, and next to it, what finding would change your answer.

"Should we launch a cold-brew line?" is not yet a research question. "Is the U.S. ready-to-drink cold-brew market growing fast enough, and fragmented enough, that a new entrant with a $2M budget can reach $5M in revenue within three years?" is. The second version tells you exactly what to measure: growth rate, concentration, and the revenue that recent entrants have achieved.

Then write your kill criteria — the findings that would make you stop. For example: "If the top three brands hold more than 70% share and all are owned by national beverage distributors, we stop." Deciding this before you look protects you from the strongest force in fast research, which is your own desire for the answer to be yes.

Hours 0–2: Frame the market

You cannot size a market you cannot name. Spend the first two hours establishing vocabulary and boundaries — this is the step people skip, and it is why their numbers come out wrong by a factor of ten.

Three things to nail down:

• The industry's own terminology. Every market has words insiders use and outsiders do not. "Cold brew" to a consumer is "RTD coffee" in a beverage industry report and sits under NAICS code 311920 in government data. Find these synonyms early; they are the keys to every subsequent search. Trade publications and the "Competition" section of a public competitor's annual report are the fastest place to harvest them.
• The boundary. Write down explicitly what is inside and outside your market. Does "project management software" include general-purpose task apps? Spreadsheets? A market-size figure is meaningless until the boundary is stated, and published figures vary wildly precisely because their boundaries differ.
• The value chain. Sketch who makes, distributes, and sells the thing, and where money changes hands. Ten minutes on this reveals whether your real competition is other products or the distributors who control shelf space.

Output of this block: a page of definitions, ten to fifteen search terms including the industry jargon, and a rough value-chain sketch.

Hours 2–6: Size it, twice

Market sizing has one rule that matters more than all the others: do it two independent ways and compare. A single number tells you nothing about its own reliability. Two numbers from different methods tell you whether to trust either.

The top-down pass

Start from a published total and filter down. Search for the category plus "market size", "market report", or "industry outlook". Research firms publish headline figures in press releases and free report summaries even when the full report costs thousands — the headline number and the growth rate are usually all you need.

When you find a figure, do not write it down yet. Answer three questions first: Who originally produced it? What did they include in the definition? What year is it for, and is it actual or forecast? A figure whose provenance you cannot establish is not evidence — it is a rumour with a decimal point. In practice, a surprising share of widely-cited market numbers trace back to a single vendor-sponsored study.

Then apply your filters explicitly: geography, segment, channel. Show your arithmetic so a reader can disagree with a specific assumption rather than the whole number.

The bottom-up pass

Build the same number from units. Number of potential buyers × how often they buy × what they pay. Each input should come from somewhere you can point to: census or industry-association counts for the population, public pricing pages for price, competitor disclosures or reasonable analogy for frequency.

Bottom-up is slower and it is where the real understanding lives. If the two numbers land within roughly 2–3× of each other, you have a usable range — quote the range, not a point estimate. If they differ by 10× or more, something in your framing is wrong, and finding out what is more valuable than either number was.

Hours 6–10: Map the competitive set

Now find out who is already doing this. The goal is not a list of logos; it is a structured view of how the market is served and where it is not.

Build a simple table with one row per competitor and these columns:

• Who they serve — the segment they actually target, in their own words from their homepage.
• Positioning — the claim they lead with. Cheapest? Fastest? Most trusted? Most specialised?
• Pricing — public prices where available; "contact sales" is itself a data point about deal size.
• Scale signals — headcount from professional networks, funding history, review counts, app-store ratings volume, job postings.
• Weak spots — the complaint that recurs across their negative reviews.

Aim for eight to fifteen competitors: the obvious leaders, two or three recent entrants, and — importantly — the non-obvious substitutes. In most markets the largest "competitor" is the incumbent behaviour: the spreadsheet, the agency, the in-house hire, or doing nothing at all. If you cannot articulate why someone would switch away from doing nothing, the market size number does not matter.

Two high-yield shortcuts. First, read one- and two-star reviews of the market leaders on software review sites, app stores, or retail listings; recurring complaints are the clearest map of unmet need you will find in a day. Second, read the risk-factors and competition sections of any public competitor's annual filing — companies are legally obliged to describe their market honestly there, which makes it the most candid industry analysis available for free.

Hours 10–14: Understand the buyer

Market size tells you whether the prize is big. Buyer understanding tells you whether you can win it. In a 24-hour sprint you will not run a survey, so you mine what buyers have already said in public.

Four sources, in rough order of value per hour:

• Community discussion. Industry forums, professional subreddits, and trade-association discussion boards. Search for the problem, not the product — people describe their pain in their own words before they know a solution category exists.
• Review text. Not the star ratings, the prose. Look specifically for what triggered the purchase ("we switched after…") — that is your buying trigger, and it is worth more than any demographic profile.
• Search behaviour. Keyword tools and the autocomplete on any major search engine reveal the questions buyers ask and their relative volume. Rising queries are a leading indicator that published market data will not show for a year.
• Three real conversations. If any part of your day can hold them, three fifteen-minute calls with actual buyers will outweigh the other twenty-one hours. Ask what they do today, what it costs them, and what they tried before. Never pitch, and never ask whether they would buy — people are unreliable about hypothetical purchases and completely reliable about describing what they currently do.

What you want out of this block: the buying trigger, the current alternative, the approximate budget, and who has to say yes. That is a usable buyer picture.

Hours 14–18: Trends, drivers, and timing

Markets are not static, and the interesting question is rarely "how big is it?" but "why is it changing, and is now the moment?"

Look for four kinds of movement, and for each ask whether it favours entrants or incumbents:

• Regulatory. New rules create markets and destroy them. Check the relevant regulator's recent announcements and consultations.
• Technological. Has something recently become 10× cheaper or easier? That is usually what opened the window you are standing in.
• Behavioural. Durable shifts in how people work, buy, or live — distinguishable from fads by whether they persisted through at least one hype cycle.
• Capital. Where money has flowed in the last 18 months, and where it has stopped. Heavy recent funding means the window is closing; a long drought can mean the opposite of what people assume.

Be disciplined about the difference between a trend and a story. A trend has at least three years of directional data behind it. A story has a compelling narrative and one vendor's survey. Both appear in headlines; only one should appear in your analysis.

Hours 18–21: Try to kill the idea

This is the block that separates research from advocacy, and it is the one people skip. Spend three hours deliberately searching for evidence that you are wrong.

Concretely:

• Search for failure. Company name plus "shut down", "post-mortem", "wind down", or "acquired for parts". Founders write remarkably honest post-mortems, and a market with a graveyard is telling you something specific about why it is hard.
• Invert every assumption. Take your three load-bearing assumptions and write the strongest case against each. If your case rests on "buyers will switch for a 30% price saving", spend twenty minutes finding out what switching actually costs them.
• Check your kill criteria. The ones you wrote in hour zero. Honestly.
• Find the structural reason. If a market looks obviously attractive and underserved, assume there is a reason it is underserved — regulatory burden, distribution lock-in, brutal unit economics, a long sales cycle — and go find it. Sometimes there genuinely isn't one. But the default assumption should be that someone smart already looked.

Hours 21–24: Write the one-pager

Research that stays in your notes has not been done. The final three hours turn findings into something a colleague can act on and challenge.

The structure that works, in this order:

• The answer, in three sentences. Lead with your conclusion and your confidence in it. If a reader stops here, they should have what they need.
• Market size. The range, both methods, and the assumptions each rests on.
• Competitive picture. The table, plus one sentence on where the gap sits.
• The buyer. Trigger, current alternative, budget, decision-maker.
• What would have to be true. The assumptions your conclusion depends on, stated so they can be tested later.
• What we still don't know. Explicit, and ranked by how much it would change the decision. This section is what makes fast research trustworthy rather than reckless.
• Sources. Every one, linked, with a note on which you would not bet on.

Two habits make the difference. Attach a confidence level to every material claim — high, medium, or low — so readers know where to push. And keep the unknowns section honest: a report that admits its gaps gets used, while one that projects false certainty gets quietly discounted the first time a reader catches an overreach.

Where to actually look

The sources below are, in our experience running this process repeatedly, the highest yield per hour. Note that free access to any specific database changes over time — verify before you rely on one.

• Government statistics. National statistical offices, census bureaus, and trade departments publish industry revenue, establishment counts, and employment by sector. Free, methodologically transparent, and typically one to two years behind. The best possible anchor for a bottom-up estimate.
• Public company filings. Annual and quarterly reports contain segment revenue, stated market sizes, named competitors, and — in the risk factors — an unusually candid account of what makes the market difficult.
• Trade associations. Nearly every industry has one, and they publish member surveys and sector statistics that no general-purpose source carries.
• Trade press. The niche publication that industry insiders actually read is worth more than any general business title for vocabulary, deal flow, and who matters.
• Research-firm press releases. Useful for headline size and CAGR. Always check who commissioned the study.
• Review platforms and app stores. The best free source of buyer language, switching triggers, and unmet need.
• Job postings. An underrated signal. What a competitor is hiring for tells you what they are building, which markets they are entering, and roughly how fast they are growing.
• Academic literature. Slower, but for questions about adoption dynamics or price elasticity, a single good paper can replace a day of guessing.

One discipline that pays for itself: keep a running source log as you go — link, what you took from it, who originally produced the figure, and your confidence in it. Writing it at the end means reconstructing provenance from browser history, which is where most citation errors are born.

Seven mistakes that ruin fast research

• Citing a number whose origin you never checked. Market-size figures circulate for years detached from their methodology. Trace it to the original measurement or label it as unverified.
• Sizing top-down only. Top-down numbers are easy, fast, and routinely wrong by an order of magnitude. The bottom-up cross-check is the whole point.
• Defining the market to fit the answer. If you widen the boundary until the number looks good, you have written marketing copy, not research.
• Ignoring the do-nothing option. Your biggest competitor is usually inertia, and it does not appear on any competitor list.
• Mistaking a vendor's content marketing for industry data. A survey commissioned by a company that sells the solution is a data point about that company's positioning.
• Confirmation bias. If you did not spend real time trying to disprove your thesis, you did not test it. Hours 18–21 exist for this reason.
• Precision theatre. A figure to two decimal places built on three guesses is less honest than a stated range. False precision is the fastest way to lose a reader who knows the market.

The 24-hour checklist

Frequently asked questions

More from our guides: browse all research guides →