Market research

How to Do Market Research (Without a $5,000 Report)

A concrete method, not a definition of what market research is.

Most guides to this topic spend the first third explaining what market research is and why it matters. You already know why it matters — that's why you're here. This is the how, for the version you can do yourself before paying anyone for anything.

What you're trying to produce

Not a document. An honest answer to three questions: does anyone want this, will they pay for it, and can you actually get in. Everything below exists to answer one of those three.

Step 1 — Check search demand (30–60 minutes)

Pull search volume for your core terms and their obvious variants. Look for two things: the raw number, and the trend over the past 6–12 months. A small but growing number often beats a large but flat or declining one — you'd rather catch a wave early than paddle into one that's already breaking.

Free options exist (Google Trends gives you relative direction, not absolute volume) and paid options exist (absolute monthly volume, competition, and cost-per-click, usually cents per keyword through a data provider). Start free; upgrade to paid data once you know which terms are worth measuring precisely.

Step 2 — Read where people already complain (2–3 hours)

Go to wherever your future customers already discuss this problem — relevant subreddits, forums, review sections of existing alternatives — and read, don't skim. You're looking for specific, repeated language: the same complaint showing up independently across multiple people, in their own words.

One complaint is an anecdote. The same complaint from five unconnected people is a pattern, and patterns are what you're actually looking for.

Step 3 — Map the competitive field (1–2 hours)

Search your core terms and note who ranks, whether the same names repeat across variants (concentrated market) or rotate (fragmented market), and whether ads are running — sustained ad spend on a term is a real signal, since businesses don't keep paying for clicks that don't convert. Full method in our competitor analysis guide.

Step 4 — Talk to 10–15 real people (1–2 weeks, mostly waiting on replies)

Find people who actually match your target customer — not friends being polite — and ask about their current behavior and workarounds, not whether they'd hypothetically buy your thing. “How do you handle this today?” gets better data than “would you use an app for this?” every time.

Step 5 — Write down what you actually found

One page. What the search data shows, what the complaints say verbatim, who holds the field today, and what the conversations revealed. Force yourself to write the honest verdict — including “this doesn't look promising” as a legitimate outcome. A negative finding after a week is a much better outcome than a negative finding after eight months of building.

When this method isn't enough

If the decision at stake is large — a funding round, a major pivot, quitting a stable income — the stakes usually justify paying for more rigor than a solo afternoon-and-a-week pass can deliver. This method is built for the much more common situation: deciding whether an idea is worth a few months of your own time before you've committed anything else.