A five-day market scan answers one written-down decision using evidence that mostly already exists: competitor pricing, job ads, reviews, forums, and five short buyer conversations, delivered as a single page with a recommendation. It works because it is scoped to the decision, not the market. The hundred-page report answers questions nobody asked, three months after the decision was made. The scan gets used, because it arrives while the question is still open.
Day one: write the decision down
Not the topic. The decision. "Should we launch the premium tier this quarter?" is researchable in a week. "Understand the market" is not researchable at all, which is why projects briefed that way run for months and land with a thud. Write the decision as a question with a yes or no in it, name who makes the call and when, and add one more line: what evidence would change your mind. If nothing would, you don't need research, you need permission, and that's cheaper.
That sentence becomes the filter for everything that follows. Every source you consult for the next four days either moves that decision or gets dropped without guilt. Scope discipline is the entire method; the rest is legwork.
Days two and three: collect what already exists
Public data first, because most decisions don't need a survey; they need someone to read what's already visible. Competitor pricing pages, and how they've changed, tell you where the market believes value sits. Job boards tell you what rivals are building months before any announcement. Reviews tell you what customers praise and what they punish, in their own vocabulary, which is worth stealing for your copy later. Industry forums and communities carry the complaints nobody says to a vendor's face. The same signals are how you read a market when your competitors go quiet.
Work with a simple rule: capture the source and the date for every claim, and separate what you observed from what you're inferring. An observed fact ("three of five competitors raised prices this year") and a constructed guess ("the market is moving upmarket") are different grades of material, and a scan that labels the difference is worth ten that don't.
Research the decision, not the market.
Day four: talk to five people
Five short conversations with people who match your buyer beat five hundred survey responses you'll argue about later. Twenty minutes each, three questions: what do you do today for this, what have you tried that didn't stick, and what would you pay to stop doing it? Past behaviour and money questions produce truth; hypotheticals produce politeness. And resist the urge to pitch. The moment you start selling, they start being nice to you, and nice is the enemy of information. When you do run a formal test instead, know how A/B tests produce false winners before you trust the result.
Finding five people in a day is easier than it sounds if the groundwork is real: existing customers who fit the profile, a lost prospect who'll take a call, a LinkedIn message that asks for advice rather than a meeting. You're not building a statistically valid sample. You're checking whether real humans recognise the problem you're about to bet on.
Day five: one page, one recommendation
The output is a single page with five parts: the decision as written on Monday, the evidence for, the evidence against, the recommendation, and what would change your mind. If it doesn't fit on a page, the thinking isn't finished; length is where unfinished analysis hides. Ship it Friday, decide Monday, and file the page where the next scan can find it, because the evidence-against section is tomorrow's early-warning list.
What a scan can't do
Fair limits, stated plainly. A five-day scan won't size a market to a defensible number, won't validate demand for a genuinely new category where no behaviour exists to observe, and won't replace ongoing tracking of a market that moves monthly. Those are different projects with different price tags, and pretending a scan covers them is how quick research gets a bad name. What it does, it does better than the big report: it answers the question in front of you, this week, while the answer can still change what you do.
