Founders ask us, more than any other question, whether they should pay for market research or do it themselves. The honest answer is: it depends on three things, and you can usually figure out which side of the line you’re on in twenty minutes.

This post walks the decision and gives you the DIY toolkit for when “do it yourself” is the right call.

The three things that determine the answer

  1. Is the data public? If your market is well-covered by IBISWorld, Statista, or sector reports you can buy for $500 to $2,500, the heavy lifting is finding and synthesizing the public sources. That’s something a careful founder can do.

  2. Are you building the market or pitching the market? If the report is for internal decision-making, a thorough DIY pass usually gets you to a good-enough answer. If the report is the document that goes into a data room and gets read by a partner before the Series A meeting, “good enough” isn’t.

  3. What’s your time worth? Producing a thorough market research deliverable takes 30 to 60 hours of careful work. If your hourly cost is $50, that’s $1,500 to $3,000 of time. If it’s $200, that’s $6,000 to $12,000. The break-even against a paid SKU is mechanical.

The DIY toolkit, for when DIY is the call

If you’ve decided to do it yourself, here’s what works.

Sources that are actually useful

  • US Census Bureau, surprisingly current, free, and the most reliable source for industry size by NAICS code.
  • Bureau of Labor Statistics, employment data by industry, useful for triangulating business count.
  • County Business Patterns, Census product giving you firm counts by employee bucket. Critical for bottom-up TAM.
  • SBA Office of Advocacy, small-business industry data, especially useful for service businesses.
  • State licensing boards, for regulated industries (healthcare, legal, beauty, food), the licensing board is the most accurate count of operators in the state.
  • Trade association membership directories, when membership is opt-in but high-coverage, you get a near-complete list of operators in a category.
  • SEC EDGAR, for any market with public competitors. 10-Ks contain segment revenue, customer counts, and management commentary that’s better than most paid research.
  • App store and Play Store APIs, for consumer markets, download estimates from Sensor Tower, App Annie, or even just public ranking data over time.

Sources to use carefully

  • Statista, IBISWorld, Grand View Research, useful as anchors, dangerous as sole sources. Their numbers are produced for enterprise customers with different needs than yours. Trust the order of magnitude, distrust the precision.
  • Crunchbase, PitchBook, funded-status data is reliable; revenue estimates are not.
  • Glassdoor, Indeed, useful for triangulating company size via employee counts and active job postings.

Sources that don’t work

  • Twitter / X threads from VCs about market size. Hot takes, not research.
  • Industry blogs without methodology disclosure. If they don’t show their math, you can’t use the number.
  • AI-summarized market reports. They hallucinate citations. You’ll spend more time fact-checking than you would have building from primary sources.

The DIY workflow that produces a useful document

A defensible market research workflow, doable in 30 to 60 hours over a week or two:

Day 1: Scope. Write a one-page scope statement. Industry, geography, audience, time horizon. Stop here and check it against what you actually need; the right scope is usually narrower than you think.

Days 2 and 3: Sizing. Bottom-up: count the addressable customers (Census, licensing data, trade directory). Price the average customer. Multiply. Top-down: find the corresponding industry report. Reconcile.

Day 4: Competition. Five to seven named competitors. Positioning, funded status, defensibility, vulnerability. SEC filings for public ones, public reviews for private ones.

Day 5: Customer evidence. Three independent demand signals (search volume, job postings, regulatory filings, trade press). Triangulate. Discount any single signal by default.

Day 6: Synthesis and methodology appendix. Write the conclusions last, against the evidence collected. Then write the methodology section that cites every source line.

Day 7: Internal review. Hand it to someone whose judgment you respect. If they push back on three claims, fix those three before showing it to anyone outside.

When DIY breaks down

A few patterns where paying makes more sense than DIY:

  • Your market isn’t covered by public sources. Emerging categories, regional markets in countries with thin industry data, niche B2B segments. The DIY toolkit fails because the sources don’t exist. You need either primary research (operator interviews) or a researcher whose existing network includes the relevant operators.

  • The audience is institutional. Family offices, SWFs, IC desks, DFI grant reviewers, they expect a methodology section that defends itself. Self-produced reports rarely have the methodology rigor that matches.

  • You don’t trust your own bias. If you’re pitching the market you’re building, your DIY report will unconsciously emphasize the upside. A third party writing the same evidence will produce a more sober document, sometimes a more honest no-go than your own version produces.

  • Time is genuinely the constraint. A founder raising a round in three weeks doesn’t have 60 hours to produce a market section. Outsource it; spend the 60 hours on customer conversations.

How to evaluate a research provider

If you’ve decided to pay, the questions worth asking before committing:

  1. Will the methodology be defensible? Ask for a sample document. Look for: bottom-up sizing with the math shown, source-line appendix, named competitors with funded status, explicit conversion-rate assumptions. Reports without these things aren’t research, they’re presentations.

  2. Who specifically does the work? Named researcher, with specializations, ideally with at-least-some public profile. “Our team of analysts” is the freelancer-aggregator pattern; the actual analyst is probably underpaid and overbooked.

  3. What’s the revision policy? Two revision rounds at minimum. Time-bound (48 hours per round is typical). Unlimited revisions sound great in marketing, in practice they signal scope-creep.

  4. What happens if your industry has thin public data? A good provider tells you within 24 to 48 hours of intake, before charging. A weaker provider takes the money and produces a thin report.

  5. How long? 5 to 10 business days for a serious mid-market report is realistic. Less than that, with claimed rigor, is the rush charge or the under-research.

The Pondera position

Our Market Research Comprehensive at $550 covers the case above: a focused report for a single fundraise decision, with bottom-up TAM, named competitors, and a methodology section that survives a partner-level read. Five business days.

But the honest position is: if you’re early enough that the data is public, the audience is internal, and you have a free week, do it yourself. Use the workflow above. You’ll save the money and learn more about your market in the process than any outsourced report would teach you. We’re not the right answer for that case, and we’ll tell you so.

When DIY breaks down, emerging market, institutional audience, time-constrained fundraise, bias risk, that’s when paying for it earns its keep.