Define the market before collecting figures
Specify product category, customer type, geography, currency and reference period. Revenue, users and shipments measure different things. A global software forecast is not automatically a local opportunity estimate. Record each source’s definition so figures from different reports do not appear comparable merely because they share a category name.
Build a small evidence register
For every important number, record the issuing organization, publication date, period measured, units and methodology if available. Distinguish official statistics, first-party company disclosures and third-party estimates. Treat forecasts as scenarios based on assumptions, not as observed outcomes. If the method is unavailable, make that limitation clear.
Brief the Research Agent around the decision
Choose a market-analysis output in Humanizily Research Agent and describe the segment and evidence requirements. Review its outline for demand, buyer needs, competitors, constraints and open questions. Ask it to leave unavailable values unresolved instead of producing plausible-looking figures.
- Use one consistent geography and time basis.
- Label estimates separately from observations.
- Explain differences between source definitions.
- Keep recommendations conditional on the evidence.
Review the report for comparability
Check the largest numbers and the claims that drive the recommendation first. Open sources and confirm whether values were quoted, calculated or inferred. For a calculation, show the inputs and formula. If evidence cannot establish a segment’s size, recommend a specific follow-up investigation rather than presenting uncertainty as a precise total.
Worked example
Starting point
Tell me the market size for productivity apps.
Suggested wording or approach
Assess demand signals for task-management software used by small design agencies in one selected country. Define the category, cite original data where available, distinguish company statements from estimates, and identify evidence gaps.
Example research brief. No market size, growth rate or commercial outcome is asserted.
Common questions
Can I combine market-size estimates from different reports?
Only after checking category, geography, dates, units and methods; otherwise present them separately.
What if a forecast has no public methodology?
Label the limitation and avoid relying on it as the sole basis for a decision.
Should the agent fill in missing figures?
No. Missing data should remain visible, with a suggested way to investigate it.
Start with a focused research brief
Define your question, review the outline and check the evidence before sharing the report.
Open the Research AgentHow the agent works