Primary vs Secondary Research

Secondary research (analyst reports, broker research, regulatory filings, public databases): broad landscape mapping, established categories, quantitative benchmarks at scale, history. Primary research (expert calls, customer interviews, surveys, channel checks): specific operator reality, emerging categories without coverage, ground-truth on assumptions, anything where the published narrative is suspect.

Mostly secondary first. Read the existing analyst coverage, scan public competitor filings, build a baseline of the category's size and structure from desk research. Then run primary to test the 3-5 most fragile assumptions. This sequence works because secondary is fast and cheap; you don't want to burn primary research budget mapping ground that's already mapped.

Mostly primary. Customer reference work, supplier interviews, off-list management refs, channel checks. Secondary is supporting context — you've already decided the deal is worth investigating, so the work is now around testing the management narrative, not mapping the category.

Almost entirely primary. Published pricing data is unreliable in B2B because real prices are negotiated. You need customer and channel interviews to understand actual willingness-to-pay, discount behaviour and price-sensitivity thresholds. Secondary research on pricing typically tells you the list price, which is the price almost nobody pays.

Mixed. Use secondary for ongoing structural monitoring (10-K filings, press releases, analyst notes). Use primary for the questions that secondary can't answer: where is the competitor actually winning deals, what's their channel partner sentiment, what's the practitioner view of their product. Continuous CI programmes typically blend both.

Two cases: (1) speed-to-decision matters more than completeness — you have 48 hours to inform an IC and there's no time to build context properly; (2) the category is so new that no secondary research worth reading exists. In both cases, run primary directly and accept the lower context.

Secondary research costs: subscription to one or two databases (analyst, broker research, regulatory) — typically €10–€50k/year/seat. Primary research costs: €500–€1,500 per call; project packages €10–€40k. For a typical strategic decision, you'd spend €5k on secondary and €15–€30k on primary. Reversing those ratios usually means you're under-investing in primary.