Research Methodology
Every CDI estimate is built the same way: establish context from published sources, validate it against people who operate in the market, then size it twice by two independent routes and reconcile the difference.
Research approach
A market size is a claim about the world, and a claim is only as good as the route to it. Every study is built upward from units and prices — what is actually produced, sold or performed, at what it actually changes hands for — rather than from a headline figure divided downwards. Disclosed company revenue is then used to check that build, not to produce it.
The four stages
The same sequence runs behind every published study, whatever the industry. The order matters as much as the steps: the segment axes are fixed before any number is collected, so the model is never reshaped to fit whatever data happens to turn up.
What the build rests on, and what checks it
The two are not interchangeable. The left column produces the number; the right column tests it. When the check disagrees with the build, the answer is to find which bottom-up assumption is wrong — a unit count, a price, a take-up rate — not to split the difference between them.
- Volume actually transacted — units produced, installed, dispensed or procedures performed, counted at the level each is genuinely recorded
- Realised pricing by tier and channel, rather than one blended average applied across the whole market
- Take-up and frequency: how much of the addressable base buys, and how often it repeats
- Disclosed revenue of the companies serving the market, where filings separate it far enough to be usable
- Buyer-side spending totals — capital budgets, procurement lines, or the output of the end market the product is bought against
- Trade and customs flows, where the product crosses borders in a separately recorded form
Data sources
Published data establishes what happened. Only the people transacting in a market can say why, and what is about to change — so the two are collected separately and weighted differently.
- Commercial and product leadership at the companies that supply the market
- Procurement and specification leads at the organisations that buy it
- Distributors, integrators and channel partners, where the market is served indirectly
- Regulatory and standards specialists, where approval governs what can be sold at all
- Company filings, annual reports and investor disclosure
- Government statistics, customs records and regulatory registers
- Trade association output and standards-body publications
- Technical and peer-reviewed literature, where the market rests on a clinical or engineering claim
Desk research runs across proprietary research databases including Factiva, OneSource and Hoovers alongside the public sources above. Modelling and statistical validation are run in SAS and SPSS.
Triangulation and validation
No figure enters a report on the strength of one source. Where the two sizing routes disagree the difference is not averaged away — the assumption causing it is isolated, tested against a third independent measure, and either corrected or carried forward as a stated limitation. Historical years are back-tested against the growth actually recorded before any forecast is allowed to run forward from them.
What every report covers
Scope varies by market, but the following are present in every syndicated study we publish:
- Market trend analysis, with the drivers and restraints behind each trend stated separately
- Segmentation by application, technology or product
- Geographic segmentation — North America, Europe, Asia Pacific, Latin America, and the Middle East & Africa, broken down to country level
- Market sizing and forecasting across the study period
- Growth drivers and the opportunities they open
- Industry issues and challenges
- Market share analysis
- Company profiles of the key players covered
Seeing it applied
Each report's own Methodology tab carries a market-specific version of this — the sizing approach actually used for that market and who was interviewed to validate it. Open any report in the catalogue and select the Methodology tab, or request a free sample to see the full working in context.
Questions about how a number was built?
Analysts are available to walk through the assumptions behind any estimate in any report we publish.