Every number, traced back to its source.
A market-size figure is only as good as the process behind it. Here's exactly how we gather data, size a market from two directions, run it through statistical modeling, and validate it three separate times before it ever reaches a published report.
40+ paid databases
Company financials, trade statistics, and industry journals feed every model.
3-level validation
Normalization, expert panel review, and QC sign-off on every dataset.
Primary interviews
CXOs, technical experts, and channel partners interviewed for every study.
Two-way sizing
Every market is sized bottom-up and top-down, then reconciled.
One pipeline, four stages, every report.
Whether the study is a 300-page global report or a scoped custom engagement, the same four-stage pipeline runs underneath it — only the depth and inputs change.
Data procurement
Paid databases, filings, journals, and primary interviews, gathered in parallel.
Sizing & modeling
Bottom-up and top-down estimates, built in parallel and statistically modeled.
Validation
Normalized, checked against expert panels, then signed off by quality control.
Delivery & support
Published with a named analyst on call for scope questions after you buy.
Six channels feed every dataset, not one.
A single source is a guess; six independent sources cross-checking each other is a market estimate. We run all six in parallel for every study, so gaps in one are caught by another.
Paid databases
Subscription access to financial and industry databases for company revenue, trade flows, and sector benchmarks that aren't publicly indexed.
Primary interviews
Structured, one-on-one conversations with CEOs, CTOs, and channel partners to validate estimates and surface trends before they show up in filings.
Government & multilateral bodies
Published statistics from national statistical offices, trade ministries, and bodies like the WHO and World Bank for regulatory and macro context.
Company filings & investor decks
Annual reports, 10-Ks, and investor presentations for revenue segmentation, capex plans, and forward guidance from the companies that make up a market.
Journals & trade magazines
Technical papers and trade-press coverage that surface emerging technology, regulatory shifts, and competitive moves ahead of formal filings.
Proprietary archive
Our own accumulated data points from prior reports and fbi_research_methodology_active market monitoring, giving every new study a historical baseline to build on.
How we conduct expert interviews.
Primary research is the one channel that can't be bought off a database — it's a structured conversation with someone inside the market. Here's who we talk to, how the conversation is run, and what it's used for.
Who's on the other end of the call
Every study draws from the same eight stakeholder categories, weighted by relevance to the market in question — a medical-device report leans on regulatory experts and clinicians; a components report leans on procurement and channel contacts.
Three interview modes, matched to the audience.
The format changes with who we're reaching and how deep the conversation needs to go — not every stakeholder needs the same instrument.
Telephone interviews
Structured, analyst-led calls for executives and technical experts where follow-up probing and nuance matter more than sample size.
In-person / video interviews
Face-to-face conversations for on-site channel checks, plant visits, and conferences where visual context adds signal.
Structured web surveys
Scaled online questionnaires for reaching a broader panel of buyers and end users quickly, feeding the quantitative side of a study.
We map the stakeholder list before we pick up the phone.
Analysts build a target list against the report's segmentation — by role, geography, and company size — so the panel actually mirrors the market being sized, not just whoever answers.
Every interview runs against the same core question set.
A discussion guide keeps responses comparable across dozens of interviews, while leaving room for the interviewer to probe on anything unexpected the respondent raises.
Responses are logged and tagged against the model's variables.
Qualitative color and quantitative estimates (growth rates, pricing, share) are separated out and coded so they can feed directly into the Stage 02 sizing models.
No single interview moves a number on its own.
A claim from one respondent is checked against others in the same panel and against the desk-research figures before it's allowed to influence the model.
Secondary sources, by category.
The primary interview network is covered above. Below is the secondary-source stack behind every study — government bodies, standards organizations, trade associations, and more — grouped by category and specific to the vertical you select.
We size every market from both directions.
Bottom-up and top-down estimates are built independently, then reconciled. If the two land far apart, that's a signal to dig deeper — not a number we publish and hope for the best.
Bottom-up
We start at the smallest measurable unit — demand per product, per country — and sum upward through segments to the total market.
- Demand estimation per country/product, summed to a total
- Built from paid databases and company financials
- Revalidated through primary interviews
Top-down
We start from an established global or parent-market figure and split it down into segments and countries using penetration and regression models.
- Used for new-product forecasting and penetration analysis
- Multi-variant regression against public & paid databases
- Cross-checked with vendor-side primary research
Combined
Both estimates land on the same desk. Where they agree, confidence is high; where they diverge, we dig into the variable causing the gap before publishing.
- Reconciled via commodity flow and consumption models
- Full supply-side and demand-side coverage
- Default method for our published industry reports
Value chain-based sizing & forecasting.
Alongside bottom-up and top-down, every model is cross-checked from both ends of the value chain — what suppliers can plausibly produce, and what buyers can plausibly absorb.
Supply-side estimates
- Company revenue estimation from annual reports, investor presentations, and paid databases
- Segment revenue determined via variable analysis and penetration modeling
- Competitive benchmarking to identify leaders and their collective share
- Forecasting via commercialization rates, pipeline, and distribution reach
Demand-side estimates
- Identifying parent markets and ancillary/adjacent markets
- Segment penetration analysis to derive addressable revenue and volume
- Heuristic forecasting developed with subject-matter experts
- Forecasting refined through variable-level sensitivity analysis
See how bottom-up and top-down sizing actually move.
A simplified, illustrative version of the two core models from Stage 02 — adjust the inputs to see how each approach arrives at a market size. This sandbox is for explanation only; published figures go through the full validation pipeline above.
= Segment total, summed across geographies to the global figure
= Split down into countries & segments via regression
Eight parameters run through every forecast.
Whichever sizing approach applies, our models are iterative and re-run as new data arrives — refined against the same eight parameters on every project, regardless of sector.
Growth drivers & restraints
Drivers, opportunities, restraints, and emerging trends specific to the market.
Regulatory & political data
Policy, compliance regimes, and government programs affecting adoption.
Application-level dynamics
How demand differs by end-use, application, and buyer type.
Trade statistics
Import/export flows and tariff exposure across key trading regions.
Raw material & supply dynamics
Input availability, pricing, and supplier concentration risk.
Technological development
Innovation cycles and technology shifts that change addressable demand.
Macro & micro economic factors
GDP, industrial output, and sector-specific spending patterns.
Demographics & value chain
Population, income shifts, and where value concentrates across the chain.
Nothing publishes on one person's say-so.
Every dataset runs through three sequential checks. A study only moves to publication once all three return a validated result.
Bottom-up and top-down estimates are reconciled.
The two independent sizing estimates from Stage 02 are aligned into one working figure, and the report's scope and data-representation format are finalized before deeper review begins.
Industry specialists stress-test the numbers.
Reconciled estimates and forecasts are sent to external industry experts and subject-matter specialists, who challenge assumptions and flag anything inconsistent with what they're seeing in the market.
An internal QC manager reviews the finished study.
A quality control manager independent of the research team reviews the complete report for internal consistency, sourcing, and formatting before it's cleared for publication.
Confidence isn't a claim — it's a set of thresholds.
These are the standing thresholds every study is measured against before it publishes. A report doesn't move to delivery until each one is met.
Sources per data point
Minimum independent corroboration before a figure is accepted.
Studies fully validated
Every report clears all three validation checkpoints, no exceptions.
Forecast variance target
The reconciliation band we hold bottom-up and top-down estimates to.
Independent source channels
Every dataset is cross-checked across all six procurement channels, not just one.
The methodology doesn't stop at publication.
Every report ships with a named point of contact, because a market study should be something you can ask questions about, not a PDF you're left to interpret alone.
A dedicated analyst
Each project is assigned a point person who can walk you through scope decisions, data sourcing, or how a specific figure was derived.
Post-sale clarification
Questions about methodology, scope, or how to read a table don't require a new purchase — they're part of what you're already paying for.
Ongoing market monitoring
Published markets stay on our proprietary archive's monitoring list, so the next edition builds on tracked data rather than starting from zero.
Methodology, answered.
Do you use the same methodology for every industry?
What happens when bottom-up and top-down estimates disagree?
Who conducts the primary interviews?
How often are published market figures updated?
Can I request the underlying data behind a specific figure?
Does this methodology apply to custom research too?
Related policies & resources.
How to Order
Every channel and step for turning a report page into a PDF in your inbox.
Read how to order →Custom Research
Commission a bespoke study scoped to your exact market and questions.
View custom research →