Scope is locked at Product × Technology × Application × Channel × Region (7 regions, 30 countries). Units (value/volume) and currency basis are set up front. Monthly signals inform quarterly baselines. We produce our forecasts at constant FX unless scenarios require alternatives, and we keep our segment trees fixed before data ingestion with all revisions routed through change control.
Unit shipments count new products dispatched by vendors to channels or end users. OEM hand offs are not double counted, and attribution follows the brand on the finished product. Subsidiary inventory is excluded until it enters the channel. Known re export corridors are reassigned to the country of final use when determinable. Manufacturer self-consumption and bona fide donations are included to protect installed base integrity.
At FMI, we capture ASP at distributor level. Freight, insurance, and import or export related taxes embedded in vendor or channel pricing are included while point of sale taxes remain excluded. Outliers are normalized. Prices are weighted across international, domestic, and China based vendors using geography and application sales mixes. For bulk intermediates such as base chemicals, we use bulk distributor prices.
Value equals Units multiplied by ASP. Installed base equals cumulative shipments minus retirements using survival curves by class. Apparent capacity is derived from production, net trade, and a capacity factor that commonly ranges between 80 and 90 percent given operating day and hour assumptions. Historical values are expressed in current period USD and forecasts run at constant FX unless scenario work specifies another basis.
Brand share is calculated as Brand Value divided by Category Value for the defined geography and channel scope. When EPOS data are unavailable brand value is triangulated from filings, tender disclosures, distributor catalogs, and public multi signal proxies. Corridor tests set bounds using minimum share implied by visible capacities or shipments and maximum share implied by price and volume feasibility.
Trade flows are compiled on HS codes with mirror statistics to detect gaps. Adjustments capture re exports through known hubs, FOB and CIF valuation differences, and reclassification events. Apparent consumption closes through Production plus Imports minus Exports with stock change added or removed when observable. Outliers are reviewed using partner statistics, company disclosures, or customs notes.
Revenue and volume are allocated across direct, distributor or wholesale, retail or aftermarket, e commerce or marketplace, system integrator or EPC, and tender or government channels according to the category structure. Margin stacks are modeled for each channel. Leakage across grey and parallel routes is bounded through price corridor checks and availability anomalies. Attach rate models are applied for services and consumables where relevant.
Supply chains are mapped from input to end use capturing capacity, utilization, cycle time, yield, scrap, lead time, lane reliability, and inventory rules. Supplier and buyer concentration indicators are computed. Single points of failure are stress tested using demand and supply shocks. Scorecards monitor quality, on time delivery, cost variance, and risk signals.
Method 1 uses regression models with predictors such as production, utilization, trade flows, capex cues, orders and backlog, pricing corridors, approvals or standards, and policy variables. Transformations appear only when out of sample performance improves. Models undergo stability checks and break tests and are retained after backtesting with diagnostics.
Method 2 applies driver growth rates to the most recent audited baseline where series are sparse or volatile. Weights are tuned to realized outcomes without overfitting and updated when drivers rebase or shocks emerge. Triangulation draws from Product Category Analysis, n per Population intensity, and an Economic Envelope to ensure feasibility under macro and budget boundaries.
Accuracy for regression structures is monitored through MAPE with residual checks and stability diagnostics. Growth rate models are validated through correlation measures and backtests versus prior vintages with tolerance bands that reflect category volatility. Apparent consumption is reconciled, and price corridors and capacity or FX boundaries are tested. A second analyst signs off on outputs, and all changes are logged with clear rationale.
The following domain signals are prioritized, each with a defined cadence and lineage card. These augment generic measures to capture sector-specific realities for brand share, channel mix, and forecast behavior.
Value = Units × ASP; Installed Base = Σ Shipments − Retirements; Apparent Capacity = Production / Capacity Factor; Brand Share = Brand Value ÷ Category Value; Channel Mix = Channel Revenue ÷ Total; Export–Import Balance = Exports − Imports; Apparent Consumption = Production + Imports − Exports ± Stock Change.
UN Comtrade / ITC Trade Map; Eurostat/PRODCOM; US Census/BEA/BLS; OECD; IEA/EIA/USGS/UNCTADstat; national statistics offices; EU TED / SAM.gov tender data; sector regulators and standards bodies. Lineage cards include pull dates, access type (public/licensed/consent), transforms, confidence tiers, and caveats.
Quarterly baselines; monthly micro-updates for high-frequency indicators; immediate notes for policy changes, outages, recalls, or spikes. Deliverables include an executive memo, workbook/models, and 16:9 dashboards with lineage chips and a change log.
Methodology
Thank you!
You will receive an email from our Business Development Manager. Please be sure to check your SPAM/JUNK folder too.
