Econometric Modeling & Driver Analysis
Patient behavior, market access, price, promotion. We isolate what's actually moving your demand, measure each driver, and turn it into decisions your commercial and planning teams can act on.
Life Sciences
Brand volume is the product of patients starting, staying on therapy and getting reimbursed. We model each step from patient-level data so forecasts hold up in brand planning and finance review.
What we factor in
- New-to-brand starts (NBRx)
- Persistency
- Adherence
- Formulary and payer access
- HCP promotion
- Co-pay programs
- Competitive entry
Patient persistency
Survival curves (Kaplan–Meier, Cox models)
From longitudinal patient data (APLD), we estimate how long patients stay on therapy and which factors, such as line of therapy, payer or starting dose, shorten or extend it.
What it gives you: Forecast patients on therapy month by month and target where support programs extend duration.
Adherence and compliance
PDC, MPR and DACON
Proportion of days covered, medication possession ratio, refill gaps and daily average consumption translate patient counts into realistic units by dose, strength and pack size, cohort by cohort.
What it gives you: Convert patients into units and revenue without assuming perfect compliance.
Gross-to-net drivers
Rebates, chargebacks, 340B and Medicaid
We model how formulary position, Medicaid and 340B mix, chargebacks and contracted rebates turn gross sales into net revenue across retail, specialty and government channels.
What it gives you: Forecast the net revenue finance reports, not just gross demand.
Co-pay program impact
Co-pay cards, accumulators and abandonment
We measure how co-pay cards and patient support change first-fill abandonment and refill rates, and how accumulator and maximizer programs shift out-of-pocket cost across the benefit year.
What it gives you: Size program budgets on the starts and refills they actually protect.
Market access effects
Formulary tier, prior authorization and step edits
We quantify how formulary wins and losses, tier changes, prior authorization and step edits move new starts and approval rates, plan by plan and region by region.
What it gives you: Put a volume and revenue value on each contract and access decision.
Promotional response
HCP detailing, samples and carryover
Response curves link calls, samples, speaker programs and non-personal promotion to new prescriptions by prescriber segment, with carryover, so you see where extra effort still pays and where it saturates.
What it gives you: Reallocate field and promotional spend toward the segments that still respond.
Consumer Goods & Retail
Volume moves with price, promotion, distribution and the calendar, often all at once. We combine syndicated point-of-sale data with internal shipments and deal costs to separate each effect so commercial, revenue management and supply teams plan from the same numbers.
What we factor in
- Base and promoted price
- Promotion depth and frequency
- NielsenIQ and Circana POS data
- Feature, display and deal costs
- Distribution (ACV)
- Seasonality and holidays
- Competitor pricing
- Media and advertising
Promotional lift decomposition
Baseline vs. incremental volume on POS and shipments
Combining NielsenIQ and Circana point-of-sale data with internal shipments and deal costs, we split each event's sales into baseline and incremental volume, net of the post-promotion dip and forward buying, so lift isn't overstated by pantry loading or retailer stocking.
What it gives you: See which events paid back their trade spend and which only shifted volume in time.
Price elasticity
Own- and cross-price elasticity
Log-log demand models estimate how volume responds to price by SKU, channel and region, and where that volume goes when a competitor or sister product moves price.
What it gives you: Set list and promoted prices with a known volume trade-off.
Cannibalization and halo
Source of volume
We measure how much of a promoted item's lift comes out of your own products, and which items pull related purchases along with them across the category.
What it gives you: Judge promotions on portfolio margin, not single-item volume.
Calendar and weather effects
Moving holidays, trading days and weather
Shifting holidays such as Easter and Thanksgiving week, trading-day counts and weather are modeled separately from trend, so a late Easter isn't read as a demand problem.
What it gives you: Compare periods fairly and stop reacting to calendar noise.
Key-account profitability and matched-market tests
Retailer P&L simulation and test-vs.-control
We model what a promotion or price change earns key retail accounts (such as Walmart, Costco, Kroger and Whole Foods) as well as the manufacturer, and design matched-market tests to prove incremental lift before rolling events out chain-wide.
What it gives you: Go into key-account negotiations with a calendar and P&L that work for both sides.
Marketing mix
Media ROI and adstock
Marketing mix models separate the effect of TV, digital, retail media and trade spend on sales, with adstock carryover and diminishing returns, so each channel gets credit for the volume it drives.
What it gives you: Shift budget toward the channels with the highest incremental return.
Related case studies
Johnson & Johnson: gross-to-net forecasting and reconciliation
Johnson & Johnson partnered with RoadMap Technologies to unify class-of-trade demand forecasting with Gross-to-Net financial reconciliation, closing the gap between supply chain planning and net revenue realization across its healthcare portfolio.
Retail, hospital, specialty & govt. · GTN reconciliationKraft Foods: trade promotion forecasting
Kraft Foods partnered with RoadMap Technologies to replace sales-driven overrides with statistical baseline forecasting, econometric promotional lift modeling, and retailer-level profit simulation across its trade planning process.
Multi-account retail · Trade promotionEnergy & utilities: forecasting load and consumption under weather-driven volatility
RoadMap applied its econometric and time-series forecasting models to energy and utility operations, where consumption is driven by weather, seasonality, and long planning horizons, and where forecast error translates directly into grid reliability risk and capital cost.
Load, weather & multi-horizon capacity
Frequently asked questions
What is econometric modeling & driver analysis?
Econometric modeling and driver analysis measures how patient behavior, market access, price, promotion and seasonality each move demand. RoadMap Technologies helps life science and consumer goods teams by building statistical models on their own data, from survival curves to elasticity models, so every forecast comes with the why behind the number.
How do you measure price elasticity?
With demand models fitted to your sales history by SKU, channel and region, controlling for promotion, distribution, seasonality and competitor prices. The result is the percentage change in volume for a 1% change in price, plus where the lost or gained volume goes.
How is patient persistency modeled?
Persistency is how long patients stay on a therapy. We model it with survival curves, such as Kaplan–Meier and Cox models, fitted to anonymized patient-level data (APLD), and break it down by cohort, line of therapy, payer and starting dose.
What is DACON?
DACON is daily average consumption, the average number of units a patient uses per day, measured from prescription data. Combined with proportion of days covered and refill gaps, it converts patient counts into realistic units by dose and pack size.
How is promotional lift measured?
By estimating the baseline, meaning what would have sold without the promotion, and treating the rest as incremental volume. We net out the post-promotion dip, forward buying and cannibalization of your own products, so lift isn't overstated.
Start with a pilot
Measurable outcomes agreed upfront, a live model on your data in about two weeks, and the ROI proven before you scale.