do(price) on the data you already have.
Pricing decisions move real revenue, run on tight calendars, and rarely permit a fresh field cycle. Simulacra fits a causal model to your historical pricing and promotion data, then allows you to run interventional do(price = x) scenarios where unit volume, share, promo lift, and segment mix shift together.
Pricing reads always come too late.
Late MMM cycles
Market-mix models deliver elasticity months after the strategic decision needed it. By then the window has closed.
Conjoint cells are underpowered
A solid total-sample conjoint can still be unreliable for the priority subgroup or regional cut leadership wants.
Stress-tests miss the cascade
"What if competitor moves price 7% in Q3" doesn't have a comfortable home in either regression or conjoint outputs — neither tool cascades the intervention through the response structure.
Predict the whole demand curve under do(price).
When you intervene on price in Simulacra, the variables downstream of price move with it: unit volume, share, promo lift, and segment mix all respond. Simulacra predicts the demand curve under the new price.
do(price = x) cascades
Interventional inference, not P(volume | price). Variance, skew, and segment composition update under the intervention.
Competitor scenarios
Set both your price and competitor price simultaneously. Trade-off in volume reads as a coherent system, not two independent slices.
Region & segment splits
Run the same intervention across regions or segments to surface heterogeneous response, returned as CATE estimates with confidence intervals.
Set the price. Add the competitor move. Preview the curve.
Start with a product and market you already track, change your price and competitor price, and preview how unit volume, share, promo lift, and segment mix move together.
Product / market
Indexed to the observed shelf-price range for the selected product.
Competitor moves are modeled inside the same product-week context.
Your price moves +4% while competitor price moves -2%; unit volume, share, promo lift, and segment mix respond together.
- Unit volume
- -5.6%
- Share
- -1.3 pts
- Promo lift
- +0.5 pts
- Segment mix
- -0.7 pts
- Readout
- Price-volume curve, share movement, promo response, and region mix.
- Export
- CSV scenario table, response curve, assumptions log, methodology appendix.
- Guardrail
- Values outside the observed shelf-price range are flagged before delivery.
Workflow preview. Your run uses your product history, price ranges, promotion calendar, markets, and measured outcomes.
A Fortune-class FMCG ran the blind validation. We passed.
In a 2026 blind pricing & promotion validation, an East-Asia FMCG manufacturer held back 2023 weekly volumes. Simulacra trained on 2021–22 NielsenIQ pricing-and-promotion data and scored 195 held-out product-week cells: 172-unit median weekly miss, 83% inside product-level weekly volatility, and correct price-response direction for all four products.
172 median miss, 83% within weekly volatility, 4 of 4 slope signs
Blind validation methodology, per-product held-out plots, price-response curves, and residual calibration profile.
Read the validation→
Bring two years of weekly POS. Run do(price) on your existing data.
Bring your historical pricing and promotion data. Simulacra fits a causal model and lets you run interventional price scenarios where unit volume, share, promo lift, and segment mix move together.