SEE EXACTLY WHAT’S DRIVING THE BUSINESS.
Get one unified causal view that shows WHY the business grows or shrinks. See every internal and external growth driver from enterprise to granular level decisions, across every time horizon.
Why are we growing (or losing)?
Until now, no single model could answer:
- Marketing says the campaign. They have an MMM to prove it.
- Sales says the distribution wins. They have the data.
- Finance says pricing. The margin numbers confirm it.
- Trade says the promo calendar. Lift studies back it up.
- Strategy says the category was up. Macro data supports it.
This isn’t a people problem.
It’s a math problem.
Each team’s correlation model can only analyze one function at once – they were never built to unify all the data.
POEM635 is different.
It’s built on causal inference, not correlation.
One complete answer – clearly broken down into cause and effect.
understand what drives your business
From the enterprise P&L down to the individual tactic. Every driver is connected in one model that shows you exactly what caused your business to move.
One unified growth decomposition
See what's actually driving growth across the entire business as one connected system. Every driver, every interaction, every offset. One view. No reconciliation.
Cross-functional interaction effects
See how pricing ripples into trade, how distribution gains amplify marketing ROI, how competitive moves reshape your landscape. These interactions are where data can be misread — and where unseen value lives.
Enterprise to tactic — one drill-down
Start at the P&L. Drill into a brand, region, channel, or tactic. Every altitude is connected by the same causal truth. No more "the brand model says X but the enterprise model says Y."
Causal scenario simulation
What happens if you raise price 3% and the competition follows? What if you shift $15M from broadcast to retail media? What if interest rates climb 100bps? The model simulates what would happen based on cause and effect.
True incrementality for every driver
The model separates what was truly incremental from what would have happened anyway. No more crediting the campaign for sales that were already coming. No more blaming the price increase for losses driven by the competitor.
Continuously learning — always current
New data flows in. The model sharpens. Annual truths become quarterly insights become weekly recalibrations. Same model. Same truth. Never stale.
How the model shows the causation of growth and loss
Total Revenue Growth
+4.2%
Pricing Execution
Price increase held, competitor did not follow
Media & Advertising
New campaign drove incremental awareness-to-purchase conversion
Distribution
200 new store authorizations, amplified by media weight in those markets
Trade & Promotion
Promo calendar delivered, but pricing offset affected trade lift -0.4% — shown as an interaction effect
Competitive Retreat
Key competitor pulled back on spend and lost distribution in 3 regions
Category Tailwind
Macro-driven category growth — not your action, the economy carrying you
Macro Tailwind
Inflation-driven consumer trade-down reduced premium tier volume
HOW your model is built
THE FOUNDATION
THE ENGINE
THE MODEL
Powered by a pre-trained model that already understands the deep patterns of how enterprise growth works and fine tuned to your business.
Faster to value. Deeper accuracy. One model that holds everything.
Fount: Large Causal Architecture
Unlike traditional models which rely on correlation, Fount is built upon causal inference — the mathematical framework pioneered by Judea Pearl. It provides a formal system for identifying what causes what, how causes interact, and what would happen under conditions that haven't been observed.
POEM365: a pre-trained Large Causal Model
POEM365 is built upon Fount’s architecture. It’s a transformer-based causal model trained on hundreds of billions of transactions across industries, categories, and markets. It already holds deep causal intelligence about how growth drivers behave and interact across enterprises.
Your model: tailored to your enterprise
The model then ingests your custom data on media, pricing, trade, distribution; as well as your competitive landscape, retail environment, and macro conditions.
It’s becomes is your customized model built for your business. Your causal truth.
What Lives Inside the Model
Four layers of growth drivers. One causal structure. Every driver interacts with each other.
Internal Growth Drivers
Competitive Dynamics
Retail & Channel Environment
External & Macro Factors
How the Layers Connect
POEM365 is built on causal inference which is what makes unification possible – unlike models that are built on correlation. This is essential because growth doesn't happen in isolated layers. It happens across layers simultaneously.
- Price change
- affects trade sensitivity
- affects promotional effectiveness
- affects competitive response
- affects share dynamics
- affects retailer behavior
HOW IS THIS DIFFERENT?
| POEM365 | MMM | TPM / TPO | RGM Systems | IBP / S&OP | BI Platforms | Consulting | |
| What it does | One unified causal model — every driver, every interaction, one architecture | Models media channel effectiveness | Manages & optimizes trade promo spend at retailer level | Models pricing, pack/price architecture, promo ROI | Aligns demand, supply & finance plans. Not a causal model — a planning process | Visualizes data — no model underneath | One-time frameworks & analyses |
| Scope | Total business: media, pricing, trade, distribution, competitive, retail, macro — unified | Media channels only | Trade & promo only | Pricing & promo only | Demand-supply-finance. No causal model of what drives demand | Whatever data you connect | Strategic-level |
| Causal vs. Correlation | Causal inference — knows what CAUSES growth | Correlation | Lift-model based. Correlation | Elasticity-based. Correlation | No causal model. Consensus & assumptions | No model | Frameworks, not causal models |
| Interaction effects | Every driver interacts — pricing × trade, distribution × media, competitive × elasticity | Mostly additive | None. Trade modeled in isolation | Limited. Price independent of media & trade execution | None. Aligns plans, doesn't model interactions | None | Sometimes hypothesized |
| Simulation | Change any input — whole system recomputes causally | Spend-response curves, media only | Promo simulation, trade only | Price simulation holding everything else constant | Demand scenarios on assumptions, not causal simulation | None | One-time scenarios |
| Continuously learning | Yes — sharpens with every new data signal | Quarterly or annual refresh | Updated with promo results, no cross-driver learning | Updated with pricing results, no cross-driver learning | Updated with actuals, no causal recalibration | Real-time data, no model | One-time deliverable |
| Speed to value | POEM365 pre-trained. Fine-tunes in weeks | Months from scratch | Months to configure per retailer | Months to build per category | Months to implement | Fast — but no model | Months of engagement |
| The fundamental gap | This is the unified model | Cannot see beyond media | Cannot see beyond trade | Cannot see beyond pricing | Not a causal model — cannot explain what drives growth | No model. No causation | Not live. Not continuous |
Who is this for?
CEO / President / GM
One answer to "what's driving growth?"
Across every function, every factor, every time horizon.
CFO / Head of Strategy
A growth model you can plan and forecast against – explanation and projection from the same architecture
Division /
Business Unit Head
Your P&L as a connected system. The interactions, the trade-offs, and where the real leverage sits.
Built for your world.
Our pre-trained models work within three industries, speaking your business language fluently.
.jpg%3F2026-04-21T14%253A04%253A51.561Z&w=3840&q=75)
automotive
Dealer performance, sales forecasting, market penetration, competitive analysis, territory optimization.
CPG
Brand performance, distribution optimization, market share strategy, category growth, competitive positioning.
.jpg%3F2026-04-21T14%253A04%253A53.979Z&w=3840&q=75)
Retail
Store performance, expansion planning, inventory optimization, competitive positioning, market share growth.
The model doesn’t replace functional expertise. It gives every function the same causal foundation — so the debates end and the decisions begin.
See what Data Poem can do for you.
Let’s talk about how we can help you grow your business.