02 · PILOT
Prove the value before the platform. Prove the platform before the program.
A 12-week value pilot on real data, a real decision, and a real P&L target. The output is a defensible value case and a platform decision the enterprise can defend in any board meeting.
Why a pilot, not a POC
A proof-of-concept proves that the demo works on sample data. A value pilot proves that the platform changes the P&L on your data. These are different exercises. We only run the second one.
Most platforms can be made to look good in a demo environment. Far fewer survive contact with the enterprise’s actual data, actual variability, and actual decision-making cycle. The pilot is where we find out — at a scale small enough to be cheap, large enough to be defensible.
What we produce
| Artifact | Description |
|---|---|
| Scoped decision | A single consequential decision — inventory policy for a SKU family, sourcing for a part class, capacity allocation across a plant network — that is large enough to matter and small enough to ship in 12 weeks. |
| Real-data load | The pilot runs on the enterprise’s real master data, real history, real variability. Not a sample. Not a sandbox. |
| Decision policy v1 | A signed-off policy describing what the platform may auto-commit and what must route to a human. The artifact the platform actually enforces in production. |
| Defensible value case | A pre/post comparison on the scoped decision, modeled against the existing operating cycle, with the assumptions audited by the customer’s finance function. |
| Platform decision | A go / no-go on the platform, with the reasoning documented. If the answer is no, the enterprise is no worse off than before — and considerably wiser. |
How we run it
- 01·
Weeks 1–2 · Scope lock
We pick the decision, the data scope, the value target, and the acceptance criteria. The customer’s finance partner signs off on the assumption set. This is the gate.
- 02·
Weeks 3–6 · Load and calibrate
Data extraction, semantic mapping, variability calibration. The platform learns the customer’s actual operating reality. We do not fake the data.
- 03·
Weeks 7–10 · Run the decision
The platform runs the scoped decision against the existing operating cycle in parallel. We compare. We document divergences. We refine the policy.
- 04·
Weeks 11–12 · Readout
The value case, the policy, the platform decision. A single document. A board-ready conversation.