Engagement
What a first engagement involves
One live decision that has not been made yet. The people who own it. Your numbers. Weeks, not months. There is no enterprise AI programme to launch first.
Five steps
The decision, its constraints, the outcome expected — and what makes it hard to run.
The method, which engine fits the decision, and a demonstration on a situation like yours.
Inputs built with your team, then the analysis run on the real decision — not a sample or a sandbox.
The answer, the assumptions under it, and which of them matter most.
Inputs revised against the discussion, the analysis re-run, the decision ready to sign.
Each step is scoped and priced before it begins. Nothing is open-ended, and the engagement can stop at any point with something usable already delivered.
What it costs
From US$35,000 for a first engagement: a fixed fee, agreed before work begins, and typically a small fraction of the value at stake. Ongoing use of an engine is licensed separately.
What you bring
- A decision not yet made, with a date attached.
- The three or four people who own it, including whoever signs.
- Your own numbers — volumes, margins, costs, timing.
What you will not need
- No IT project. Nothing integrates with anything.
- No data cleanup. Missing figures become assumptions and get tested.
- No adoption programme. The people accountable are the people who run it.
What you are left with
- A decision you can sign — and the record of how it was reached: every assumption, and who confirmed it.
- Continued use, under licence. A deck is a photograph of one moment. When a competitor moves or an assumption proves wrong, the engine re-runs; only what changed gets changed. The licence level depends on the functionality needed — a separate conversation, and only after the method has proved itself.
The same decision, six months later, does not start from nothing.
Depth to suit the decision
- Full configuration for decisions carrying real consequence — every criterion, the competitor's position modelled alongside yours, the challenge and audit layers, the full governance record.
- Stripped back for the routine ones — a price reviewed each quarter, a mid-sized award, a bid too small to justify a week of analysis and too large to guess at.
- Same structure underneath. That is what makes the light version trustworthy rather than merely quick.
The flagship decision was always going to get analysed. The value is in the forty that were not.
Why weeks, not months
The method is already built. What normally consumes the first half of an engagement — deciding how to approach the problem, designing the analysis, building the model — is not work you are paying for. The time goes into your decision instead.
Three ways an engagement runs
- Direct. Decisiums runs the engagement with the client team, on their decision, in their environment.
- Co-delivery with a partner firm. The partner leads the client relationship and brings industry context; Decisiums brings the method and runs the engine. Methodology depth and relationship depth are different capabilities, and the decisions worth this treatment usually need both.
- Methodology transfer or licence. Partner firms and in-house consulting groups adopt the engines as a permanent capability, calibrated against real engagements rather than handed over in a slide deck. Licensing suits firms with the decision-analysis depth to run them unaided.
Nothing to host or integrate
Each engine runs as a self-contained application, with no infrastructure to set up and nothing to connect to existing systems. Continuity of the capability can be written into the licence agreement.
Discuss a live decision →