Making sense of the decision maker's instinct
8 April 2026·3 min read
Most organisations don't really have a data problem. They have a decision problem.
Data is everywhere. Dashboards, reports, models, forecasts. But when it comes to making a decision, especially a complex one, things slow down, fragment and become opaque.
Who decides? Based on what? Using which data? With what level of confidence? Those questions are often harder to answer than the technical ones.
Decisions often come from senior leaders using their experience, and their gut, then finding data to reinforce that opinion.
This is how the world goes round.
But what if AI tools have introduced a new way of supporting decision making?
From information to decision (the missing layer)
A lot of systems are built to provide information. Very few are built to support decisions. That gap matters.
Information tells you what is happening. A decision requires you to choose what to do next.
That involves:
- trade-offs
- uncertainty
- accountability
- time pressure
Most tools stop just before this point.
Decisions are structured (even when they don't look like it)
Decisions can feel messy, political, or intuitive. But underneath, they usually follow patterns.
- similar types of problems come up again and again
- the same actors are involved
- the same kinds of data are used
- the same risks are considered
Once you start looking for it, you can see structure. Not perfect structure. But enough to work with.
Decisions apply across organisations (not within them)
In complex systems, especially infrastructure, water, energy, cities, decisions are rarely made in one place.
They sit across:
- regulators
- operators
- investors
- delivery partners
Each group sees the system differently. Each uses different data. Each applies different rules. Each has a different instinct for what is right. And of course, there's the profit motive.
The result is inconsistency. Not because people are doing the wrong thing, but because there is no shared way of structuring the decision itself.
Decision making as a system
If you treat decision making as a system in its own right, you can define:
- the types of decisions that exist
- the steps involved in making them
- the data and models they rely on
- the roles and responsibilities of different actors
This is where approaches like ontology (framing concepts) and decision patterns (showing strategies) become useful.
They don't remove judgement. They make the structure visible.
There's a growing push for:
- transparency
- consistency
- explainability
- better use of data
- value for money
These goals cannot be met by improving datasets or models. They depend on how decisions are made.
AI makes this more urgent, not less.
If decisions are poorly structured, AI will scale that problem. If they are well structured, AI can support them.
AI can find decision patterns and make ontologies by mining case studies.
What changes in practice
For Visioning Lab, this leads to a renewed focus on context.
For us it was never about better data, models or dashboards.
It was about understanding context, seeing the landscape and making it recognisable.
AI tools help us establish:
- clearer decision types
- reusable decision patterns
- explicit links between data, models and outcomes
- shared language across organisations
Decisions already exist everywhere in a system.
They just aren't always visible.
Once you start to map them properly, you can improve them. Then everything else, data, models, systems, can add real value.
Making sense of the decision maker's instinct captures the expertise that we really need.