Intelligence analysts almost always decide on partial, contradictory and outdated information. Over the years they developed tools and ways of working that help them avoid locking onto one explanation, and keep moving even without the full picture. These are the nine tools I present in my keynote for organizations, and most of them work just as well in a boardroom.
1. From intelligence to management
In intelligence, every decision goes through four stages: collection, cross-checking, assessment and decision. In civilian organizations these stages happen too, but usually all at once, in the same meeting and the same sentence.
Separating them changes the quality of the decision: first collect, then check whether the sources confirm each other, only then assess what it means, and finally decide. Everyone can see what is known, what is assessed, and what exactly is being decided.
2. Solving problems under uncertainty
The full picture almost never arrives. Whoever waits for it never decides. Instead, ask: what is the minimum I need to know to decide? And what would make me change the decision if it came to light?
That lets you move step by step, set checkpoints in advance, and update your assessment as new information arrives, instead of stopping everything.
3. Creative, out-of-the-box thinking
When every familiar path has been tried, the problem is usually not a lack of effort but the question itself. Reframe it, look for sources nobody has checked, and ask who else holds a piece of information you haven’t thought of.
Sometimes the solution is not where everyone is looking, precisely because everyone is looking there.
4. Ownership for employees and managers
A long task with no promise of success needs an owner: someone who holds it over time, even when there is no visible progress and the organization’s attention moves to more urgent things.
A manager’s job is not only to take that ownership but to pass it on to the team, so everyone understands why their work matters to the outcome.
5. Collaboration inside and outside the organization
Important information is almost always scattered across several bodies, each with its own interests, language and priorities. Each holds a piece, and nobody sees the whole picture.
Connecting them takes a shared goal everyone agrees on, trust built slowly, and an orderly way to share information without any side feeling it is giving something up.
6. Handling pushback from the top
Senior management has often already decided that something is impossible. Insisting without evidence won’t change that. What works: present orderly facts, show that you also tested the possibility that you are wrong, and propose a small step that can be checked.
Just as important: knowing when persistence is the right thing, and when it is time to stop.
7. Advanced research methodologies
Devil’s advocate (Ipcha Mistabra)
In Aramaic, “ipcha mistabra” means “the opposite is likely”. In this approach someone is given a defined role: to argue deliberately against the accepted assessment and see whether it holds. Israeli intelligence adopted it after the Yom Kippur War, to guard against fixed conceptions.
In business it can be simple: before a big decision, one participant presents the strongest case against it.
Analysis of competing hypotheses
Analysis of Competing Hypotheses (ACH) was developed by Richards Heuer of the CIA and described in his book Psychology of Intelligence Analysis. Instead of picking one explanation and looking for support, you write down every possible explanation and test each piece of evidence against all of them.
The preferred explanation is not the one with the most supporting evidence, but the one contradicted by the least.
A Popperian approach: look for what would disprove you
The philosopher Karl Popper argued that a scientific theory is tested by whether it can be falsified, not by a collection of confirmations. Working with information, that means asking: what evidence would prove me wrong? And then going to look for it.
A mathematical model (Leading Factor)
When many factors affect the outcome, a mathematical model helps identify the leading one, and focus collection and checking on it instead of spreading the effort thin. In the keynote I show how it worked in practice.
The Rumsfeld matrix
In February 2002, US Secretary of Defense Donald Rumsfeld described three kinds of knowledge: things we know we know, things we know we don’t know, and things we don’t know we don’t know. Over time a fourth quadrant was added: things we know without realizing it.
Mapping a project or a decision onto the four quadrants quickly shows where it is worth investing in collecting information, and where the real risk lies.
8. Sticking to the truth, and fighting for it
Long work brings failures: directions that led nowhere, assessments that turned out wrong. What matters is not giving up, and not abandoning the truth just because it is inconvenient.
As the Dalai Lama put it: when you lose, don’t lose the lesson.
9. Separating facts from assessments
The most basic rule, and the easiest to forget. “The supplier won’t meet the schedule” is an assessment. “The supplier was late with three of the last four shipments” is a fact.
The trouble starts when an old assessment, once voiced in a meeting, gradually becomes a “fact” that nobody checks anymore. The fix: always go back to the original raw data, and ask where every statement came from.
From the field: the search for the missing of Sultan Yacoub
I present these nine tools in my keynote “Bringing Home the Missing Soldiers of Sultan Yacoub”, through the search I led for eight years: facing partial and contradictory information, several research bodies each holding a different piece, and people who were sure there was nothing left to look for. About the keynote.
