Behaviour Intelligence: Understanding Behaviour Before Deciding What to Do
Behaviour is visible; the cause isn't. Learn how to separate what you observe from what you assume, and decide what to do with better understanding.
Executive Takeaways
- When someone's behaviour changes, your first decision isn't what to do. It's how much you need to understand before you act.
- An observation describes what happened. An interpretation explains why. A useful test: could someone who disagrees with your explanation still accept your description?
- A person is vivid and a setting is mostly invisible, so the explanation that mentions the person comes easily. The one that mentions the setting needs a deliberate question.
- Scores, surveys and annual reviews are all snapshots. A pattern tells you which way things are moving, and that's usually what the decision needs.
- Move from behaviour to pattern to context to conversation before a decision is made, so understanding comes before action.
Seen, explained, decided
Imagine a strong team member who has gone quiet in meetings. They used to challenge ideas; now they nod and move on. By Friday their manager has a theory: they've lost interest, or they're unhappy in the role.
Nobody has asked a question yet.
The behaviour is visible. The cause isn't. Yet the theory is already shaping what happens next: a quiet word about engagement, perhaps, or a different project, or nothing at all.
We make a quick jump: we see something, decide why it happened, and decide what to do, sometimes within minutes. Each step feels reasonable. Together, they rest on very little.
This guide sets out another way to work, which we call Behaviour Intelligence. Before you decide what to do about behaviour, understand it.
We tend to jump from "I saw something" to "I know why" to "I know what to do". Behaviour Intelligence slows that jump down. Observe first. Look for the pattern. Ask what changed around the person. Have the conversation. Then decide.
The decision you are really facing
When someone's behaviour changes, your first decision isn't what to do. It's how much you need to understand before you act.
Many people decisions are interventions chosen on the back of an explanation. A development plan assumes a skills gap, a warning assumes a choice, and a new project assumes boredom. If the explanation is wrong, the intervention misses.
Imagine three cases: training for someone whose workload has doubled, a formal conversation with someone unsure what their role now includes, and a survey sent when a direct question was needed. Each response is reasonable for one cause. Applied to the wrong cause, each can make things worse.
None of this means you should delay everything. Some situations need an immediate response, such as safety or legal issues. The point is smaller. Know what you're assuming, and know when you're acting on a guess.
What we mean by Behaviour Intelligence
Here is the definition we use:
Behaviour Intelligence is the practice of understanding people's behaviour at work through observation, patterns over time and context, so that people decisions rest on understanding rather than assumption. It moves from behaviour to pattern to context to conversation before a decision is made.
The phrase is used in other ways, so it's worth being clear. One usage, in Wikipedia's entry, describes behavioral intelligence as "an individual's capacity to comprehend and impact social interactions" (Wikipedia). A survey of AI research uses "Artificial Behavior Intelligence" for machine analysis of human posture, expressions and behavioural sequences. The authors say: "The term Behavior Intelligence has not yet been standardized and is interpreted in various ways" (Jo et al., arXiv).
Some HR technology is marketed under the same term, typically for assessments and profiles (a vendor's public page). Our use differs: it looks at patterns over time and in context, not at a snapshot. It's about how people decisions at work get made, and how to make them on better understanding. It's about understanding, not watching.
Observation is not interpretation
An observation describes what happened. An interpretation explains why. The two feel almost the same in the moment, which is why they get mixed up.
A useful test: could someone who disagrees with your explanation still accept your description? "They went quiet in the last three meetings" passes. "They've disengaged" doesn't. It already contains a theory.
The table below shows how a single observation can support several explanations. It's illustrative, not exhaustive.
| What you observed | A common interpretation | Other explanations worth checking |
|---|---|---|
| Went quiet in the last three meetings | They've disengaged | Workload has spiked. They disagreed with a decision and held back. Their role changed after a reorganisation. They're waiting for direction. |
| Missed two deadlines this month | They're not committed | A dependency arrived late. The scope grew. Priorities shifted without the deadline moving. |
| Didn't put themselves forward for the promotion round | They lack ambition | They don't think they're ready. The criteria weren't clear. The timing didn't suit them. |
Separating the columns takes a minute. It doesn't tell you which explanation is right. It tells you that you don't know yet, which is the useful part.
Why we jump to the person
Psychologists have a name for the shortcut. The fundamental attribution error is the tendency to explain someone's behaviour through their personal characteristics while underestimating the situation around them. Jones and Harris reported early experimental work on the effect in 1967. Lee Ross gave it its name in 1977 (Jones & Harris, 1967; Ross, 1977).
It shows up in judgements about promotion too. Swift, Moore, Sharek and Gino tested it with 129 executive education students, who had an average of 18 years of work experience. Each played an airline CEO choosing a candidate for a senior role. They saw ten years of on-time data for each candidate's airport (Swift et al., 2013, PLOS ONE).
Some airports were easy, with about 85% of flights historically on time. Others were difficult, at about 70%. Every candidate's results were 5% below, at, or 5% above their own airport's average. So relative to their own setting, candidates in easy and difficult airports performed identically.
They weren't rated identically. At the same relative performance, candidates from easy airports were rated higher for promotion worthiness. The gap was between about 0.5 and 0.65 points on a seven-point scale.
The authors read this as a pattern. Decision-makers treat strong nominal performance as evidence of ability without discounting enough for how easy it was to achieve.
Be careful with what this does and doesn't show. It's one experiment. The participants were students role-playing a CEO, not managers deciding about colleagues they know. It shows how the bias can appear in a promotion judgement. It doesn't tell you what happens in your organisation.
It does suggest where to look. The study shows candidates from easy airports rated higher than those from difficult airports at the same relative performance. Our reasoning, which the study didn't test, is that the same habit could judge the same person more harshly after their conditions get harder. A person is vivid and a setting is mostly invisible, so the explanation that mentions the person comes easily. The one that mentions the setting needs a deliberate question.
Patterns, not snapshots
A single observation is a moment, but decisions are rarely about moments. They're about where someone is heading.
Go back to the quiet team member, and suppose we look at the previous three months instead of the last three meetings. Imagine the team was reorganised in the summer. Their remit changed, and two of their usual collaborators moved elsewhere. Their written updates are as frequent and sharp as ever, so they're still contributing. The room where they used to contribute has changed around them.
The picture is different now. It isn't a person who has lost interest but a person whose working context shifted at a specific moment.
Patterns raise better questions than moments do. Is this new, or has it been building? When did it start, and what else changed then? Does it show up everywhere, or only in one setting, and does it come and go?
Scores, surveys and annual reviews are all snapshots. They tell you where someone was on a given day. A pattern tells you which way things are moving, and that's usually what the decision needs.
The model: Behaviour, pattern, context, conversation, decision
We use one sequence to hold these ideas together, as our way of thinking about the problem rather than a proven method.
Behaviour. Describe what you saw, in terms someone who disagrees with you could accept. Leave the explanation out for now.
Pattern. Place it in time. Is this a one-off, a trend, or a return of something earlier? What does the shape suggest?
Context. Ask what changed around the person. Role, workload, priorities, relationships, expectations, and wider organisational change all shape behaviour. Most of these are invisible from the outside.
Conversation. This is the human bridge. You can't read a cause off someone's behaviour, so you ask. How you ask matters. A first question that already contains a diagnosis can steer the answer, so the first question is worth planning. Compare "I've noticed your performance has dropped. What's going on?" with "Help me understand what's changed for you recently." The first already contains a judgement: it treats the drop as settled and invites a defence. The second makes no claim and leaves room for the answer.
Decision. Now you decide, and you decide knowing which parts of your reasoning are evidence and which are hypothesis. Then you watch what happens and adjust.
You don't need software to use this. A notebook and honest questions will do. What matters is that the order holds, so understanding comes before action.
What this changes in your next decision
Next time someone's behaviour changes, try five questions before you act.
- What exactly did I observe, and could someone who disagrees with me accept that description?
- What else could explain it?
- What changed around this person, and when?
- What's the pattern over time?
- What do I need to ask before I decide?
You may still reach the same decision. If so, you'll reach it knowing why. And sometimes the question you didn't ask turns out to be the one that mattered.
That is the thinking behind Opptimise, which helps people leaders and managers understand behaviour well enough to make better people decisions. Better understanding comes before better action.
Questions people ask
What is Behaviour Intelligence?
Behaviour Intelligence is the practice of understanding people's behaviour at work through observation, patterns over time and context, so that people decisions rest on understanding rather than assumption. It moves from behaviour to pattern to context to conversation before a decision is made.
What can Behaviour Intelligence tell you about a person, and what can't it?
It can help you describe what you observe, see how it has changed over time, and ask better questions about context. It can't tell you who someone is, label them, or reveal the cause of a behaviour by itself. A signal prompts a question; it doesn't answer one.
Is Behaviour Intelligence an assessment or a personality test?
No. An assessment or test gives you a snapshot at one point in time. Behaviour Intelligence looks at patterns over time and in context, and it uses conversation to test explanations.
Sources
- Swift, S. A., Moore, D. A., Sharek, Z. S. & Gino, F. (2013). Inflated Applicants: Attribution Errors in Performance Evaluation by Professionals. PLOS ONE, 8(7): e69258. doi:10.1371/journal.pone.0069258. Full text on PubMed Central
- Jones, E. E. & Harris, V. A. (1967). The attribution of attitudes. Journal of Experimental Social Psychology, 3(1), 1-24. doi:10.1016/0022-1031(67)90034-0
- Ross, L. (1977). The intuitive psychologist and his shortcomings: Distortions in the attribution process. Advances in Experimental Social Psychology, 10, 173-220. doi:10.1016/S0065-2601(08)60357-3
- Wikipedia, Behavioral intelligence, retrieved 2026-10-08.
- The Predictive Index, vendor page, retrieved 2026-10-08. Cited only to show the term is used for assessments.
- Jo, K. et al. (2025). Artificial Behavior Intelligence: Technology, Challenges, and Future Directions. arXiv:2505.03315, retrieved 2026-10-08.