The Retreat That Made It Worse
Michael Teo built a multi-seven-figure business out of a ragtag team of five or six people. TEDx curator. Global speaker. Flown to more than 120 countries. On a stage he was magnetic, and the room assumed the business ran the way he spoke.
In 2024 he brought me a problem that had been building for two years. His team was fighting. Nothing dramatic, no shouting matches. Something quieter and more corrosive. People had stopped feeling safe with one another. A remark from a colleague landed as a wound, and the wound stayed. The low-grade kind of damage that never shows up on a profit and loss statement and touches everything underneath it.
He and his operations head had a plan. A retreat. Take the team away for a weekend, let them have fun, and trust the bonding to do the work.
I asked him one question. "What will be different in the Monday meeting that was not true in the Friday meeting?"
He went quiet. Then he said, "Nothing. They will have had a good time."
That was the moment the real conversation started. Michael was not dealing with a morale problem. He was dealing with a thinking problem, and the retreat was aimed at the wrong room.
So we changed the intervention. Not team-building. A facilitated conversation with one rule: bring thoughts, not complaints, and not feedback. Thoughts. The kind of exchange where you hear how a colleague processes a problem and discover that you had been assuming a completely different internal logic for years.
Halfway through the second hour, one of his people said out loud what the room had been carrying privately. That they never knew which version of the truth was safe to say in that meeting.
Michael looked at me across the table. His face changed. He had been organising weekends and dinners and trust exercises around a symptom, and the cause had been sitting in the meeting cadence the whole time.
His team did not need a better weekend. They needed a shared way to think in public. Once that got named, the hope came back, and so did the clarity. Nobody went zip-lining. They went back to work knowing how each other makes a decision.
Symptom and Cause Are Different Rooms
The retreat would have been pleasant. The mistake was never the retreat. The mistake was its address. Michael had aimed his time, his money and his care at the symptom, and the thing producing the symptom was untouched.
I want to give you the structure I use to stop that from happening, because every founder I work with runs some version of it, and the cost is measured in quarters rather than hours.
The structure is called SCORE, and it separates a problem into five fields: Symptom, Cause, Outcome, Resources, Effects. Five fields, five different rooms of the same building. Walk into the wrong room and you can clean it perfectly while the leak carries on behind the wall.
Here is the distinction that carries the most weight. A symptom is what you can observe. A cause is what produces and maintains the observation. They live in different places, they answer to different instruments, and they rarely share an address.
Michael's symptom was visible to anyone in the room. Tension, guarded speech, people editing themselves before they spoke. His cause was structural and invisible until you went looking for it. The team had never been given a shared method for thinking together in public, so every decision turned into a negotiation about who was allowed to be right.
Treat the symptom and you buy quiet. Treat the cause and you buy a different business.
Here is the uncomfortable truth about most founder problem-solving. It is symptom management with a plan attached. The visible sign gets the attention because it is visible, and the mechanism underneath keeps producing.
Where the SCORE Frame Came From
SCORE comes out of the NLP modelling tradition. Robert Dilts set it out in Changing Belief Systems with NLP in 1990, and developed it further with Todd Epstein in Tools for Dreamers in 1991. The five letters stand for Symptoms, Causes, Outcome, Resources and Effects. The frame was built by watching how effective problem-solvers arranged their thinking, and then writing that arrangement down as a sequence anyone could run.
The history explains what SCORE is good at and what it is not. It is a modelling instrument, drawn from observation of skilled practitioners, and it has not been validated the way a clinical protocol is validated. Treat it as a structure for your attention, and take the evidence for its components from the separate research traditions that studied them. There is solid work on problem formulation, on solution fixation and on diagnostic error, and I will bring that in honestly rather than dress the frame in lab coats it never earned.
Where SCORE sits in the wider toolkit matters as well. The well-formed outcome frame answers what you want. The ecology check answers what else changes when you get it. SCORE answers an earlier question, which is what is actually going on here. Run SCORE before the outcome frame, and you stop building careful plans on top of a misread problem.
There is a piece of lineage underneath the acronym. Bandler and Grinder's original work in The Structure of Magic in 1975 was about recovering the structure beneath a vague statement, and their later training transcripts in Frogs into Princes in 1979 show the same instinct applied to behaviour. Look past the label a person gives a problem and find the structure generating it. SCORE is that instinct organised into five fields.
Why the Mind Grabs the Nearest Problem
The mechanism underneath all of this is not laziness. It is architecture, and it is worth understanding because it explains why capable founders walk into the wrong room with total confidence.
A hard question gets swapped for an easier one. Daniel Kahneman describes a move called attribute substitution in Thinking, Fast and Slow, published in 2011. When the mind meets a question that is expensive to answer, it quietly replaces it with a related question that is cheap, then returns the cheap answer as though it answered the expensive one. "Why is the team fighting" is expensive. "How do I make the team feel better" is cheap. The retreat is the cheap answer to the expensive question.
A familiar method gets reused past its expiry. Abraham Luchins ran the water-jar experiments in 1942 and named the effect Einstellung, a mechanisation of thought. Once participants had learned a method for pouring water between jars, they kept applying it to problems that had a faster solution. Expertise does this. The founder who has fixed every previous crisis with effort and energy reaches for effort and energy again, and the jar is already full.
Objects get locked to their expected use. Karl Duncker described functional fixedness in his 1945 problem-solving monograph. The box is for holding the tacks, so it does not occur to you that it is also a platform. Substitute a retreat, a bonus scheme, a new hire or a software tool, and the pattern repeats. The instrument you already own becomes invisible as an instrument.
And the first plausible diagnosis gets closed early. Pat Croskerry wrote in Academic Medicine in 2003 about cognitive error in diagnosis, and one of the errors he names is premature closure. Once a plausible explanation arrives, the search stops, and the evidence that would have complicated it never gets examined. The first explanation is not the strongest. It is the first.
Three further findings show what this costs at the business level. Dietrich Dorner, in The Logic of Failure, studied people running complex simulated systems, and the pattern of failure was consistent. People acted before they understood, then read the system's pushback as bad luck. Russell Ackoff made the blunt version of the point in his 1979 paper in the Journal of the Operational Research Society: we fail more often because we solve the wrong problem than because we produce the wrong solution to the right one. Paul Nutt, writing in the Academy of Management Executive in 1999, found that roughly half the decisions taken in organisations fail, with poor problem definition among the leading causes.
There is a version of this in strategy research that founders find harder to dismiss, because it is quantitative. Markus Baer, Kurt Dirks and Jackson Nickerson published a study in the Strategic Management Journal in 2013 on what they called the microfoundations of strategic problem formulation. The quality of the way a team formulated a problem predicted the quality of the strategic decision that followed. The thinking before the plan decided the plan.
So the first move is not to press harder on the symptom. The first move is to separate the visible sign from the mechanism generating it, and to do that before committing a single resource.
The Five Fields, One at a Time
Run the five fields in order, and notice that the order carries meaning. Symptoms and causes come first because they are about what is. Outcome and resources come next because they are about what could be. Effects close the loop, because every solved problem creates a new one.
1. Symptoms: what you can observe
Write the symptom in terms a camera could record. Replace "the team is toxic" with "in the last three weekly meetings, two people spoke and five did not, and the two who spoke disagreed about the same topic each time". Judgements compress. Observations expand, and the expansion is where the information lives.
Founder version: "Sales are down" is a judgement wearing numbers. "Inbound leads converted at 11 per cent this quarter against 23 per cent last quarter, while the offer and the traffic source stayed the same" is an observation you can work with.
2. Cause: what produces and maintains it
Ask two questions. What produces this? What keeps it going now that it exists? Those are often different answers, and the maintaining cause is the one that decides whether your fix holds. A cause that has stopped operating but left its scaffolding behind will keep the symptom alive long after the original trigger is gone.
Michael's producing cause was a hiring and management pattern that had never given the team a shared decision method. His maintaining cause was the meeting design, which rewarded people for staying quiet and made the ones who spoke pay for it.
3. Outcome: what you want instead
State the outcome in the positive and in sensory terms. "Less fighting" cannot be aimed at, because attention does not move toward a negation. "A meeting where a person can say what they actually think and the room keeps working" is an outcome you can picture, and the picture is what you will test against.
Hold the outcome light at this stage. You have not finished diagnosing, and an outcome fixed too early turns into a defended position.
4. Resources: what you already have
Founders under pressure inventory their deficits. Run the other direction first. What skill, person, relationship, information or internal state is already in the building? Michael had the answer sitting in his operations head, who could hold a difficult conversation without taking sides, and he had six people who wanted the friction gone. That is a resource base. The retreat budget had been aimed at a symptom, and the same budget aimed at the cause bought far more.
5. Effects: what else changes when you solve it
This field separates a competent diagnosis from a costly one. Every intervention produces second-order effects, and the ones you do not name arrive as surprises. Ask what else moves in the system when this gets better. Who gains standing, who loses it, what process has to change, what commitment you are implicitly making.
When Michael moved from a retreat to a facilitated conversation, the effect was that the silence itself became visible, which meant some people were going to be exposed. That was not a reason to stop. It was information he needed before he started, so the room could be held while it happened.
Running the Diagnostic: The Protocol
Five steps, and the whole pass runs inside a focused hour for a real problem. I run it with a blank page and five headings.
- Write the symptom as an observation, and read it back without a cause attached. If your sentence contains the word because, you have smuggled a diagnosis into the description. Strip it and keep the recording.
- Ask what produces the symptom and what maintains it, then keep the two answers separate. Producing causes are history. Maintaining causes are live structure. Your fix has to touch the live one, or the symptom returns within a month.
- Test the cause against evidence before you act on it. Ask what you would expect to see if this cause were true, and what you would expect to see if it were false. Then go and look. This is the door most founders skip, and it is the one that stops you spending a quarter on a plausible story.
- State the outcome positively and run the effects check. Name what you want instead as something you can picture, then name what else changes in the system when you get it. Say the trade you are accepting out loud.
- Marshal the resources and close on one commitment with a date. One action, one owner, one date. A diagnosis that ends in a plan ends in nothing. A diagnosis that ends in a calendar entry has a chance.
Run that on Michael's problem and watch what changes. Symptom: guarded speech and silent meetings. Producing cause: no shared decision method had ever been installed. Maintaining cause: a meeting design that made speaking expensive. Test: if the cause were true, the same people would speak in small groups and fall silent in the full room, which is exactly what his operations head reported. Outcome: a room where disagreement is survivable. Resources: the operations head, the will of the team, one existing weekly slot. Effects: the silence would be exposed, and the cadence would have to change to protect people while that happened.
Same budget. Same week. A different target, and a different result. Diagnosis consumes an hour up front. Misdiagnosis consumes the quarter behind it.