About the method
One problem. One constraint. One next step.
Most performance problems get solved twice — first with the wrong fix, then, eventually, with the right one. This tool exists to shorten that gap. You describe one real result that is not moving, answer a fixed set of questions, and leave with the single constraint most worth attention first.
Step one: the Four Forces scan
The scan tells you where to look. Any result depends on four forces working together, and the weakest one directs the deeper questions.
Offer
Is the work worth doing—and well designed?
Actors
Are the people willing and able?
Context
Does the environment enable the work?
Binding
Is the whole being led into aligned action?
Step two: the Six Gears deep dive
The gears tell you what is stuck. They separate look-alikes that are usually confused — a skill problem treated as motivation, a means problem treated as skill, a grip problem treated as commitment.
- Aim
- People are working hard toward different finish lines, or nobody clearly owns the result.
- Will
- People can do the work, but incentives, friction, or consequences point them elsewhere.
- Skill
- Even with time, tools, permission, and motivation, people cannot yet perform reliably.
- Means
- People know and want to act, but tools, time, resources, access, or authority block them.
- Sync
- Individual parts work, but hand-offs fail, information arrives late, or people stay silent.
- Grip
- Decisions are made, but commitments fade between the meeting and the result.
- Loop
- The team repeats familiar actions even when evidence says conditions have changed.
How the score works
The scoring is plain arithmetic, not a hidden model. Each question is rated 0 to 4. Answers are averaged per force and per gear. The lowest-scoring gear becomes the primary constraint, the second-lowest becomes the linked constraint, and the distance between them sets the confidence label:
- Strong — a clear gap of one point or more between the first and second constraint.
- Moderate — a gap of half a point or more.
- Working — the two are close; treat both as candidates and gather more evidence.
Nothing is generated by AI at this stage. Identical answers always produce an identical diagnosis, so you can re-run it with colleagues and compare.
What the result is not
It is a working hypothesis about a system, not a verdict on people. It is deliberately narrow: one constraint, a few actions, one thing to stop, and one observable signal to watch. If the result does not move within the review window, the honest response is to return to the evidence and diagnose again.
Where the AI part fits
After the diagnosis, you can optionally turn the problem into an application concept. That step uses AI to propose a feature breakdown you then edit and select from. It shapes options; it never decides your diagnosis.
Credit
The questions, signals, and interventions are drawn from the Four Forces Method (Offer, Actors, Context, Binding) and the Six Gears (Aim, Will, Skill, Means, Sync, Grip, Loop) as set out in the source manuscript Fix That First. The framework belongs to that work; this tool is a practical instrument for applying it.