VARIETYPORTAL
Route My Problem

The mechanism

Six questions. Six shifts. Six scales.

Six interrogatives × six relational modifiers × six quantitative scales produce 216 functionally distinct cells. No two cells ask the same question, because every pair differs on at least one orthogonal axis.

That completeness is the whole point. The grid defeats confirmation bias by construction rather than by discipline — you cannot skip a category of explanation because it did not occur to you.

The three axes

Every cell is a coordinate.

Interrogatives

The journalistic vector — which question you are perturbing.

Who
Which people, roles, teams or organisations are in the frame — and which were never named.
What
The object being acted on: the offer, the asset, the process, the decision.
Where
The channel, market, system or physical place the symptom appears in.
When
Sequence and timing — what came before, what is driving the deadline.
Why
The stated reason, and the reason that would have to be true instead.
How
The mechanism: how the work is actually performed, not how it is described.

Relational modifiers

The directional shift applied to that question.

Direct
The stated or obvious reading.
Opposite
The exact antithesis of the stated reading.
Other
An unrelated paradigm entirely — outside the current framing.
Keep Same
It did not change. That absence is itself the finding.
More Same
Amplified beyond the baseline.
Less Same
Diminished below the baseline.

Quantitative scales

How much of the system the reading applies to.

One
A single instance.
Partial
Some, but not most.
Many
A majority, short of everything.
All
Without exception.
Any
At least one, unspecified which — the existence claim.
None
Zero. The null hypothesis.

Worked readings

What a single cell actually does to a diagnosis.

Four coordinates, applied to symptoms we hear constantly. Each one turns a familiar complaint into a different — and testable — explanation.

Why × Opposite × All

“Our leads dried up when the new site launched.”

Direct reading
The redesign broke something and cost you leads.
Cell reading
Invert the causation and apply it universally: every lead source declined, and the redesign merely coincided. If all channels moved together, the cause sits above the website — pricing, category demand, or a change in who is searching.
What would test it
Compare pre- and post-launch volume across every channel, not just organic.

Who × Other × One

“Sales and marketing keep blaming each other.”

Direct reading
The two teams are misaligned and need a shared definition of a qualified lead.
Cell reading
Look outside both teams for a single unnamed actor. Often it is one person — a fulfilment lead, a finance approver, a single support rep — whose queue silently sets the real conversion ceiling.
What would test it
Trace ten won and ten lost deals end to end and record every human who touched them.

When × Keep Same × None

“We have to decide before the end of the quarter.”

Direct reading
The quarter-end deadline is driving the decision.
Cell reading
Nothing about the timing actually changed, and no real deadline exists. If the thesis rests on urgency that does not survive inspection, the correct move may be to spend another month gathering evidence rather than committing budget.
What would test it
Name the specific consequence that occurs on the first day past the deadline.

How × Less Same × Partial

“The automation we installed made everything worse.”

Direct reading
The automation was the wrong tool and should be removed.
Cell reading
The mechanism was right but executed weakly and only part-way. Not a bad decision — an unfinished one, which leaves a visible seam where the manual and automated halves meet. Incompetence and incompletion are real explanations and the grid forces you to consider them.
What would test it
Map which steps were automated and which were left manual, then measure the handoff between them.

Triage

You will not be handed 216 things to do.

  1. Generate mechanically

    All 216 cells, without filtering. Cheap to produce, and the completeness is the point.

  2. Kill fast

    Most cells die instantly against facts you already have. A killed cell is documented negative space — it is what protects the recommendation later.

  3. Score the survivors

    Would this change the decision, and is it testable with evidence you can actually obtain? Both must be true.

  4. Write a test for each

    A hypothesis without a test is a rumour. Every surviving cell names the specific record, dataset or conversation that would confirm or kill it.

  5. Chase the top few

    Five to eight, no more. The output is a recommended first move, not a backlog.

The engine expands the possibility space. Human judgment reviews evidence, feasibility, risk and context. It is not a fully automated system and we do not present it as one.

Bring a symptom and see which cells survive.

Start the Problem Route Scan See the assessment