1) The MATH behind why the breakdown Matrix is only 2 dimensions

2) Why you should be OCD about Breakdown Matrix Exclusions.

THE MATH

Performance analytics scores are just counts at a point in time. The numbers themselves don't understand what data composed them, so its impossible *later* to say.... of the P1 (breakdown) incidents created (indicator) which ones were also CategoryA? PA has to log a score for every single number you're interested in. In my case, a separate score logged for "incident created that is a P1 (breakdown1) and CategoryA (breakdown2)". That's called the Breakdown MatrixAt *ONLY* 2 levels deep... the number of scores created is the number of elements in each breakdown multiplied together.Example Breakdowns:Priority: 5 elementsCategory: 12 elementsGroups: 400 elementsStates: 5 elementsTime of day: 24 elements.At *ONLY* 2 levels deep, you're storing 5*12*400*5*24 =

2.88 MILLION values.

At 3 levels deep, the matrix of values would be more stored values than the rest of your servicenow instance combined.

BREAKDOWN MATRIX EXCLUSIONS

Knowing that PA is already storing millions of scores per indicator, its worth taking a good look at your breakdown matrix exclusions. If you're SURE you'll never want Groups combined with Time of Day, add that to your exlusions! Given the velocity of geometric growth, every exclusion helps.

A POSSIBLE FUTURE

You can give this explanation until you're blue in the face. It doesn't make the desire go away.

I think the only way SN could facilitate that kind of breakdown is to alter the design from Exclusion to Inclusion. Instead of waiting for people to define the combos they DON'T want, instead force them to define the combos they DO want. This would cut out incalculable permutations and ensure only mindfully selected options existed.

But even this would need guardrails... because nothing would stop someone from creating a matrix so vast it still contains hundreds of trillions of values.