4 ms·
There's a difference between evaluative metrics and diagnostic metrics. You have a handful of top-level metrics that tell you whether things are good or bad on
by Ozzie_osman 4y ago
There's a difference between evaluative metrics and diagnostic metrics. You have a handful of top-level metrics that tell you whether things are good or bad on a dashboard. When they change, you have to dig in to see why, so you need more metrics. You segment, you cohort, etc.
Otherwise, you see GMV is down, and you don't and won't understand why.
- supersrdjan 4y agoI heard an analogy that goes like this: As a driver, all I want to see is current speed and RPM. But I like knowing that my car tracks a hundred other things, too, because when it fails, those other metrics will come handy. Or alert me before it fails. I heard it in the context of people analytics[0] but it seems to apply everywhere. [0] https://youtu.be/WE2ePETzzYQ?t=1434 https://youtu.be/WE2ePETzzYQ?t=1434
- incongruity 4y agoYour car is also using metrics to keep you safe, maybe even alive at many moments. Active stability control and ABS are massively data-driven and essential parts of modern driving safety and are always on. I think there’s also an issue of layers of decision making and the requisite metrics at that layer. Abstraction is useful in business just as it is in operating a car.
- ravivyas 4y ago> evaluative metrics and diagnostic metrics I am stealing this, I called them "Opportunities & Firefighting" & "Heath / KPI Metrics" in a recent post
- CyanBird 4y agoThere is other quote Don't be data-driven, be data-informed