3 Tenets of Performance Metrics

Posted on by

I was talking with some really smart folks recently about performance metrics. We started with “what” and “why” – what metrics do you collect and why (i.e. what do you do with them). After some discussion, one of the participants said:

  1. Measure everything, because you don’t know what will matter
  2. Keep data forever, because you don’t know how far back you’ll need to look
  3. Individual Metrics are irrelevant, they must correlate with something else

Three tenets of performance metrics, if you will.

Not earth-shattering. You might even say common-sense. But I like how clearly and succinctly they capture what we should be thinking about when deciding the “what” and “why” for performance metrics.

Number 3 is key. In the words of a CTO I once worked with, it’s the “so what”. IOW, why should I care? How does (page load time, availability, etc.) impact my business?

For example, correlation (which isn’t necessarily causation, but a good place to start) may show that 1% drop in availability resulted in a 10% drop in page views – and PVs drive revenue. Suddenly, I can quantify the monetary impact of poor performance. That’s the “so what”.

One thought on “3 Tenets of Performance Metrics”

  1. Good tenets from a performance metric analyzer point-of-view. People used to always sneer at my thirst for all metrics available, or request to generate more, before figuring out which correlate to the problem at hand! I always seemed to have a much heartier appetite than most for the metrics.

    Another tenet per #3 is is the old adage of turning metrics into information, specifically useful information to those that can make a difference. I’ve always found two avenues for this:

    1. Recognize metrics you observe that important to the business and communicate (per your “so what” example).

    2. Recognize “pain” being experienced by the business/executives, and root out the appropriate metric to improve and follow. In the old CompuServe days, the “pain” was the VP of Ops being pinged whenever other executives couldn’t load the CompuServe Main Menu! The metric ended up being availability, which led to a whole treasure trove of other metrics (Capacity, publishing, server metrics…)

    Mark

Comments are closed.