The Metrics of Usability If a Recommendation Falls in the Forest, Does It Make a Sound?   March 12, 2005

Synopsis: Is usability productive work? How productive? Can we prove it to management to justify our existence? Or is the issue not one of work but results?

The subject repeatedly arises of how to quantify the efforts of usability experts. Should we measure the number of tasks that we perform... reviews, tests, recommendations, designs, mockups, reports... and then pat ourselves on the back proudly declaring ourselves the most valuable employees in the company? Or is the real issue results... starting at the end: increased sales, and working backward from there? Answer: the latter, and anything short is subterfuge or malicious compliance, pandering to corporate silliness.

Despite my conviction that attempting to reduce usability to numbers is a telltale sign of a business that is simply dying but no one knows yet, I nonetheless accept that many businesses are incapable of tying authentic judgment to business decisions, so here are some ideas on genuine metrics:

1. Change (inversely correlated) in all costs and logistics related to training, documentation, and techsupport:
   1. - duration of calls
   2. - quantity of calls
   3. - length, quantity, cost of training classes, internal and external
   4. - pages printed, in development and techsupport as well as for end users
   5. - copy paper purchased
   6. - printing costs
   7. - salary of techsupport
   8. - years experience required for techsupport
2. Change in customer referral rate.
3. Change in sales (or web "conversion" rate from prospect/free to paying customer).

These are all business (outwardly focused) results. Items such as "how many tests were conducted" are inwardly focused, and I would hope that as a relatively new field, we are not already going down that path. I understand the attraction to such objective measures, but there is a fundamental conflict: the easier something is to measure and directly correlate, the less meaningful it is. Sales, the ultimate metric, is easy to measure but difficult to apportion to usability. Customer satisfaction is easier to correlate to usability but most folks in the business hardly recognize a usability showstopper when it's right between their fingertips and their eyes. How can we reasonably expect customers to make the correlation?

Usability is about results. Educate those who imply less or ask you to perpetuate the notion by measuring something inwardly focused.