Posted Friday, January 30th, 2009 at 3:26 pm by Marissa (51 posts)
I like diagnosing problems. I love implementing solutions. And as some of my colleagues will tell you, more than anything else, I love to test. But it’s the final step that I loathe: clicking that dreaded “publish” button.
A few weeks back, we discovered a minor issue with the “print this page” functionality of a site. The issue affected only a small universe of browsers. I was able to diagnose the problem and implement a fix. However, that fix was more of a coding change than I had hoped for. I tested each several pages using each template in four browsers times two versions times four operating systems, (which translates to “I tested it a whole heckuva lot”). Everything worked. But when it came time to click that publish button, I hesitated. I was about to change this functionality on no less than 5,000 pages. Could I have possibly anticipated every possible outcome?
A few weeks earlier, we had just implemented some Google Analytics reporting that specifically tracked the use of “print this page”. So, if my change somehow caused a catastrophic failure of “print this page” on every page except the ones I had tested, the reports would show it.
So I pushed the publish button. Everything seemed fine. And then, for the next few mornings, I looked at the analytics reports. There had not been any drop off or unusual activity for the “print this page” function. In fact, “print this page” events increased. So not only could I confidently say that nothing was broken, but that we had fixed the original issue. Advanced Segmenting of the particular browser confirmed this.
That led to the discovery of yet another universe of use for Analytics – verification of cross-browser support. As always with analytics, it seems the possibilities are endless.