“Crafty” analysts will often ask, “If you could only have one metric, what would it be”.  I have a hard time answering this because I think it’s a dangerous proposition.  When you only look at one metric you don’t get a holistic picture of your website.  For instance, I’ve heard practitioners say, “Bounce is not important”.  By itself, no, it’s not that important because you don’t really know what the metric is telling you.  I’ve also heard, “Conversion is the only metric you need to worry about.”  Again, not true.  If my conversion is up +25% month over month, why?  I need more info.

Pairing metrics is crucial to analysis.  If you don’t, you’re simply regurgitating a data point to your listener.  You cannot glean actionable insights from a single metric.  Further, you can cause unwarranted panic if you just speak about a single metric.  “Our Average Order Size dropped -20% last week!”  “Oh man!” … “Oh my gosh, what does that mean?”

So now you ask, “Ok, what metrics do you recommend that I pair together?”  It can definitely vary depending on what you are trying to determine, but here are some of my more frequently used pairings:

• Traffic, conversion, AOS, and then by segment (new / return; organic / paid, etc) – mostly used when analyzing total site (ecommerce sites)

• Bounce, time on page, traffic and then by marketing source and by browser for top 5 entry pages – bounce segmentation can indicate if a certain traffic source is not being directed to the right page on your site, and as well can tell you if you might have some technical issues with a particular browser

• When measuring a video engagement: traffic to area, video category total starts, video category total completions (and thus an overall completion rate); top viewed videos by starts; top completed videos; videos viewed per visitor.  Also depending on your analytics provider you can get more specific such as pauses, average completion percentage (i.e. how many completed 50% of video or 75% of video) etc.

There are tons of metric pairings that help give you deeper insight into your projects.  My philosophy is to ponder whether the glass is half empty or half full, if I feel that it’s half empty, I look for other metrics to glean additional insights; if I feel that it’s half full and I’ve already dug deep enough, I stop before I succumb to paralysis by analysis.  Again, my key point is using one metric to analyze something is a dumb idea.  So don’t do it.