It’s easy to fall into a rut when looking at web analytics. Bounce rate, e-commerce conversion, return visits… all good stuff to know, but it’s in its application that web analytics becomes a powerful tool. So here’s my game plan for 2010 for becoming a better web analyticator. <sic>
Understand the Correlation of Bounce Rate and Visit Length
Bounce rate is a key indicator in the health of any landing page. Certainly, you want your visitors to take action when they visit your shopping cart, blog, or lead generation site. But there’s more to bounce rate in some instances than “I came, I saw, I puked”, as Avanish Kaushik so eloquently puts it.
I hope to find a “formula” based on web site topic that indicates WHY a visitor is bouncing. Is it a long visit, and then the visitor backs out? That might indicate you are close to having a great landing page, but just need a few tweaks. Or are a good percentage of your visitors actually puking when they get there, and gone in 10 seconds or less? If that’s the case, your landing page might need a lot more work, or at worst, a complete tear off.
Discover PPC Key Phrases that Simply Don’t Work
Google Analytics currently has some excellent new AdWords measurement tools in Beta. Once all of you can see them, you will need to use them. In 2010, I intend to find specific words buried in our long tail etc. that tend to not work for our site. Often, visitors search with our best converting key terms, but they add in a little something extra that either makes us or breaks us for conversion.
Coupled with my first goal for landing page value, I think I can better identify new negative words to add to our AdWords campaigns. Words that we pay handsomely for, even if unintentionally, yet never bring in the exact customer we are looking for.
Segments, Segments, Segments
If you have a popular website getting hundreds, thousands, or millions of visits per day, data can pile up pretty quickly. In 2010, I intend to set up better custom segments in Google Analytics so I can study success and failure far more granularly.
We sell 5 distinct product groups, so I should clearly have 5 segments created in GA to look at those. Similarly, I should do sub-segments for each of those groups. If I know that the average visit length for a lead generating visitor is 8 minutes, I should have a segment that studies visits of 7 minutes or less, to further help identify bad long tail spend, and get a feel for the web site flow of a bad visit.
Segments are very powerful. Be sure you are setting yours up to help your small business.
Above are my Big 3 for 2010. As we all know, the analytics industry gets new tools and new thinking almost daily. So this list is bound to be dynamic. But I think it’s important to have a framework to work against, so I built one.
What are you going to do to be a better web analyst this year? I’d love to hear!