It's one of those seemingly evergreen questions: Why is our bounce rate so high? Isn't that bad? To quote Bill Murray from “Ghostbusters”, I'm fuzzy on this whole good/bad thing.
First of all, what is a bounce? It's a visitor who comes to your site, looks at the one page he or she landed on, then leaves your site entirely.
Generally speaking, if you're selling goods on your website, you want people to stay, shop around, put items in the cart and check out. It's not too much of a stretch to say that a visitor needs to visit more than one page to accomplish that goal.
But we forget that people frequently visit or site for reasons other than to buy. Time and time again we're told to write content for people, not engines. So consider these two scenarios.
Customer Service Tracking
You purchased a bunch of furniture for your apartment from IKEA. Later in life, you move house. When moving, you need to disassemble and reassemble the furniture but you forgot where you placed the instructions. So you search for [ikea benno tv unit manual] and the first link you discover is the one and only page IKEA has for all of their products. The page is sorted alphabetically by product name. You have your entire needs met on the first page.
So, we can search, discover and land on the assembly instructions download page. We can then download to our heart's content and become even more satisfied customers who are even more inclined to buy again in the future.
Looking at the pure data, however, these are web analytics statistical bounces.
The site comes up first in your search query, gives you everything required from an after-sale customer service standpoint on one page. In and out of the site in a timely fashion, free to crack on with the important business of packing and moving. You wanted quick service and you got it on one page. Is that bad?
Lesson 1: If you aren't tagging your downloads or on-page events correctly, you could miss out on the full picture.
General Information Dissemination
I work at a large university. Our website serves several constituents, including currently enrolled students. These students sometimes need to know where and when to find a professor during office hours.
As the webmaster in charge of content and SEO, I should be able to have a student search for [professor thom craver rit] and find my office location, phone numbers and office hours in one quick search. That information should be readily available and easy to find. They search, click the first link, get the information they need and go along in grandiose ADD, late-for-class way.
This is great and all, but how do I know what those students' intents are looking at the bounce rate? Well, I don't. However, using segmentation, I can find those visitors who bounced and see what area of the site they landed on.
If the user lands on a page in the faculty or staff directory, I have a real good idea. If the visitor was from an on-campus IP address or even a mobile phone, I have an even better idea.
Either way, the student body has been served well, having happily found the information needed in a one-click process. Everyone wins, right?
Lesson #2: Segmentation is everything. It helps you discern "good" metrics from "bad" metrics.
Neither of these examples are far-fetched. If you only look at one metric in aggregate and don't compare that metric with other metrics or factors, you're missing the real story about your web visitors.