Think of two web pages you’ve created in the past. Which of them will work best?
Before you answer, you should know that you’re not entitled to an opinion. The only people who are entitled to one are your visitors, and they convey their opinions through what they do on your site. If they do the things you want them to, in increasingly large numbers, that’s a good sign; if they don’t, that’s a bad one.
Analytics is that simple. Everything else is just details.
Before you go further in this chapter, we want you to try an exercise. Find a volunteer, and ask him or her to think of something he or she wants. Then draw a picture on a piece of paper, and get him or her to reply “warmer” or “colder” depending on how close the thing you drew was to the thing he or she was thinking. Repeat until you know what they’re thinking of.
Go ahead. We’ll wait.
How long did it take for you to figure out what he or she wanted? How many iterations did you go through?
Notice that you had plenty of creative input into the process: you came up with the ideas of what might work, and you deduced what your volunteer wanted through small “improvements” to your picture based on their feedback.
The same process takes place as you optimize a website, with some important differences:
You’re listening to hundreds of visitors rather than a single volunteer, so you have to measure things in the aggregate.
What you define as “warmer” or “colder” will depend on your business model.
If you know your audience well, or have done your research, your initial picture will be close to what your web audience wants.
You can ask visitors what they think they want, through surveys, and they may even tell you.
Your visitors are telling you what they want. You just have to know how to listen by watching what they do on your site, and react accordingly. Web analytics is how you listen. Use it wisely, and eventually the website you present to visitors will be the one they were craving.
For a long time, web traffic analysis was the domain of technologists. Marketers were slow to recognize the power of tracking users and tying online activity to business outcomes. Web activity was mainly used for error detection and capacity planning.
The growth of the Web as a mainstream channel for commerce, and the emergence of hosted analytics services—from Omniture, Urchin, WebTrends, CoreMetrics, and others—finally convinced marketing teams to get involved. Today, free services like Google Analytics mean that everyone with a website can get an idea of what visitors did.
There’s no excuse not to listen.
Imagine that you own a small store in the country. You don’t need much market analysis. You can see what people are doing. You know their individual buying preferences, their names, and their browsing habits. You can stock their favorite items, predicting what they’ll buy with surprising accuracy.
Now suppose business picks up and the number of customers grows. You slowly lose the ability to keep track of it all. Clients want more merchandise, which in turn requires more floor space. You have to make compromises. The sheer volume of customers makes it impossible to recognize them all. You lose track of their buying habits. Your customers become anonymous, unrecognizable. You have to start dealing in patterns, trends, and segments. You need to generalize.
Websites face similar issues. The Web pushes both visitor anonymity and traffic volume to the extreme. You can’t know your visitors—in fact, you can’t even tell if they’re male or female, or how old they are, the way you can in a store. You’re also opening the floodgates, letting in millions of complete strangers without any limits on their geographic origins or the times of their visits.
We’re going to provide a basic overview of analytics. Analytics is the cornerstone of a complete web monitoring strategy, and many other sources of monitoring—from performance monitoring, to customer surveys, to usability, to community management—all build upon it. Only by consolidating analytics data with these other information sources can you form a truly complete picture. Only through an integrated view can you return to the confidence of someone who’s intimately familiar with her market and customers.
There are many excellent books and websites that deal with analytics in much greater depth than we’ll attempt—in particular, with the need to focus on outcomes, to experiment, and to focus on data and facts rather than on opinions and intuition. We can recommend Jim Sterne’s Web Metrics: Proven Methods for Measuring Web Site Success (Wiley), Avinash Kaushik’s Web Analytics: An Hour a Day (Sybex), and Eric Peterson’s Web Analytics Demystified (Celilo Group Media). You can find Avinash Kaushik’s blog at www.kaushik.net/avinash.