Before you start measuring your website you are in blissful ignorance.
You probably have the assumption that people come to your website through the front door of your homepage and that they look at things in an ordered manner that somehow relates to your navigation structure. They find the things they are looking for before doing exactly what they’re supposed to do (like book or enquire). Finally they leave fully satisfied, bookmarking you as they go for an easy return.
Once you start measuring your site - whether by web analytics, surveys, user testing or a combined approach - that happy illusion is shattered forever.
Chaos seems to rule. People enter the site all over the place, then leave in droves without seeming to do anything much at all. The deeper you look at the data, the more fleeting and varied that visitor activity seems to be.
Web analytics and user testing provide more information than you know what to do with, yet exactly the right bit still seems to be missing.
So how do you make sense of how people use your site - and what can you do about it?
First, forget everything you know about your site…
The navigation and linking structure, the fact that everything is explained on the homepage, those driving directions clearly given in the 34th paragraph. Your visitor knows none of this.
Things may make sense from the homepage but is perfectly possible that only a low proportion of visitors see the homepage anyway. (Believe me, I’ve seen businesses spending 80% of their effort optimising a page less than 25% of people even see. Meanwhile, visitors arrive from search engines into long forgotten, dusty corners of the site).
A first time visitor typically starts lost and often the experience just goes downhill from there.
So, when survey respondents tell you “I couldn’t find the…” they’re not lying, even if you know exactly where it is. Things are much less obvious when viewed for the first time and its critical to remember that when trying to make sense of how visitors behave. If you’re struggling, why not try some DIY user testing?
Second, know thyself…
Why do you exist? What is the website for and how does it relate to what your business is for?
This isn’t philosophising for the sake of it - when people use websites they are looking to solve problems and achieve goals.
What problems does your site exist to solve?
Think hard about what your different segments of visitor are trying to achieve when they come to your site. Given every visit starts with a problem or a goal, how good is your site at helping visitors achieve theirs? Understanding this context will be invaluable in making sense of how visitors use (or fail to use) your website.
Third, tame your tools…
Taming your tools means identifying your key questions and using your tools (from web analytics to surveys) to answer those questions. What tends to happen is that people let the tools lead them - so they find themselves measuring what the tools hand them (page views, visits etc) rather than what they really need to know.
Start with what you really want to know - think about your question different ways and break it down into smaller parts if you need to. Then look at the tools and data that you already have that may help you to answer those questions. That way, you don’t need to worry about all the information you have, you can just selective measure what really matters to you.
You can pick a few Key Performance Indicators and stick with them. Trends over time are far more important than absolute numbers and will start to give you insight into what is really going on.
But it is important to also tame your tools by understanding their potential and familiarising yourself with how they work. Making sense of how visitors behave doesn’t happen at the top level of information, it involves drilling down beyond the superficial -getting beyond page views, unique visitors and 1 to 5 satisfaction ratings.
If you only look at the front page of your Omniture or Google Analytics, or don’t filter or segment your SurveyMonkey data into subsections of responses - you’ll only being seeing a very superficial view of your visitors.
Four, do something, do anything…
The difference between good websites and bad ones is that someone is generally spending time and effort to continually improve good websites, whereas bad ones are left to stay bad, until some giant overhaul in the far away distance.
An ongoing series of small tweaks and fixes (measured of course, so you know their impact) has a bigger impact on revenues and visitor experience, than a once a decade overhaul!
Now some stuff to help you:
- Jakob Nielsen on 10 High-Profit Redesign Priorities and the Top Ten Mistakes in Web Design
- For those of you who don’t glaze over and twitch when I sat web analytics, check out Context is King Baby: Go Get Your Own at Avinash’s blog
- Getting your Google Analytics right at the outset, from the official Google Analytics blog: How you start is as important as how you finish
- And from Steve at Blackbeats blog: Why KPIs can not come from the bottom up. Management should always participate to the KPI creation and Drilling down into Analytics data
- Making sense of other travel websites’ visitors, courtesy of the compete blog
- Last but definitely not least, Matt Belkin at Omniture has a couple of great posts about knowing thyself and tying your analytics to that. See Don’t Do This! 7 Pitfalls When Deploying Analytics Part 1 and Part 2
This entry was posted on Wednesday, April 9th, 2008 at 2:34 pm and is filed under Online customer behaviour, Web measurement. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.