Mozilla Marketing

The Mozilla.com website: more than just pretty illustrations

By DR ⋅ August 18, 2008 ⋅  Print This Post ⋅ Post a comment

John Slater has written a really good post explaining the genesis of the new Mozilla.com website design, which went live with the launch of Firefox 3.  In general, the feedback from the community has been really positive — it’s a great looking site which does an excellent job of communicating the benefits of Firefox and giving visitors a feel for Mozilla.

But how has the new site performed when looking at some key performance metrics? For example, are visitors to Mozilla.com more or less likely to download Firefox since the relaunch after seeing all of the balloons, eggs, and yetis?  I took a look at the performance of the site in the six weeks prior to launch, and then the six weeks after launch (note: this is for the US version of the website) for a variety of metrics.  The first is conversion rate — the likelihood that a user visiting Mozilla.com or Firefox.com ends up downloading Firefox.

Though the conversion rate bump leveled off in late July, we’re still in positive territory compared with the older site for the Mozilla homepage.

What this conversion rate doesn’t capture, however, are the incremental conversions that have been captured from the product pages (specifically the customize, security, and organic pages) that were added to the new site and also included a download button.  Adding download buttons to multiple product pages makes a lot of sense.  It gives users the opportunity to download without having to find the one or two pages on the site that offer a download button (like the pre-Firefox 3 site).  Thousands of people have downloaded from these pages after reading more about Firefox security, customization, and why we consider it to be “organic software”.

Beyond just downloading, the new Mozilla.com is loaded with info about Firefox features, tutorials, and information about the organization.  More time spent shouldn’t come at the expense of a user being able to easily download.  But the increase in conversion rate coupled with an increase of time spent on the site should be viewed as a positive.  Users who spent less than 15 seconds browsing on Mozilla.com decreased after the website relaunch — and those that spent more than 15 seconds increased.  This was true for all segments above 15 seconds — those that spent between 30-60 seconds, 1-3 minutes, 3-5 minutes, etc — improved after the website relaunch.

Another good indicator to analyze for the pre and post launch performance of Mozilla.com is to take a look at bounce rate, which is a measure of how many visitors arrive at a certain page and then immediately leave the website (by ending their session, typing in a new url, etc.).  In other words, how likely they are to “bounce”.  The lower the number here, the better — meaning that users are finding relevant information before they exit the site. With the launch of Firefox 3, there’s a huge drop in bounce rate — this makes sense since people were arriving at Firefox.com and then downloading with a high frequency (which sends people to another page). However, even though this metric is creeping up, it’s still lower than pre-launch levels — another good sign that people are hitting an entry page and taking action on the site (whether that means viewing more pages or downloading).

There is certainly room for improvement in all of these areas — even a very small percent change can make a very positive difference.  We’ll continue to test and refine these pages over time with an eye towards improving these metrics.