Community Driven Statistics for Download Day

Thursday, June 5th, 2008

Hats off to community member Ehsan Akhgari for his awesome analysis of our Download Day pledge data (it far exceeds my feeble attempt).  Not only is Ehsan’s analysis updated in real-time, but he’s also made the source data available for other community members to work with and build on his ideas.  Great job!

Filed in process | Permalink | Comments (4)

Firefox Adoption Grows at Breakneck Pace

Thursday, June 5th, 2008

John Lilly highlighted a milestone a few months back when we reached 50 million active daily Firefox users.  Now, just more than four months later, we’ve surpassed 60 million active daily users.  Wow!

What’s perhaps most shocking is that the percentage growth of Firefox usage has continued, all the while our base number has become substantially bigger.  In other words, when Firefox had 10 million active daily users, growing at 50% per seven or eight month period meant that we had 5 million new daily users join our community.  Now, to maintain that percentage pace, we’re seeing nearly 20 million new daily users join our community over a roughly similar time period (e.g., 50% recent growth as we went from 40 million to 60 million active daily users).

This remarkable pace of adoption and community growth stands in stark contrast to some outside expectations from just a few years ago.  Here and here you’ll notice that folks were writing about a supposed decline in Firefox growth back in 2005.  Clearly, those predictions have not played out; far from it, an opposite scenario has become reality.

Thanks to our contributors, volunteers and community members around the globe… they certainly deserve the credit for Firefox’s amazing success.

Filed in results | Permalink | Comments (0)

Go Colombia!

Monday, June 2nd, 2008

Since Download Day was announced (just a few days ago), it’s been fun to watch in real-time the way Firefox community members around the globe have embraced it.  We’ve already seen pledges from community members across more than 200 countries (or regions)!  In addition, I thought it’d be interesting to see how the early number of pledges matches with Firefox adoption.  In other words, have certain regions been ultra avid in the way they’ve embraced Download Day, relative to the size of their region’s Firefox user base?

Below is a chart showing the number of pledges by region (as of earlier today), along with the associated percentage out of total pledges (# of pledges from region x/total pledges across all regions).  The last column then shows us what percentage of Firefox users hails from each country.

As one example, you’ll notice that a country such as Poland has been particularly excited about Download Day.  While about 4% of all Firefox users are in Poland, nearly 8% of all Download Day pledges can be attributed to community members in Poland.

A few other things to highlight thus far:

  • The international Firefox community has been amazingly active, with nearly 85% of pledges attributable to members outside the U.S.
  • In addition to Poland, Brazil, Japan, the Netherlands, Russia, Mexico, Argentina, Taiwan, Chile and Portugal are among the most active regions (among many, many others)
  • Colombia made the list above.  Why mention this specifically?  Colombia has also shown the greatest increase in Firefox usage, on a percentage basis, since the beginning of 2008.  Usage growth by country is an entirely different topic, but this one nugget of info seems relevant in calling out our community members in Colombia.
Filed in results | Permalink | Comments (2)

Firefox Market Share Update and a Perspective on Firefox 3

Monday, June 2nd, 2008

After last month’s hiccup, yesterday Net Applications released its latest market share report for May 2008.  Firefox showed a nice improvement in the April to May timeframe, going from 17.76% to 18.41% worldwide market share.  Net Applications provides their own interpretation here.

The interesting part of the story is that while Firefox 3 continues to show rapid adoption, most of our growth was driven by increased usage of Firefox 2.  In other words, Firefox 3 adoption is not coming at the expense of Firefox 2 usage (it’s not simply the case of existing users updating from Fx2 to Fx3).  Clearly, the overall health of the Firefox brand is amazingly strong as we head towards the official release of Firefox 3.

The first chart below details the change in Firefox market share numbers over the last five weeks, and the second chart highlights the latter two columns:  Firefox 3 adoption relative to the share of Opera (I didn’t mean to put Opera on the spot here… it just seemed like a relatively good benchmark).  Source for all data: Net Applications.

Filed in results | Permalink | Comments (0)

The Firefox Community - International Participation

Wednesday, May 28th, 2008

We’ve talked a lot here lately about the proliferation of Firefox usage around the world and about the early adoption of Firefox 3.  Today, something combining both of these subjects was announced – Download Day.

The coolest thing to watch is the interactive map.  You can see the number of pledges grow by the minute, and you can visually see the extreme international scope of our community.  Check it out!

Filed in plans | Permalink | Comments (4)

The Psychology of Downloading Firefox

Wednesday, May 21st, 2008

What attributes of a browser are most important to a prospective user when he/she is considering downloading a new browser?

This is a question we’ve set out to answer.  For marketing purposes, for example, this type of knowledge could be tremendously useful for Mozilla.  One of the key learnings that any student takes away from business school is the power of conjoint analysis, which can precisely answer our question.  However, such market research can be time consuming, expensive and often quite complex.  Fortunately, our question is much simpler than the one facing most consumer marketers for two reasons:

  1. Typically with conjoint analysis, each attribute can be broken down into different levels.  As the Wikipedia example points out, the screen format for TVs includes the levels of CRT, plasma, LCD, etc.  But in the case of web browsers, a browser generally either has a feature or it doesn’t, e.g., there aren’t different levels of tabbed browsing as a feature.
  2. The decision being made by a new, prospective browser user is fairly binary in nature.  In other words, it’s either “yes, I’ll download” or “no, I don’t want to download”.  On the other hand, typically with conjoint analysis, a marketer has to design several prototypes of a product and allow the prospective buyer to assign some dollar amount to each feature-level combination (i.e., the results are along a relatively complex spectrum).

Given the situation outlined above, and given that we frequently look for marketing solutions beyond the traditional (e.g., not “let’s hire a market research firm and let them figure it out”), David Rolnitzky and I had the idea of using search marketing to see if we could find the answers to our original question.

In other words, given a control ad variation, what if we swap out a few words of an existing advertisement with a highlight of a new feature in Firefox 3?  For example, a text ad within AdWords includes four lines (text and URLs).  If our existing (control) ad variation has a line stating, “More Customizable!”, perhaps we could rotate other phrases in its place, e.g., “Now with the Awesome Bar!” or “Now with Improved Support!”

If the results show that simply changing one line within an advertisement has a big impact on whether or not prospective new users click on the ad, then we’ll presumably have an idea as to which attributes of Firefox are most important to the user in their download decision process.  Perhaps there are some flaws in our thinking here or in our methodologies, but this exercise seems worth trying given both our situation and the question we’re attempting to answer.

We’re currently in the very early stages of this experiment, so please let us know if you have any suggestions for what Firefox features/attributes (or associated ad phrases) we should be considering.

Filed in plans, process | Permalink | Comments (3)

Life after Launch of Firefox 3

Tuesday, May 20th, 2008

As we approach the official release of Firefox 3, I thought it’d be an interesting thought exercise to start considering what life will look like shortly after that special day.  In other words, where will the Mozilla community see itself one month after launch, e.g., what will the adoption picture look like?

One approach would be to devise a statistical model to predict the future of Firefox 3.  Another, more intuitive approach is to simply layout the events surrounding the launch of Fx2 (October ‘06) and let the readers of blog of metrics take some educated guesses.  In the charts below, Firefox 1.5 is the blue line and Firefox 2 is the red line.

Here’s my own fuzzy match… In total, we saw about 22 million cumulative downloads of Fx2 in its first month after release.  That number translated to about 8 million active daily users one month later.  The Mozilla community is now more than twice as large as it was back then (e.g., in terms of # of users), so perhaps we can estimate that we could see 50 million cumulative downloads of Fx3 in its first month after launch, and that this number could translate to about 20 million daily active users of Fx3 within that first month.

Your thoughts?

Filed in plans | Permalink | Comments (2)

What www.mozilla.com tells us about our global community

Monday, May 19th, 2008

We’ve known for a long time that several Firefox product pages (e.g., download.html) are closely tied with the Firefox user experience.  What we didn’t previously know is exactly how closely our community’s interaction with www.mozilla.com speaks to Firefox’s global reach.  As recently mentioned, we’ll soon have a much more structured dialogue around the international impact of Firefox, but before we get to that, we wanted to point out another somewhat surprising fact:

International visits to Mozilla.com come very close to approximating actual global usage of Firefox

Even though the two metrics are measured by two separate systems using two different geoip technologies, and even though the product page experience can differ by Firefox build and locale, and even though we have several separate international web properties (e.g., mozilla-europe.org, mozilla-japan.org, mozillaonline.com) that we’re not including here – after considering all of these things, when looking at a breakdown of visitors to www.mozilla.com we see a pattern that almost exactly mirrors actual global usage of Firefox.

Similar to what Schrep previously highlighted, our analysis is below.  Usage of Firefox by country is in the second column and a breakdown of visitors to www.mozilla.com is in the third column (both are % of total).  These numbers are from the current month of May (thus far).

So, what’s the main takeaway here?  Web analytics is critical for Mozilla.  The ability to take a partial temperature of the Firefox user experience on a per country basis and help fix any pain points for our users should be a big win for the Firefox community.

Filed in plans, results | Permalink | Comments (1)

Hello Madrid!

Thursday, May 15th, 2008

We’ll eventually be talking a lot more about this, but for today, I wanted to present one slice of data that got me excited.  Mozilla recently implemented Omniture as our web analytics solution for www.mozilla.com and we’re just starting to poke around in understanding how Firefox users interact with our web site.   As an example, the report below shows the top 10 cities with visitors to mozilla.com so far in the month of May:

A few things that stand out:

  • Manila is #4!
  • There’s only one U.S. city in the top 10
  • Our web traffic has a verrrry long tail.  The top 10 cities represent about 10% of our web traffic and the top 1,000 cities represents 83% of mozilla.com visitors.
Filed in process | Permalink | Comments (11)

Fluctuations in International Internet Usage

Tuesday, May 13th, 2008

In thinking about broad questions such as “how do people use the internet?”, one very simple and curious problem solving approach would be considering how usage of the internet fluctuates (e.g., over time, by country, by web site, by segment of users, etc.).  Whether the cause is something as simple as seasonal patterns (e.g., perhaps usage declines during summer months) or as complex as genuine shifts in users’ behavior, an understanding here seems like a very basic and critical need for the technology arena and business world.

For example, one pattern we’ve noticed here at Mozilla is that the overall usage of Firefox seems to decline a bit during this time of year.  In 2007, for example, the number of average daily users (ADU) declined by about 2% from April to May.  Even though we’re currently less than halfway through the month of May, it looks like a similar pattern will likely hold in 2008.

So, the first question this raises is: is this pattern in Firefox usage representative of the entire internet population?  The answer seems to be “yes.”  Looking at comScore’s data (subscription required), the number of worldwide average daily visitors dropped during April and May 2007 (relative to its level that March), jumped back up in June and then declined again in July and August 2007.  This pattern tracked extremely closely with month-to-month Firefox usage.

The next question this raises is: “why?” (and why aren’t folks talking more about this?).  In short, the answer to the former seems to be holidays.  Much of the world celebrates some version of May Day or Labour Day on May 1st.  Moreover, countries like China and Japan enjoy “golden week” holidays.  While it may seem hard to understand how a single holiday can have such a macro level effect, consider this: all else constant, if overall internet usage drops by 30% within just a single day, that month will show a 1% decline in average daily users.

As cool as it is to uncover this pattern and its potential causes, the question I raise in parentheses above seems like the most interesting part of this discussion.  And it starts to touch on the recent discussion by John Lilly and Mike Arrington (and John’s follow-up here).  Many cohorts (curious users, university researchers, small businesses, etc.) would certainly be interested in the opening up of all sorts of data… indeed, much of the reason folks aren’t talking more about questions such as “how do people use the web?” primarily relates to information being locked up.

Filed in plans, results | Permalink | Comments (2)