Faulty April Market Share Statistics

After reading the statistics provided by Net Applications for the month of April 2008, I was really disappointed at Mozilla’s Firefox performance. I tried to find a logical reason for the dramatic drop, but couldn’t find any. I searched all over the Internet to find a clue, but no joy. Something really strange had happened to the momentum gained by Firefox. Those numbers were all wrong.

Today, I found out that Net Applications had a mistake in their figures. According to Computerworld Networking and Internet, “An ‘extremely large’ marketing campaign among a small number of sites that targeted only users of IE was responsible for the aberrant numbers, said Vince Vizzaccaro, the executive vice president of marketing at Net Applications.” Vizzaccaro would not name the company that owns the sites that ran the IE-only marketing campaign, but did say that it was “totally unrelated to anything with Vista, Microsoft or Windows.”

Vizzaccaro said that Net Applications had been taking a second look at the data when it was contacted by Mozilla, the company that develops the open-source Firefox. “We had a call from Mozilla, and working with them is what led us down this path to identifying that it was a very large traffic increase that was very specific to a group of sites,” he said.

When the Net Applications statistics were pulled out of the oven there were a number of red flags thrown up questioning the integrity of the data. Mozilla, confident that their internal numbers did not show such a significant drop, started to investigate the findings. Net Applications cooperated, and here are some of the things they found:

1. Usage of Firefox did not appear to decline in April.

2. Usage of IE saw an ultra unusual spike over the period of a few days in April. While there was steady traffic from IE users – as measured by page views – for a short period surrounding April 18th, there was a huge spike in traffic from IE users – as measure by unique visitors – during the same time period. When I say “huge spike”, I mean something on order of 25% to 50% greater than what could have been expected.

3. Most of IE’s spike was attributable to IE6 users on XP (with some assistance from IE7 users on Vista), and nearly the entire spike came from users outside North America.

After this stumble, Net Applications have now put additional filters in place to watch for abnormal behavior like this. According to them they were already watching for stuff like this to happen on any one particular site, but they weren’t expecting such a widespread campaign to occur.

The company, Net Applications of Aliso Viejo, Calif., has posted revised data on its Web site that shows Firefox and Safari still dropping in share, but by much smaller increments. The stats have been updated, and it looks like things have returned to normal. Firefox and Safari both dropped slightly in market share, but only by very modest amounts this time. Internet Explorer only saw a 0.03 percent increase instead of the 1.22 percent that is was at with the incorrect data.

Here are the revised statistics provided by Net Applications after their thorough investigation:

(Expressed in Percentages)

Web Browsers April March Difference
Internet Explorer 74.83 74.80 0.03
Firefox 17.76 17.83 0.07
Safari 05.81 05.82 0.01
Opera 00.69 00.69 0.00
Netscape 00.56 00.55 0.01
Others 00.35 00.31 0.04
T O T A L: 100.00 100.00 0

Besides IE’s modest gain and Firefox tiny blip, the rest of the players kept their positions. I’m glad Net Applications got its act together and came up with these revised figures. It was difficult for me to accept Firefox’s dramatic drop of 0.87 percent incorrectly shown in its original statistics.

Now you really know what happened in April in the Web Browsers arena. I hereby apologize for all inconveniences caused by inadvertently giving you wrong figures.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s