|
|
|
|
|
|
| Mozilla/Firefox takes advantage of the flaw of Internet Explorer |
|
| Internet Explorer has lost market share recently in Europe and has just experienced the repercussions of the security flaw, which was announced on the 15th January. In Germany, and to a lesser extent in France, its challenger Mozilla/FIrefox benefited from this to increase its visits share. |
| Cross section: |
- Study carried out from the 4th to the 22nd January 2010
- Cross section of 14,215 websites
|
|
| Estimated loss of 2.7 points for Internet Explorer in Germany: Mozilla/Firefox takes advantage of Internet Explorer’s flaw |
In our previous browser barometer from December 2009, we found that the decline of Internet Explorer has been confirmed in Europe for several months. Internet Explorer’s visit share has decreased by 3.6 points over the past 6 months and fell to 58.6%.
The bad luck seems to continue for the leader. On 15th January 2010, the German government and the Centre d’Expertise gouvernemental de Réponse et de Traitement des Attaques informatiques (government’s center of expertise for the response and processing of computer attacks) warned against using Microsoft's Internet Explorer to browse the web because of security flaws and recommended the use of alternative browsers until a corrected version was available from Microsoft.
Has this news had a direct impact on the browser market?
The goal of this analysis is to find out if the news about the flaw and the recommendation of the use of alternative browsers in particular, has had an impact on Internet Explorer’s visit share for websites in Germany and France.
Internet Explorer’s loss of market share observed on websites in Europe for several months has been confirmed in Germany and in France.
We chose to pay attention to week days for both countries (the distribution of visits, by browser, differs between week days and weekend days due to professional usage and private usage). This is to analyze the impact of Internet Explorer’s security flaw, without being affected by the usual week/weekend fluctuations.
We made a hypothesis of short term quasi-stability (the two weeks prior to the news and the week after) of this observed decline. Linear regression was performed on the data that was produced two weeks prior to the news by forecasting what would happen the week after the news (from Monday 18th to Friday 22nd January 2010). This data allows us to estimate Internet Explorer’s share and does not include the factors which might affect it.
Our test results show a significant decrease of Internet Explorer’s visit share in Germany and in France, in addition to the observed trend before the security flaw news.
For German websites, the decrease, which is directly associated with the news of Internet Explorer’s security flaw (additional loss outside the observed loss over several months), was estimated at 2.7 points of visit shares from Monday 18th to Friday 22nd January 2010:
- For the studied period (weeks from 4th to 8th January 2010 and from 11th to 15th January 2010), the average gap for every day, between the real share of visits recorded by Internet Explorer and the estimated share by the model, was at 0.5 points.
- Every week day from 18th to 22nd January 2010, the real share recorded, on average, was at least 2.3 points below the estimated share. On Friday 22nd January 2010, this gap climbed to 3.1 points.
In this same week from 18th to 22nd January 2010, we estimated the gain of Mozilla/Firefox’ visit share to be 2.2 points.
|
|
| Less significant loss of 0.7 points estimated in France, Mozilla/Firefox also profits from this |
As far as French websites are concerned, the decrease, which is directly linked to the news of Internet Explorer’s security flaw (additional loss outside the observed loss over several months), was estimated at 0.7 points of visit shares from Monday 18th to Friday 22nd January 2010:
- For the studied period (weeks from 4th to 8th January 2010 and from 11th to 15th January 2010), the average gap for every day, between the real share of visits recorded by Internet Explorer and the estimated share by the model, was at 0.1 points.
- Every week day from 18th to 22nd January 2010, the real share recorded on average was at least 0.6 points below the estimated share. This higher gap was observed on Thursday 21st January 2010 with 0.9 points.
- Note: in the long-term Internet Explorer‘s visit share in France is experiencing a downward trend, but remained stable on the weeks from 4th to 8th January 2010 and from 11th to 15th January 2010.
In this same week from 18th to 22nd January 2010, we estimated the gain of Mozilla/Firefox’ visit share to be 0.7 points.
The current browser leader is undoubtedly in trouble. With a decrease of its visit share for quite some time, the news of the security flaw has confirmed this trend.
Is this a momentary increase of its loss in market share (a corrected version is available) or is there a change in behaviour? Will an increasing number of European Internet users use alternative browsers?
Stay tuned as AT Internet Institute will provide a new browser barometer soon.
|
Methodology:
We are interested in Internet Explorer’s visit share, which have been recorded by a certain cross-section of websites (except mobile websites).
In order to establish the cross-section of websites to be used as part of a study, a website is only used if the selection criteria, which have been established previously, are respected. These criteria are very strict and their aim is to remove any websites which have an anecdotal audience, or a chaotic audience, or whose collected data suspects a dysfunction in the measure.
For Germany and France, only websites whose traffic has been mainly generated within the country itself are considered. In addition to this, only visits to these websites which have been generated within the country itself are considered. We calculate average traffic distribution per website. For every website, the visit share for Internet Explorer corresponds to the total number of visits of Internet Explorer compared to the overall number of visits for all browsers combined over a given period.
|
|
|
|
|
|
|
|