Site analytics and statistics

Recently, I wrote a post posing the question, Is IE6 dead? I described how critical it is for web designers and clients to understand who is visiting their site. Only by analysing the statistics available for site traffic, or estimating it for new sites, is it possible to decide whether or not to cater for older technology. In this post, I will show you the ‘industry standard’ tool to use and show you what will be useful to you in making decisions about the development of your site.


There’s a range of analytics tools available. The current industry standard is Google Analytics. It is easy to set up and use and provides a fairly comprehensive data set from which information can be gathered.

Google develops constantly the tool to ensure it meets the requirements of web developers and other users. So, there is additional functionality available for ‘advanced users’, making it a flexible working tool for most web sites. This is the tool I use for analysing the traffic to my own websites and Redcentaur’s clients.


I regularly check the statistics for my sites using Google and it is through this analysis that I make decisions about how to develop my sites in the future. It is one of the first things that I ensure is in place when I am asked by a client to improve their sites; it shows me how visitors to the site are able to interact with it and can show where there are serious problems.

I think it might be useful to share some of the key findings from my own site with you to show how I might change my site as a result of these stats.

1. Browser use

The biggest question on people’s minds at the moment seems to be whether or not designers/developers should still spend time on workarounds for Internet Explorer 6. The position I have taken in previous posts is that the statistics for a particular site should tell you whether or not this is necessary: if a sizable proportion of site visitors are still using older browsers (which might be the case for a number of corporate, technological or age-related reasons,) then this should inform any decision about ditching support for these browsers.

My statistics for May 2009 shows some really interesting results. IE as a whole only catered for a total of 18% of visitors, whereas Firefox catered for a whopping 58%. Safari catered for 13% of visitors. I have little doubt that this is broadly related to the types of visitors that my site attracts, particularly this blog, which is suited to design and development / technology-aware individuals who are more likely to upgrade their browsers and to use current technology themselves. This is not likely to be the case for all of my clients’ sites, which is why it is important to relate decisions directly to the findings for the site in question.

Browser share for Redcentaur May 2009

Browser share for May 2009

Delving further into this shows that the majority of Firefox users were using the latest version of the browser (3.0.10); although there was some latent use of earlier versions. Of the Internet Explorer users, 15% of all visitors were using IE 7 and 3% IE 8. None of my visitors were using IE 6 during May 2009. Good news! If this persists, I would consider it reasonable to remove support for IE 6 as a browser from this particular site. I won’t base this decision on one month’s findings though (last month showed 9% of traffic used IE 6).


Browser versions May 2009

Browser versions used on Redcentaur May 2009

2. Screen resolutions

Screen resolution remains an issue, especially with mobile browser technology gaining ground and the likelihood that smaller screen resolutions will be used to access the site. My analytics shows that all my visitors are using large screen resolutions, which suggests I’m not currently getting mobile browsers or people using very small (old) screens. I may wish to look to improving mobile browser experience on my site in the future as this is a growing segment of my market that I appear to be losing.


Screen resolutions chart May 2009

Screen resolutions chart, May 2009

3. Flash versions

If your site uses flash, you can also gather information about the flash versions your visitors are using. This is useful when developing flash on your site. As my site does not currently use flash, this is not important for this site but it would be a useful dataset if I were asked to develop flash site for a client.

Flash versions May 2009

Flash versions used by visitors in May 2009

4. Java support

As Google runs on javascript, it does not have a data set showing how many visitors didn’t have javascript enabled; all visitors identified in Google are javascript enabled. For this reason, it is always useful to compare raw visitor figures against those obtained from your hosting service or another source. Any differences should show how many visitors had javascript disabled in their browsers. This is important if you are planning to implement any javascript functionality that are core to the site — generally, it is better to use javascript for progressive enhancement of design and for back-of-house analytics evaluations.

Google does tell you how many visitors had Java supported on their platform. In my case, 10% of visitors did not have Java supported, which is quite a significant number and might affect any applications I wished to develop in the future using this technology.

Java support May 2009

Java support in May 2009

5. Connection speeds

You might be forgiven for believing that today everyone uses broadband and therefore connection speed is not important. However, for a site that predominantly uses images or multimedia, this is still a factor to consider (and not all areas of the world have high-speed broadband connections). Other factors to bear in mind are that visitors who do have broadband may still have a throttled service, might be visiting on mobile devices using 2G or 3G bandwidth or may not want to use all of their bandwidth allocation downloading a page of your site.

May 2009 data suggests there are still a small percentage of visitors using dialup. Google was not able to identify the connection speed of 35% of visitors to my site, which means a larger proportion of visitors could be disadvantaged by bandwidth munching files or repeated requests to the server for additional files.

Visitor connection speeds May 2009

Visitor connection speeds May 2009

6. Country of origin

Something I always find fascinating is where visitors come from. Unsurprisingly, a large proportion of my site visitors come from the UK, USA and Europe. There are a number of visitors from other countries and this could lead me to think more about the number of conversions that are lost through not supporting native languages. Google provides further analysis on bounce rates (the number of people leaving the site on the first page) and time spent on site, which I have not covered here, and the language settings for the operating system. Together these figures could lead me to improve my site translation or to provide dedicated sites in the key languages.

Countries of origin May 2009

Countries of origin May 2009

7. Sources of traffic

Analytics can aid understanding from where people have landed on your site. This provides a useful analysis for the value of ad campaigns, search results, link value, etc. My data shows that I have had a number of hits from Google organic searches (from search requests), from direct hits (typing the address into the browser), from social media, such as Twitter and LinkedIn and from other referrals, blogs and listings. This information is more useful when assessed in conjunction with other data to identify any parts of a site that visitors are landing on and bouncing from. Improvements should be targeted at these areas.

Traffic sources May 2009

Traffic sources May 2009


A vast array of information can be obtained from simple data. This post just touches the surface of what you can gather about the usability of your site from your visitors. Check out Google Analytics for yourself and see what surprising information it can tell you.

Is IE6 dead?

.net campaign to bring down IE 6

.net campaign to "Bring Down IE 6"

There has been much debate in recent weeks about the death knells of Internet Explorer 6; particularly since .net magazine ran an article in its February 2009 issue ( ), which has been supported by some designers.

IE6 is well known for its quirky interpretation of web standards and for requiring significant code hacks to enable web sites to appear as they should. I remember in the Dark Ages of the internet, when Microsoft and Netscape were battling for browser supremacy, that designers started putting “This web site is best viewed in Internet Explorer [or Netscape Navigator]” messages on their sites. At the time there were only a fraction of the people online compared to today, and the internet was not a commercial entity in the way it is today. The decision by Microsoft to gallop headlong into the distance when everyone else was going in another direction has led many designers to pull out what little hair they had left and throw their arms up in despair for many years. I, like most web designers, keenly await the day we can finally bury IE6 for good. However, frustrating and annoying as IE6 is for designers, a word of caution needs to be sounded before ditching the browser from design considerations.

It is important for designers to remember that they do not make browsers; they make web sites for visitors to view in the browser of their choice: a browser is a user-defined tool over which the designer has no control. In the same way, a designer has no control over whether or not JavaScript is enabled on the computer a visitor is using. To say, therefore, that you are not considering IE6 in the design of your client’s site is effectively saying that you are willing to turn a proportion of your client’s customers away.

It is indeed fortunate that more recent browsers have been more closely aligned to web standards and the trend appears to be towards meeting the standards rather than creating their own. Even Microsoft appears to have seen the light with IE7 and 8. So, the question is whether or not IE6 is obsolete now that there are alternatives and updates to preferred browsers available.

On the face of it, this is a compelling argument as IE6 users have diminished in number, seemingly moving to IE7, IE8, or one of the other browsers, such as Firefox, Opera, etc. However, there is a core of users who still rigidly seem to be sticking with IE6 and it might be worth considering who these people are.

Traditionally, they are thought to be dullards who can’t or won’t update their browsers and who need to be forced to do so. Looking at the statistics available from W3Schools, there is a sharp move away from IE6 over the last twelve months: in April 2008, over 29% of visitors used the browser; in April 2009 this figure was 15%. This downward trend can be rationalised by the increases in Firefox, Safari, Opera, Chrome and IE8.

Table of browser usage for 2008 - 2009 from W3C Schools

Browser usage 2008/2009 by W3C Schools

Market Share shows a similar tale. In June 2008, IE6 controlled approximately 26% of the traffic. By April 2009, the proportion dropped to 18%. Again, this trend can be attributed to the massive growth in popularity of Firefox, the growth of Safari and the release of IE8.

Browser trends graph and table by Market Share

Browser trends by Market Share

While these figures are often disputed because they rely on site traffic to the sites that publish the results, what cannot be denied is the continuing popularity of the old browser, IE6 still commands about 15 – 20% of the market share.

But why might this be the case? Are up to 20% of internet users dullards? Of course not. Many of the continuing users of IE6 are large corporate enterprises who deliberately wait before upgrading, to make sure that the latest versions of operating systems and software applications (especially Microsoft) work and are reliable before upgrading themselves. Ending consideration of IE6 at this stage could alienate a large proportion of sizeable corporate clients.

Designers should be led by what the users of their clients’ web sites need. If the client has a large number of corporate customers, chances are that some of them will still be using IE6. Similar chances exist for a large proportion of general users, who are not necessarily technologically clued-up, or might not know how to upgrade (or change) their browser.

It is not the designer’s responsibility to dictate who can view their clients’ web sites. That is a decision that rightly can only be made by the client. However, the designer’s responsibility is to help their clients to reach an informed decision by explaining the problems of IE6, identifying the relevant user statistics for the site (or general stats for new sites), identifying trends over time and demonstrating the pitfalls of not including IE6 as a supported browser.

Taking the statistics from my own site, IE6 accounts for just over 20% of visits in the last year to April. Amazingly, there were even some users still clinging on to IE5.5! However, over the last six months, the trend has completely changed. No-one now is using IE5.5 and IE6 has fallen to 9% of visitors. Firefox has actually overtaken IE as the browser of choice with over 45% market share, compared to IE’s market share of 37%.

Redcentaur's own user analysis

Share of visitors to Redcentaur using IE 6

What this shows me is that if I suddenly switched off support for IE6 now, I would alienate 9% of my visitors. This is something I am not prepared to do as that is equivalent to saying that 9% of my customers are not valued.

The key is for the designer not to unnecessarily alienate a proportion of potential customers for their clients. If it is clear from research specific to the site being designed, that a proportion of visitors are using earlier versions of browsers, then as a designer it is your responsibility to include them in the design of the site.

Clearly, this will require a discussion with your client, where the pros and cons of including IE6 and other browsers are identified. This will also involve a discussion about the additional cost of hacking the site for IE6, which should be costs included in the estimate you provide your client. It is your clients’ decision whether or not to include IE6 within the design – afterall, it is their site, they are their customers and it is their risk.

Personally, I would prefer not to start seeing sites saying “This site is best viewed in Firefox 3.2”. It is really not a professional image to portray for your design or for your clients’ corporate web site.