Before I sat down to write this short blog I looked into Amazon.co.uk to check how many books there were on the subject of web analytics, which is 56 (including Kindle versions) if you really want to know. Google alone returns just over 200 million mentions of the topic so if you want to know everything about web analytics you might as well resign yourself to spending several years of your life on the subject, with the likelihood you may have to start all over again once you have finished just to catch up with new developments.
With a such monumental amount of data to research it would be impossible for me to sum up every aspect of this important subject with a blog, so I will instead attempt to provide potential web analytics users with some common sense suggestions in the form of a quick ‘to do’ list.
Find the right tool
Google analytics is the most obvious choice; it must be a reliable service, not just because it comes from Google but since it seems that 50% of all web masters in the world are currently using it. But Google analytics isn’t the only service available. Yahoo offers a similar tool, with a few more in-depth analytics options, too. Other options are also available – some free, some paid. Quite why you would wish to pay for web analytics when Google offer a free service is beyond me, unless of course you want to go for more sophisticated approaches like comparing your site with that of the competition (e.g. compete.com). Don’t dismiss, however, the possibility that in future Google may indeed offer a two-tier service – a basic free one, and a higher level (and paid for) one, though this may be some time away.
Set it up correctly
This may sound obvious but you will need to have access to the source code of your site and insert the appropriate snippet of information for it to work. If you don’t know how to do it, or don’t even have access to the back end of your web site, the rest of this blog is academic and you may as well make yourself a cuppa instead.
Plan and be data selective
Once it’s all set up you need to collect the data. This is where the fun starts. You will see that in most cases the amount of available data is vast. In all instances you will only ever need a small proportion of it. For example, unless you are a software developer you would rarely need to know what version of a browser your visitors are using. Setting up the right data can be a complex exercise and you really need to know what you are doing. It pays to pay an expert to set it up correctly so that you can really get at the data your company needs, rather than what may look pretty on a spreadsheet.
Use all the other free tools
If you use Google Analytics you will have access to a plethora of other free web tools. For example you can use google.com/webmasters to check that your site is functioning correctly, with no crawl errors and such likes. Here things can become a little technical and unless you are the geek in your company, or you have plenty of spare time in your hands, you should also leave this to the experts. More experienced users can even set up experiments by publishing test pages that are then checked by Google, reporting back on which best case scenario you may wish to go for. This is a particularly good approach for e-commerce sites or where there is high competition in terms of products (and therefore key words and search terms).
Don’t limit yourself to basic analytics
So you have seen how many visitors you are getting in a period, where they land, how long they stay on a specific page, where they are coming from and where they go from there, and so on. This is all good stuff as one of the points of analytics is to improve visitors’ experience – if you see a decline, or a peak, you can take appropriate action, as well as evaluating the extent of your campaigns. But what if you want to get more useful data? There are now commercial packages on the market that consolidate a number of perfectly legitimate sources to offer you a proper interpretation of your visitors. Instead of just getting a figure, you can actually see which specific company has landed on your site and when (and sometimes even more). If you are prepared to pay a monthly fee for this kind of data through something like Leadforensic, returns on your investment can be great indeed and a well-planned system of this kind will normally pay for itself in a couple of months.
Finally, the old 90:10 rule applies to web analytics. That is, while it may take 10% of your time to set up the system, 90% will go on evaluating the information. As time is money you may be better off seeking the help of experts. At least instead of receiving a truck load of unsorted hardcore, you will end up with some nice and usable building blocks.