(Preferred title: What the hell is web analytics?)
Web analytics is Business Intelligence (BI) for your website.
Let’s start with an example. You own a shoe shop, and you want at the end of every month, quarter or year to know stuffs like:
- How many shoes were sold?
- What design/designer sold the most?
- What influenced the buyers?
- Do buyers come back? When buyers come back what do they buy? (Do they buy from the same designer or from shoes displayed where the one they bought the last time were displayed?)
- And if you have multiple store locations – why do people buy more shoes from store A than from store B
And answers to these questions will definitively help you understand your customer’s behaviour and thus you can identify areas of your stores strength and weaknesses (and if you really mean business, mitigate your weaknesses whilst you grow your strengths) and plan your business for maximum exposure and subsequent increase in sales.
Now think of the store as your website! Whether or not you own an ecommerce store!
And buyers? Visitors to your website!
So (a) how do you track visitors to your website? And (b) make meaning of the data collected from your tracking effort(s)?
How to track visitors and their behaviour on your website
Unlike in the physical store described above, where you can designate an intern or a clerk to count visitors to your store and record their purchases or otherwise, it is practically impossible to put up a human clerk to count traffic to your website. But there exist a myriad of tools out there to do your dirty job and these are broadly categorized in two based on the approach to tracking your visitor behaviour. Thus:
(a) Logfile analysis (Raw server logs analysis)
(b) Page tagging
(a) The precursor to modern day web tracking; raw server logs track and records every call to your web server by visitors to your website.
The figures from raw server logs often leaves one under the misguided impression that one is getting a lot of traffic whereas this may not be totally true – this as a single visit to a page is counted multiple times depending on what he page contains – images on the page will each count as additional visits to that page for example.
Whilst this approach does not provide you totally accurate figures, it does provide you a general idea of how your website is performing!
Page tagging based analytics is often offered as a service as opposed to a product and an example of this is: Google’s Analytics (Http://google.com/analytics).
The tracking service makes you complete a form detailing your domain, domain setup (stufss like multiple domain names, sub domains etc.) and some additional information and then provide you a string of code you are expected to add to your website and a dashboard from where you can monitor what is happening and perform all sorts of manipulation to juice out as much sense as possible from that data collected from your website.
Figures from this approach to web tracking are often saner and smaller than obtained from raw server logs but when compared in percentages or rates to those collected via raw server logs, a similar patern of visitor behaviour is obtained.
Some Web analytics Jargon and their popular meanings!
All web tracking tools collect and present data using terms that may be new to you. While some may be intuitive and their meanings easily guessed, others are bland and meanings ‘unguessable’.
Following are some of the important ones:
Unique visitors – this is refers to the number of visits from individual IP addresses in the period for which you are measuring. Most tracking software will log as one unique visit the first visit from an IP address and all other visits as return visits from the same IP. The downside to this is that visits by different (unique) people from inside a corporate network might be recorded as a single unique visit as most corporate networks have one or two public IPs which is shared via a system known as subnetting amongst any number of computers.
It is also possible for you to configure some analytics software to treat as unique a visit from an IP after a certain time lapse.
Return visitors – refers to all additional visits from an IP address after its first (unique) landing on your website. This gives you a sense of people who found value in their first visit and thought to make a return visit.
It is good for this to exceed the number of unique visits. And performing this little arithmetic:
will give you a sense of average returns per visitor (think of that as Visitor Index…now I must rush to patent that before somebody else does J)
Total visits – refers to total visits by all visitors to your website.
Page views – All websites are set up as pages (think about a website as a book, with hyperlinks linking to other pages of the book and to other books – so reaers can jump between pages and books without neccessarliy reading to the end of a page or book). Page views will give you a sense of total number of pages of your website that was viewed by visitors in the period under review. Some tracking tools will let you hone down to see things like average page views (i.e. total page views/total visitors), page views per visit etc.
Average time on your website – average time spent by all visitors to your website. This can be calculated thus:
Total time spent on your website
total visitors to your website
Bounce rate – refers to one page web visits. Think of it as the visitor coming to your website and having looked at the page of interest (often the entry page) bounces of to another website or closes their browser window without clicking through to other pages of your website. There is an argument about this and exit rate, but the general rule of thumb is to keep the bounce rate as well as the exit rate as low as possible and in the event of a marketing effort requiring some actions, place the call for action and if possible the action triggers on the entry page, that way audience will have the opportunity of seeing the call for action and possible trigger the action before bouncing off and exiting your website!
Entry page – is the page through which traffic arrived on your website. This is often not your home page. Yes, most of us want people to arrive on our home page first before taking a trip down the other pages but for several reasons this is often not the case. Go back to the shoe store example, some people will land in your store via the display window and only when triggered by something irresistible will they step into the store via the door offcourse (your homepage perhaps) for a fitting and perhaps a purchase.
Thus, it is important that you keep your other webpages as attractive as possible to retain your target audiences’ trust to stay and perhaps do business with you.
Exit Page – refers to the page via which visitors left your website. This may be due to an increasing loss of interest in your content or an exciting link taking them out of your website to another (rule of thumb is: when linking to a website outside of your own, instruct the link to open in a new window/tab or frame, that ensures your website remains open in the visitors window for additional interaction and actions!).
Traffic sources – This informs you of the source of traffic to your website and some tools will even provide you a chart to help you visualize the traffic sources. Some common traffic sources include: Search engines, organic (people or visitors who know your web address and typed it straight into their browsers), referring websites (other websites where your website has been linked to).
Some web analytics software also allow you see, the operating systems of computers from which your website has been visited and the type of browsers used.
What do you do with information collected?
Read my next post!