Like your food, after a while your data can go off

This article, written by Neil Mason, was originally published on Clickz.com on 16/07/10 and is republished here with permission.

As web reporting and analytics gets more mature as a technique and as an industry, I sometimes now hear comments from organisations along the lines of “We’ve finished our web analytics implementation now and we’re on to the next project”. It’s as if the implementation of web analytics in a business is seen as an event rather than as a process. Whilst the introduction or the upgrading of a web analytics system can be seen as a project, the adoption and the ongoing use of web analytics data in an organisation is something that requires ongoing management and maintenance.

I sometime think it’s like building a house to live in. A great deal of time and effort goes into the initial construction and then you move in. At the beginning everything is wonderful and everything works. The paint is fresh and you enjoy the new environment that you’ve created. But after a while you need to start paying a bit of attention to maintenance, fixing a few things here and there, replacing this or that. Eventually you may decide it’s time for a major overhaul and to redecorate. That process might take a few years but you don’t move in thinking that you’re never going to have to do any work on the house ever again and you may even sensibly think about putting aside some money to cover some of the ongoing maintenance.

It’s the same with your web analytics system. At the beginning there’s a lot of attention and effort and maybe even some excitement of the prospect of the new data that’s going to be coming on stream. Perhaps new processes are developed and implemented to make sure that the right kind of data is being collected and reported in the right kind of way. But once the new system is launched and is up and running what happens then? Do you have a maintenance programme in place? Have you got the right resources available to make sure that the paint doesn’t start to peel and the doors begin to stick?

There are a couple of areas in web analytics where I see from our work with companies that the lack of maintenance begins to cause problems. One areas is campaign tracking and another is content management. I think that campaign tracking is of the areas that is hard to do well in web analytics. There are too many moving parts. You don’t get campaign tracking out of the box with a web analytics system; you have to set it up. Adword integration with Google Analytics makes it easier than most and if all you’re doing is tracking search through GA then that’s great. If you want to track any other channels or you use any other system then there’s effort involved. You may need to decide what type of attribution model you want to use, you will need to think about the structure of your campaign tagging and you will also need to set up a process to make sure that all your marketing links and landing page URLs have the right parameter strings attached to them. There’s also some configuration work to be done in the tool itself to generate the right types of reports for your business. Depending on your scale you may be using a marketing agency to do some of this work for you or you may even be using several.

To keep your campaign tracking at the top of its game, it’s necessary to have strong policies and processes in place and to make sure that those processes are being managed properly. If not then after a while the quality of the data will start to deteriorate. Campaigns will be misclassified or missed out completely. Potentially people will start to use a different structure to the original one that was implemented. The result is that the reporting will become messy and difficult to interpret. Your campaign data will start to go off.

In a similar way, with some sites and some systems there will have been some initial effort involved in making sure that you can report on content properly. Content aggregation is important for heavy content sites so that you can understand the overall patterns of behaviour rather than be constantly working at the detail of the individual page level. In some cases this configuration might be relatively straight forward but often it will require a process to design a content reporting hierarchy and to assign individual pages to relevant content sections. This will have been done when the system was implemented but sites are dynamic, content is continually being created, so an ongoing programme of work is needed to ensure that content is being reported on in the right way.

So, implementing web analytics into an organisation is the beginning of a journey not an event. As part of the implementation process, the right questions need to be asked. How is the maintenance programme is going to be managed? Have we got the money put aside? Have we got the right resources in place to manage the process?

Don’t let your data start to go off!