Analytical web analytics

This article, written by Neil Mason, was originally published on Clickz.com on 14/01/10 and is republished here with permission.

In my last column I reflected on 10 years in digital analytics and how far the industry had developed in decade in some ways and how there was still room to grow in others. I commented that I thought that one of the issues was that the online marketing world had been “data rich and analytically poor” and this week I want to explore some of the areas where I think there is work to be done to enhance the quality of insight that digital marketers get from their investments in data capture and reporting technologies.

For example, considering that web analytics and campaign data is often used to make resource and budget allocation decisions around online marketing spend, the continued use of the “last click attribution model” seems to me to be bizarre. I recognise that many organisations are seemingly locked into this approach through the measurement and reporting technologies that they use but it seems to me that this is an area where little “analytical” progress has been made over the years. Many businesses and their agencies know that this model is a sub-optimal solution to managing marketing spend but have limited options given the systems available to them. Some work has been done by some agencies to improve the way that campaigns are measured and optimised but it’s seemingly required large investments to get the position where the influence and impact of different channels can be tracked and measured properly. What I would like to see is the web analytics providers providing greater analytical horsepower in the area of campaign analysis by providing more flexibility in defining attribution models and windows and also the ability to look at the impact on multiple campaigns on conversion.

Another area where it would be great to see some more progress is in the development and application of econometric and predictive techniques to understand online marketing effectiveness and particularly the interaction between online and offline activity. These sort of techniques have been used for years in brand and consumer marketing to assess the impact of TV campaigns and the like and there is probably the role for these types of analytical approaches to supplement or enhance the direct measurement techniques that are most often used in the digital space. Again, some work is being done in this area by some businesses or their agencies but i feel we see to see more debate about the use of these techniques in the online space and the type of problems they can help to solve.

This leads me on to my final point and that is I would hope to see the development of more “science” in online marketing. For me marketing as blend of art and science, a combination of left-brained and right brained approaches. That’s not to say that there isn’t any science in online marketing analytics. Multi-variate testing is a good example where analytical and statistical techniques are used to predict the best probable outcome from a series of experiments. There are other examples as well in behavioural targeting. However I think that there are many other areas where there are the opportunities for a more scientific and statistical approach to understanding online behaviour in the same way that marketing scientists developed theories and models to explain such things as TV viewing behaviour, shopping behaviour and purchasing behaviour 20 to 30 years ago. One of the earliest marketing texts sitting on my bookshelf is called “Marketing Decision Making: A Model-Building Approach” and it was published in 1983! For this to happen I guess there will have to be a greater interaction between academia and commerce. I’m sure there is some really interesting research in this area being done in instructions around the world but I’m not sure about how much of it is reaching the commercial world in a consumable format. One good initiative in this area is that the Web Analytics Association has secured access to 5 leading marketing journals for its members to provide an opportunity for industry analysts to connect with the work being done by marketing academics and researchers. It’s one step along what I hope will be the road to making web analytics more analytical.