Digital analytics

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

Last week I was over in San Jose for the Emetrics Marketing Optimisation Summit for three days of input and output on all issues around digital analytics. As usual there was a wide variety of content to absorb and a breadth of issues to discuss. As I sit and write this at the airport on my way home (got to meet those ClickZ deadlines…) I get a chance to reflect on some of the key themes that came out of the conference.

This was my fifteenth Emetrics conference in the last six years on both sides of the Atlantic, so I have had the chance to see how the event has grown and developed over the years. Whilst some of the fundamentals remain the same (such as how to get analytics embedded into an organisation), the debate has moved on. I attended sessions covering mobile analytics, social media measurement, voice of the customer programmes, marketing mix modelling and data integration.

One of the sessions I found most interesting was from Joe Megibow from Expedia. Megibow outlined how they have invested huge amounts of time and effort into integrating their various sources of customer experience feedback, specifically OpinionLab and TeaLeaf. This gives them the ability to link comments left by users on the site with their actual sessions, so they can see the comment and then replay the user’s session to more fully understand the user’s experience that led to the comment. The integration is interesting but what was impressive was the detailed analysis that went into understanding and solving very specific customer experience issues. Megibow certainly washed some of his dirty laundry in public by taking us through a number of examples of how they had uncovered problems with the site through the voice of the customer, the impact that the problems were having on the customer experience and the steps they had then taken to solve them. What struck me was the fact that often these were very small specific problems affecting perhaps a relatively small number of people but when you added them all together they were having a significant impact on the user experience. In fact Megibow said that since they had embarked on this systematic programme that conversion rates were being steadily improved but that the major success had been “winning the cultural shift of listening to customers, institutionalising analytics in the business and executing against the outcomes”.

Putting customers at the centre of the company’s strategy was a theme taken up by Greg Dowling from Nokia as well. He took us through the challenges of developing and implementing a global measurement strategy for a business that is looking to develop consumer data as a strategic asset. These challenges that had needed to be addressed included the lack of a common language around metrics, the fragmentation and quality of the data, the lack of competencies in certain areas and the fact that data and insights were not part of the business processes. Anybody who works in a large global organisation will probably recognise one or more of these challenges. Over a two year period Dowling and his team had worked to address the challenges and is his words “behavioural data is at the heart of our relationship with customers”.

One of the tough parts of implementing the Nokia measurement strategy had been around getting the mobile analytics strategy sorted. This was taken up in more detail in a further session by Dowling and Gary Angel from Semphonic. I touched on the issues around mobile analytics in my last column but in this session I really discovered how tough it is at the moment to get decent data on user behaviour on mobile devices, let along integrate that with the same users’ behaviour on the fixed web. All the issues that impact our ability to measure on the fixed web 10 years or so ago plague the mobile web such as the data collection methodologies and browser standards. There are challenges around visitor identification and measuring mobile applications usage. However given that it’s expected that by 2012 more mobile phones in the world will be accessing the internet than PCs it’s clearly an area that organisations need to start to address and given the evidence I saw in San Jose, the sooner they start to think about their mobile measurement strategy the better.

Having said that, there is still a ways to go for many organisations in term of sorting out their fixed web measurement strategy and WebTrends contribution to the debate was the release and publication of their Digital Marketing Maturity Model (DM3). This model provides a framework against which organisations can assess the maturity of the measurement capabilities on a number of different dimensions. I haven’t had the chance yet to look at it in detail, so something for next time. Till then…