The Importance Of Context Data In Web Analytics

Web Analytics

The whole buzz around ‘Big Data’ has created a sense of urgency for decision makers to leverage all the available data to its maximum potential. Every enterprise wants to hear a story behind an observed trend and every customer wants an experience which is relevant to his/her context. At the same time, 37% of mid-senior managers believe that generating insights from analysing big data is the most challenging need for mid and senior management across the business landscape.

Information in itself is not relevant. It is the context attached to the piece of information that stands to be the game changer. With the shift from Mass Marketing to Targeted Marketing and from Targeting to Personalisation, Contextual Marketing has emerged as the next big thing. Contextual data is about having the right data at the right time, in the most relevant situation and in a sharable format.

In the web analytics scenario, Conversion and Optimisation are the key objectives of most business decisions. Companies often run into the quantitative aspects of things quickly and end up designing improvement measures based on trends emerging out of a customer’s web journey. This is just the first layer. Adding qualitative aspects to the trends in a relevant manner is something which can differentiate the design of solutions and the way we think about conversion and engagement.

For example, a trend showing a decrease in conversion is not enough to tell what action needs to be taken and how it should be designed. It only reveals that there might be a need for action, if confirmed. Add the customer attributes to this data, tied up with the context around it and you should be able to answer the ‘why’ and ‘how’ of the action required. To add context to your web data, it is important to think of these premises from a perspective of maximizing the interactions with every touch point a customer meets with.

The Ubiquitous Web

The accessibility of web is no longer a constraint and customers rarely relate to the idea of a destination website. The digital connect is on their fingertips and hence the variations in the context of using a website are innumerable. It is important to understand in what environment a customer is accessing your website and whether the environment variables around him/her influence the purchase decision/end action.

For example, weather has a direct relationship with how a customer accesses web, depending upon other demographic attributes. The online sales can go up or down on a rainy day, depending upon the behavior of people in a particular region. A customer watching TV and browsing the web simultaneously would exhibit different behavioral traits than a customer who is listening to music and browsing the same content. It is therefore important to understand the interaction with various touch points at the same time in an online search or purchase journey.

From Data Collection To Data Selection

All sophisticated web analytics solutions aim to collect the maximum amount of data generated from clicks, impressions and other actions. According to a Krux research study, data collection volume has gone up by 400% year on year, from 10 events to 50 events per page. Preventing data loss and capturing maximum dimensions of data is what defines the rating of a web analytics tool. To have a contextual flavor of all the user generated data, it is important to have discretion in the data collection process followed with intelligence around the way data is stored.

It is equally important to understand the correlations between different data pieces in advance to avoid collecting redundant data and include more contextual data which serves as building blocks for customer journey fables. This can be possible by thinking through efficiency in tagging, cross flow of identity and information between different web properties one has as well as incorporating context in the data collection process

Top Down Versus Bottoms Up Story Creation

From a measurement and insights perspective, there can be two ways to think about story creation. One, to think of a high level trend and try to analyse all the related data focusing on the ‘Why’ or ‘What’ aspects of the trend. Another way of looking at it is to form layers bottoms up by tying up different data elements in the right context at the right time, to seamlessly build a story which has relevance to many interconnected chapters. A customer having exposure to multiple digital channels at different time scenarios has a tendency to exhibit different behaviour at every interaction, attributing it to the context around.

Connecting these dots of contextual behaviour helps in understanding the evolution of customers through various web interactions and helps identify personalised traits. This gives us a chance to validate the right context and explore many possibilities before we make a limited decision based on the behavioural and qualitative attributes of a customer or by scrutinising a customer’s actions.

Technology Aspect Of Context

Technology has made life simpler for analysts. There are millions of ways a customer can interact with businesses anytime, anywhere, across a variety of platforms. Customers can interact with you through the web in innumerable ways using different technology elements adding to the experience. A customer reaching your website through mobile vs. desktop and landing on the same page, has a very different context in terms of how he/she would react to your offering, content and purchase funnel. With the exponential usage of internet enabled mobile devices and the availability of myriad mobile apps, businesses have to continuously be on their toes to deliver a quicker, smarter and contextually personalised customer experience.

History Is Important But ‘Now’ Is The Key

Looking at the historical trends and deciding the future of customers is inevitable when we try to derive the life time value of customers. However, with the business landscape changing faster than ever, it is critical to understand the ‘now’ of a customer’s journey and respond in real time, clubbing it with data from past. Inbound experience design has started taking precedence over outbound experiences. Businesses should be ready for a customer with the right experience, and complete relevance, in the right time. Context is both ‘historical’ and ‘Just in time’.

Omni-Channel Is The New Magic Spell

Customers are changing their preferences every now and then and creating new paths for purchase. In the two-world scenario where the real world and virtual world co-exist every moment, the Omni-channel presence of customer is something which will define the way we look at customer behaviour. A customer browsing a product in the physical store or its reviews online is much more contextual for ‘push advertising’ or a customised discount than a customer who sits at home and browses your website. The former is more likely to be a serious buyer and leaves a trail that ought to be suitably capitalised by businesses. Online behaviour sets the background for context data a customer would generate in an offline experience and vice-versa.

The World Of ‘Interconnected’ Problems

Problems are not independent; they are interconnected. So, the solutions should also be thought through in a connected manner with all possible factors. The universe of web data has a constellation of different contexts which are related to each other.

The journey of an online customer has so many cues and footprints which can be matched with context data, such as time and place of access, demographic attributes and environment variables. It is crucial to ask pre-emptive questions around an observed online behaviour. These questions lead to hypotheses which are connected with each other and aim to solve multiple interconnected problems. For example, channel information, when combined with device and consumption information for an exhibited online behaviour, helps us understand the ‘Why’ of a specific customer behaviour.

To conclude, with all the context setting around the context data, it is imperative to mention that the right perspective and planned strategies around using context information are the key to better business decisions. They can lead to better predictive analyses of online customers’ behaviour, scaling up qualitative aspects of the data deluge and shaping up the new face of Personalisation and Content Optimisation.

Gaurav Kumar is a Manager at Mu Sigma. He comes with over 7 years of experience in Marketing Analytics, Web Analytics and Digital transformations, in various domains including Retail, Telecom, and Insurance. He has an MBA degree from MICA, Ahmedabad.