Customer 360: Personalising The Customer Journey With A Single Customer View
How digital is enhancing customer service expectations within the retail sector
Individuals as data sources People are emitting ever more data through social networks, devices and online searches on a daily basis. Previously, much of this data has been dispersed in silos, because the connections between different datasets have not always been direct, hampering attempts to generate value from it. New technology has addressed these disconnects with great strides forward in interconnecting and understanding a wide range of data sources.
People-centric differentiated wisdom The truly digital enterprise now has an advanced level of insight – of people-centric differentiated wisdom – that effectively puts customers at the heart of business strategy and actions. Much of this is achieved by new levels of connectivity that provides better tracking and reporting of individual behavior from each point of interaction, which can then be analysed in unison to provide a much finer-grained detail of an individual’s behaviour. As we will demonstrate, this is already happening in retail, but it is also gaining momentum in other sectors too. One of the main enablers of this people-centric focus has been made possible by the move beyond basic machine learning to far-reaching deep learning techniques. The result is much more personalized and engaging operations.
People as customers: lessons from retail | A change in approach In order to achieve a fully integrated ‘Customer 360’ view, retail organisations need the right tools and environments. Previously, we have had variations of this approach, beginning with the single customer view (SCV) – an outsidelooking-in approach developed over 15 years ago and based on a deterministic, single solution perspective. A new approach is required; one that focuses on the probabilistic considering a number of potential outcomes using behavioral analysis and scenario extrapolation.
The current state of digital retail Much of today’s customer focus in retail involves basic matching that links datasets together based on common characteristics. Hypothetically, this is easy to do and can deliver value, but each retailer has a multitude of systems that can make this a complex undertaking. Advances in machine learning play a part here – by training a data model to highlight potential matches, which can provide the foundation for automatic matching and profiling of customers. This can be as sophisticated as tracking a user through multiple websites through usage analysis or as basic as data matching with a partial address or surname. Even with basic information, retailers will make assumptions. For example, the type of device used to access a website can directly affect the buying experience – a more premium computer brand could potentially result in a higher price being set. Now, retail profiling is becoming more targeted. Dynamic pricing is a good current example that demonstrates how retailers are attempting to quantify abstract customer values, such as the propensity to buy. In this instance, an estimate is made based on the number of visits an individual makes to a specific part of a website. This is used most often for dynamic pricing of very specific, time-limited products such as airline tickets. However crude, this is a tactic with sound psychological reasoning: raising the price can suggest the flight is close to being sold out, which provides an added incentive to purchase.