I want to deliver great customer experience not work on data models!Published: Tuesday, 21 December 2017 by Matthew Tod, Head of Analytics
OK, so our customers never actually admit this to us, but we at Celebrus know what you are thinking, and it is not ‘I want to spend my time building rich, deep, data models!’ It’s natural to focus on desired businesses outcomes rather than the mechanism through which they are achieved. BUT could I please ask for three minutes of your time because the key to delivering great customer experiences is ensuring that you have great customer data, and then using it creatively to power differentiated experiences.
We recently commissioned a study from Forrester Consulting (which you can download for free here) that asked 200 business executives globally about the challenges of delivering great customer experiences and, specifically, about collecting the data they need. This is what they reported:
Interesting, right? The limitation is not technology, it’s not securing investment and it’s not regulation but instead what might appear to be two very simple issues get in the way – the data model and the organization. Ill cover organizational design issues in another blog, but let’s address the data model today.
What is a data model? Well it refers to the definition of data, the logical inter-relationships and data flow between different data elements involved in the customer experience. It is, in essence, the organization’s data policy or manifesto, which defines how customer data is collected, processed, profiled, stored and used to improve both experiences and sales. Does your organization collect data in different ways, using different systems via the different channels which your customers use to interact and engage with you? Does this data reside in a range of systems and databases which are not fully integrated with one another? If so then your business, like 26% of respondents to our study, lacks a standard data model, meaning an accurate single view of the customer is hard to achieve.
Without a standardized data model there is no common language, which makes it extremely difficult to understand your customers as individuals and to build a truly personalized, omnichannel customer experience.
In the case of Celebrus, we have a defined data model that covers more than one hundred different classes of event, each of which have over 400 attributes. Our data model can be assembled from customer data collected from an almost infinite number of digital devices such as web, mobile, tablet, kiosk and ATM. Events range from behaviors such as search term inputs, button presses, mouse movements or screen swipes, and Celebrus captures ALL behavioral data, so NOTHING is lost.
As a result, large amounts of extremely rich, granular data are collected, and this is incorporated into a standardized data model. The creation of such a data model is the best way of providing complete visibility of all customer interactions across all touchpoints. Without a detailed data model which is fed by real-time data, it is extremely difficult to build a truly effective customer experience solution which extends to every touchpoint.
So by all means keep the omni channel utopian vision within sight, but please make sure it is underpinned by a robust, standardized data model, which incorporates and the most detailed customer data possible, collected from all of your customer touchpoints.
LinkedIn: Matthew Tod