by Monica Woolmer, Teradata
I am sure you have heard it said before that fashion and trends repeat every twenty years. As it happens, in 1997 I joined the Marketing Workbench team at Harrah’s Entertainment. Marketing Workbench was designed to be marketing’s analytic environment to drive customer segmentation through predictive analysis. It was this career move that started my focus on customer analytics.
I decided to have a look at what has changed (or not) in customer analytics from then to now. I actually left Harrah’s in 2003, so my “then” includes six years. I decided to limit my trip down memory lane to three key areas:
– Data Platform,
– Analytics, and
– Customer Interactions.
Data Platform – The underlying database technology utilised (no surprise here) was Teradata. Since then many features and functions have been added and the amount of data supported has increased exponentially. The biggest change to me with regards to the data platform for analytics available today is the introduction of Hadoop (HDFS) and the recognition that there is a now an analytic ecosystem rather than a single data platform.
Another current capability that was not present back then is cloud-based technology. According to Teradata’s recent Data Warehouse Survey, more than 90 percent of our customers aim to have a hybrid cloud environment by 2020. Hybrid Cloud is a computing environment which uses a mix of on-premises, managed cloud and public cloud orchestrated to work together. Hybrid Cloud is all about flexibility and deployment options. Read more Hybrid Cloud in the CITO Research article ‘Analytics without Borders‘.
Analytics – Analytics were a core strength of Harrah’s and many predictive models (e.g., customer life time value, share of wallet) were developed and utilised to improve customer segmentation. Predictive models continue to exist today and they benefit by being able to run against more data points than ever before. As illustrated in the diagram below, analytics has evolved from Descriptive (e.g., executive dashboard) to Predictive (e.g., propensity score) to Prescriptive (e.g., automated decisions):
Another new capability today is the advancement of visualisation techniques. This includes not only using visualisations (e.g., geospatial representations) as the actual user interface to the data but also includes new types of visualisations. For customer analytics we rely heavily on Flow visualisation for Customer Path analysis; Hierarchy visualisation for Relationship analysis; and Affinity visualisation for Product Affinity analysis.
Above: Flow Visualisation Example showing the paths to in store purchases.
Customer Interactions – At Harrah’s, Marketing Workbench evolved from daily batch loads to message-based loads to ultimately publishing customer insights onto the message hub. In many ways Harrah’s was on the leading edge of using the results of analytics in customer interactions. For example, utilising the customer’s predicted life-time-value score in determining at what rate a hotel room should be offered.
The major difference to today is the ubiquitous use of mobile devices. Customers are now in charge of when and where they initiate interactions. It is up to each company to ensure that insights are shared across all channels. What is even better is in providing the visibility to all customer transactions and interactions as well all the ability to orchestrate any response such that the right message is given (or action is taken) that considers both the customer need and the priorities of the company.
So what hasn’t changed? Getting the chance to work with smart people and working together as a team to use analytics to improve customer experiences.
If someone is just starting their customer analytics career journey, I wonder what they will find in another twenty years. Perhaps customer sentiment will not solely be based on the words a customer uses, but rather include analysing their facial expressions. Perhaps robots become more prevalent in customer service.
Whatever the future holds there will be a role for customer analytics. From then to now as well as from today into the future, customer analytics will continue to provide valuable insights – for the companies that use them.