Anyone faced with the task of turning customer data into meaningful insight will be aware of the significant challenges. What should really be an enjoyable and creative experience is all too often turned into an overly-technical, time-consuming and draining process.
Turning raw data into actionable insight is nothing new, of course. I started my career as a data analyst in a full-service advertising agency in the mid-90s when agency data planning was in its infancy.
It was useful to see how advertising and media planners used data to help justify their proposals. Typically, the data they used was locked up in software packages which would only produce specific outputs. The software could identify some really interesting market trends and recommend ways of addressing them. The more inquisitive planners could see that limitations in the applications prevented them from tapping into value they could sense in the underlying datasets.
I soon discovered that with CRM analytics, none of the traditional rules applied. You were free to create and enhance the datasets you wanted to analyse. You could design surveys to collect the information you needed. There were no limitations imposed by a rigid software interface. However, the price of this freedom and flexibility was inefficiency and complexity in achieving your objective.
The process of turning customer data into meaningful insight has always been arduous at the best of times. There is no standard for how the data is structured and there is no standard interface for interrogating the data. Typically, the solution won’t be found in a single tool. A range of software and techniques is usually necessary to produce what you need.
The back-to-basics approach involved in developing insight from customer data helps create a unique mindset. When confronted with rigid software interfaces, CRM analysts will be thinking of how they can export the underlying data into their analytics package of choice.
The analytics process hasn’t changed since the early days of CRM analytics:
•Data sourcing - generic scripts are usually written to create standard analytic data extracts from customer databases.
•Data transformation - this typically involves functionality such as aggregating data in relational databases at the customer level, creating RFM segments, recoding existing fields, etc.
•Data exploration - examining the data from a range of perspectives will allow a story to emerge to shed light on the issue being investigated.
•Report creation - a report providing insight into the issue being investigated will typically be created in PowerPoint or Word. This will usually incorporate a range of visual outputs including graphs, maps, Venn diagrams, word clouds, etc.
In the mid-90s, CRM-focused analytics software promised little more than lightning-fast queries on multi-million record data sets. Now, they are moving more towards simplifying the whole process outlined above.
In addition to allowing experienced analysts to be more focused and productive, software which successfully simplifies the process will allow non-technical users to solve business problems using their customer data resources. If someone can identify a business issue which they believe they can help solve, they should be able to do so without worrying about writing code or juggling a myriad of software applications.
All customer databases contain vital clues to help shape strategic direction and channel marketing budgets, but they can sometimes be difficult to unearth. Some firms have spent huge sums on large teams, hardware and software to ensure that no stone is left unturned for insights from their customer data. At the other end of the scale, there are huge opportunities for companies to realise efficiently the potential of their untapped customer data assets in a cost-effective way.
Striking the right balance between human and computer effort is still a real challenge for the CRM analytics industry. Ultimately, getting it right will help more businesses harness under-utilised customer data resources to become more competitive.