A huge dataset and a team of data scientists are not sufficient to seize value creation opportunities. I propose here a check-list for value creation.
The bank, the data scientist and the branch manager
To introduce the check-list, let me share a personal experience I had with one client. This leading bank had just recruited a chief data officer and his task was to build solutions with data and analytics to solve the bank’s business problem. Nothing really original, but challenging though given the maturity of the company regarding analytics.
The CDO does the job perfectly: he starts by interviewing the managers in charge of the business and operations to identify the business problems and then he prioritizes them according to the probability of solving them with data.
“Churn” makes it to the top of the list: a lot of clients are going to the competition, which is becoming a major problem for that bank. Stimulated by this interesting challenge (how to reduce churn) the team works hard to source past data, identify clients churning, running models to identify the variables to explain why they churn, … And then apply the model to the current list of clients and predict which would churn.
They end up with a list of clients with a high probability of churn in the next quarter.
Pretty cool, right? So the team goes to the branches to meet the managers and proudly reveal that secret list of clients that will churn in the next three months. They had 10 meetings with 10 branches and they all developed the same sequence. 1) the team explains the problem 2) they explain how they worked 3) they announce they have the list of the clients that will churn and then …
the branch managers interrupts and says “I also have the list, here it is”.
The team was smiling first but stopped when they saw the list was 80% identical to theirs. And this occurred during each of the 10 meetings. How could this be possible? The branch manager had a very clear answer: “we know they will churn because they all are having their second child and are moving to a new house, they need us to finance the acquisition and the problem is that our competition does a better job than us. The problem is not to know they will churn, our problem is to have the good services to keep them with us”.
A check-list for value creation
This story is interesting because it captures the necessary elements to create value with data and analytics:
- a clear problem related to competitive advantage: here, the churn
- an explicit value creation KPI to act on: here, the lifetime value of clients
- data resources and capabilities: data of the clients, talented data scientists, IT infrastructure, …
- operational capabilities to respond. This was the missing part of the example I described previously. Having the prediction didn’t help at all because the company couldn’t respond to the signal.
The first three elements are critical, you can’t do anything without them.
But they are not sufficient and the fourth one is less present when it comes to identifying strategies and opportunities of value creation with data.
Using this framework helps to capture the promises of of value creation with data and analytics.
A key question you could ask yourself before making any investment decision or launching any project is: “how will we respond to the signal?”.
But let’s be clear, you shouldn’t interpret this as a Go/NoGo decision matrix. It’s a guide for better setting the scope of the project and envisioning the processes or operations you need to change to have an impact on performance, beyond data and analytics processes.
In the next editions, I’ll be more specific on these 4 components.