Building a strategy for data management
When I talk with our insurance customers, many ask me how they should manage their data and get better insights from it. In this blog post, I intend to provide some perspective.
At its most useful, data can provide workers across all areas of the business with the information they need to carry out their jobs in the most effective manner. To do that, the data must satisfy a number of criteria:
- It needs to be descriptive; it needs to tell us what is happening.
- It needs to be diagnostic; it should be able to tell us why a particular event happened.
- It needs to be predictive and provide some insight into what will happen in the future.
- It also needs to be prescriptive and tell us what we should do next.
Role-tailored insights
Generally speaking, there are three types of user that require access to data.
We have the information worker that we want to enable with self-service and exploration by using tools such as Microsoft Power BI.
There’s the IT professional who builds and supports the data modelling, data warehousing functions, and enterprise data management strategy. These are the ones who are more concerned about the overall infrastructure and the way in which enterprise data is managed.
And then you have the data scientists who are doing the advanced analytics. They’re just looking for a tool to enable them to do this. They don’t care if it scrapes the data on the fly or if it is part of a larger enterprise data strategy. They’re also not that concerned by master data management; they just want to be able to do advanced analytics and get access to the results faster for better informed decision-making.
So from an information worker and IT professional perspective, it’s all about business intelligence (BI) enablement. And from an IT professional and data scientist’s perspective, it’s about providing a relevant and reliable capability for advanced analytics.
Both require an enterprise data management environment.
What is enterprise data management?
When we talk about enterprise data management, it boils down to two areas: master data management – making sure there are consistencies across the business with data; and data lifecycle management – where the data came from and, if it changed, we want to know where that change occurred, what the impact of that change was, and if it was an error then it goes back to the system of record. Data lineage is a key aspect of master data management.
Enterprise data management must also take into account data integration, data governance and, from a reporting perspective, BI, predictive analytics and machine learning.
At Microsoft, we support insurance carriers with self-help business intelligence solutions. The aim is to help them gain insight into their business performance and profitability, improve their combined profit and loss ratios, and provide an integrated view of enterprise-wide risk exposure. And that could be through gaining real-time insights into business exposure with risk management analytics, driving business performance insight with financial analytics, or embracing market and customer insight with big data, and turning customer sentiment into opportunity with social analytics.
Data management choices
In basic terms, insurance carriers have two options when it comes to how they manage their data and gain insights from it. 1) They can buy a tool that will scrape and consolidate their data. 2) They can build an enterprise-wide strategy to improve the way they ingest, control and manage their data.
So what approach should you take for your business and which is best? Ideally, you need to work on a future-proof data strategy from an enterprise perspective and give your workers the tools they need to work with the data and gain insights from it as it becomes more relevant and useful to them. The aim is to take information from all areas of the business and create a single reference point where all data is reliable and accurate. Just think what you would be able to do: what-if insights, channel optimization, understanding customer churn and customer value, determining real-time cash positions, portfolio management, actuarial and risk modelling, the list goes on…
The other thing to remember is that you don’t necessarily have to replace your existing technology to achieve this. We can work with you to make the most of your existing investments and help you gain more control over your data lifecycle.
And what happens if you do nothing at all? Well, it’s no secret that the world’s data is growing exponentially. If you don’t try and get control of the data you have now, how are you going to manage it in future?
For those of you who do master your data now, not only are you addressing this problem head on, but in the process you will gain better business insights and, as a result, become a much more nimble and adaptable company.
