Reinsurance is a complicated industry and it's only getting more complex.
Reinsurers provide insurance for other insurance companies to keep those companies' risks to a minimum and according to a recent report from Deloitte, their industry is facing a problem.
They're putting together increasingly customized, individually-negotiated contracts for clients, incorporating data from several disparate sources — business partners, administrative systems, internal documents and reinsurance terms — and they're doing it manually.
According to Deloitte, 92 percent of the executives surveyed said their company’s reinsurance contracts have a medium or high level of complexity.
These complicated contracts offer better risk management and work better, financially, for all the companies involved but the process is risky in a different way — when customized reinsurance products are put together, processed and administered manually, that can lead to errors.
The way to solve this problem is technology — but there's another problem: the reinsurance industry hasn't kept up with IT or data management. Reinsurers are using a patchwork of solutions to put together each policy, sometimes using spreadsheets or even a pencil and paper to pull together data housed in several different silos.
It's an awkward approach to business intelligence and data analytics that opens reinsurers up to risk and it doesn't scale.
Something has to change.
Data management, data analytics and the reinsurance industry
Data management and analytics is a huge issue for reinsurers.
When asked to name the top three pain points for their organizations, Deloitte's respondents listed data quality at the top. Operations came in second, followed by technology. The study concluded that data management needs to be the reinsurance industry's top priority, but data analytics is a challenge in any industry that requires access to so many siloed data sources.
Reinsurance contracts often cover multiple insurance products and businesses, and it's a business intelligence challenge: the data for each of those products or businesses usually live in different systems, agents might run reports on each system, download the reports into spreadsheets and compare them manually. The process might take hours.
But what if there was a way to store all the information relevant to a policy in one system? If a reinsurance organization could channel information from all its data sources into one repository? That would be the first step in creating a system that would help automate parts of the reinsurance process and simplify data analytics.
The tool that can help is Microsoft Azure Data Lake.
What is Data Lake?
The first thing you should know about Azure Data Lake is that's it's not exactly a database.
Rather, a data lake is a very large repository that stores unstructured data in any format. A data lake can take in web logs, spreadsheets, information from the Internet of Things (IoT) or any other kind of data an organization needs to store. That data doesn't have to be configured as it would for a traditional database; it's stored as is and assigned an identifier and a set of metadata tags that will allow the information to come up in searches.
As with an actual lake, several streams of data can feed into it. This makes it ideal for organizations like reinsurers, which require information from several disparate systems to administer their policies. The number of data sources used by reinsurers might overwhelm more commonly used analytic software like Power BI and Tableau but an Azure Data Lake can handle terabytes of information, which makes it a business intelligence dream for industries like reinsurance.
However, data lakes do have some limitations. Just like a real-life lake, it's very easy to dump things into them but it's a lot harder to get the right things out.
Data analytics and the data lake
A data lake isn't for your average business user. There are a few big analytics challenges involved in using one.
- Making sure all the data in the lake is current
- Deciding which insights you want from all that data
- Writing the queries that will bring up the right information
An organization typically needs a data analyst or a specialist in business intelligence to make sure these things happen.
For example, while putting data into the data lake is easy, current information from all the systems the organization uses to compile reports needs to be uploaded into the lake regularly in order for Azure Data Lake to be a good data analytics tool for a reinsurer. That's an automated process that needs to be built by a specialist.
But what if a company doesn't have a specialist on staff? That's where an IT consultant comes in. A consultant with a background in data management and business intelligence can set up a data lake that regularly takes in the information it needs.
A good consultant with a business intelligence background can also help a company understand which handful of insights it needs to glean from the information in the data lake and set up specific queries that will get that information out of the lake.
That way, agents can run the query whenever they need to, generating reports in minutes that currently take hours. It's a data analytics approach that scales, helps reinsurers reduce human error and lets reinsurers devote their energy to the work that needs to be done by a human — like negotiating bespoke policies for their clients.