News & Articles
Hiring a Data Scientist for the Insurance Industry – Part Two | 2019 Overview
4th March 2020
In Part 1 of our article, we discussed that many Insurance businesses rushed into hiring Data Scientists and for some this was not a successful adventure. The 3 themes that emerged in our conversations with Senior figures in the market that the main challenges were around:
- The Strategy
- Strength of Data
- Assessing the Candidates Capabilities
We discussed and looking at the strategy challenges in Part 1 and in Part 2 we shall look at Data and the Candidates.
How strong is the Data?
One of the main challenges implementing effective Data Science into any business let alone an Insurance Company is Data. In order to gain effect and usable insights you need to have a large amount of easily accessible and clean data and the reality is that most insurers have the complete opposite to this.
“Most insurers only have access to 10-20% of their own data”
In theory, because Insurance is one of the longest standing industries in Financial Services, they should have some of the largest and best data stores. However, accessing this data is challenging for many.
It isn’t digitalised, there is no common unity in the various divisions of an Insurer to store the data in the same way, there are many different legacy systems in each company (often lots of data is lost when transferring to a new system) and large amounts of it is unstructured.
For many it’s cheaper and easier to partner with 3rd party data providers to access data they can use, rather than getting it from their own internal sources. Although this information is available internally and is more relevant than the data they get from the 3rd Parties, they simply can’t access it.
As we move into a new decade, The “Data War” in insurance won’t necessarily be won by those who buy the most data and who has the most, it will be won by those who can start to access their own data with the greatest efficiency in a cost effective manner.
Companies who invest in their Databases, by building effective Data Warehouses and then look to apply Data Science techniques will and do see the most benefit. In the market today there are only a handful of businesses (that we have met) who have taken that approach. It’s taking one step backwards to take two steps forward.
This is one of the reasons InurTechs have been able to disrupt the market and are able to be innovative more easily.
The data they collect, and store is readily accessible and allows the Data Scientists to spend their time doing what they are good at, finding patterns and insights which can be used advantageously to improve and grow a business. For many Data Scientists at larger or historic Insurers, they will spend 70-80% of their time sourcing and cleaning data so they can use it. This is not the best use of all their training and experience and is a contributor to why those who made an investment in Data Science do not see the return they want.
There is large supply of Data Scientists in the Financial Services market, but expectedly so, those who have Insurance exposure are harder to come by and will cost more.
When many clients look to hire a Data Scientist the common theme is to hire someone with 3-5 years’ experience, typically from a non-insurance background.
Candidates without direct industry exposure are attracted to come into Insurance because it’s a chance to develop and progress their careers quickly, if it works that is.
Its widely known that the Insurance Industry is 10-12 years behind other Financial Services businesses in how they use their Data and how they employ effective Data Science techniques. For many candidates, this is an interesting challenge and one which they feel they can conquer.
The reality though is that hiring someone with only a few years of working experience means you will get someone who is technically strong and well educated. However, these individuals are not equipped to deal with the commercial and political challenges that these roles pose. Knowing how to work in environment where data and tools are limited compared to what they have known and started with in their careers makes it a particularly difficult challenge.
If a business is to hire a junior Data Scientist, they must be realistic in their expectations and must give the individual the best platform to succeed, otherwise it will inevitably be a wasted investment. More pressingly if there is no clear strategy for what they want this person to achieve and how, it will only compound the issue.
If a business is unable to change their systems, improve their data or define a strategy then they need to hire someone with enough experience who can operate effectively in this environment. Typically looking at a more seasoned professional with 10+ years’ experience would give them that. An individual who can define the data strategy for either a team or whole business
In order to build and benefit from a successful Data Science team, there must first be a clear strategy of what business unit it will be implemented into. Then the business needs to be realistic its expectations taking into account all the resources that will be available, and the experience of the candidate being hired. Then finally if and where possible making Data as accessible as it can be.
The intention of this article is not to paint a picture of doom and gloom rather one of caution. The Insurance industry could provide a much better service to its customers and in turn increase their profits if they approach Data Science in the right way. Taking a step back, getting the house in order and creating a clear and definitive strategy will create successful results and give the business what it expects. Going into it poorly planned and with a “give it go” attitude, won’t work and will see the business fall behind others.
Follow us for more articles…
Hiring a Data Scientist for the Insurance Industry – Part One | 2019 Overview
11 February 2020
Where has all the talent gone?
25 November 2019
What makes a good recruitment consultant?
20 November 2019