Riding the insurance big data wave
Riding the data wave!
Insurance has always depended on data. In 1693, scientist Edmond Halley (of the famous comet) produced a Royal Society paper on the mortality rates in Breslau, Poland. This was only made possible due to data because the Breslau authorities recorded the ages of citizens at their death. Thus, modern actuarial tables were born. And since then, that initial data ripple has grown into a big data wave of tsunami-like proportions.
IBM claims that businesses around the world generate nearly 2.5 quintillion bytes of data every day, the vast majority of global data surging in the last two years. Insurance as a sector can lay claim to at least 3% of that, which is still an awesome amount of data.
The urgent question for insurers now is how can you make sense of the tsunami of data coming your way? How can you profitably ride the data wave and avoid being swamped by it? Or worse, be swept away by nimbler competition that successfully harnesses the power of the data wave?
The challenge of managing the 3Vs – volume, velocity, and variety
Ever trying drinking from a fire hose? Some say that is the equivalent of coping with the volume and velocity of data today
Legacy data systems, based on spreadsheets or in siloed systems, were not designed to manage the quantity of data available to insurers now. The variety of data is also expanding far beyond the finite fields that were originally envisaged by insurers. But the challenge is not simply a technical one.
Amazon’s CEO Jeff Bezos is on record saying that “speed matters in business” and “being slow is going to be expensive”. People, process and cultural limitations can hold your company back from a speed and agility perspective – no matter how fast you collect and process data.
Others suggest adding validity and value to the V3 list. MetLife Asia’s CMO, Sanjeev Kapur, says that companies who use data to add value to consumers’ lives will be able to generate more profit. “This shift from volume, variety and velocity is the real pivot that we are trying to make from big data to smart data,” he said.
“We ask ourselves, ‘Do we want to be in the business of only helping customers when they are running into an adversity? Or do we really want to be in the business of not only helping them but also preventing them from having an adversity?”
The benefits of smart data
Using data better will potentially open up a cornucopia of benefits for both insurers and customers. There are a number of potentially positive societal outcomes from better data analytics.
Better risk signalling
The more data you have about an individual, the more they can understand their risks and how to mitigate them. This feedback loop allows customers to modify their behaviour to reduce their risk exposure.
Increasing lifestyle and health information
Lifestyle and health data collected by insurers is being used to influence behaviour and to reduce the risk for both the individual and the insurer. Research suggests that 93% of retail customers (in Australia, France, Germany, the UK and US) are willing to share personalised data if they can save money or receive customised offers.
Greater premium dispersion
Improved data will allow insurers to price more precisely, down to the individual insured. Current pricing factors will move from broad segments to more accurate predictors of individual risk characteristics.
Tailored car insurance
Car insurance is being transformed by telematics devices that measure how, when and where a car is driven. Instead of charging extras for all drivers under age 25, you can price using the specific driving behaviour of each young driver.
Improved genetic information
Genetic testing can be used by insurers to better understand the risks to which individuals are pre-disposed, though caution is required as there is no absolute certainty the risk will eventuate. The insurer’s ability to combine test information with other client data and medical expertise can assist the individual to recognise and reduce their health risk.
Connected Homes and the Internet of Things
Connected homes can benefit insurance policyholders by monitoring risk factors such as temperature, smoke, water usage, and by acting to prevent risks from occurring.
How to start small – where is the greatest impact to be felt?
If we consider big data an elephant, sound advice is – don’t try to eat it in one bite. You need to tackle the challenge in small increments.
Use the info at hand
Start with your current policy information and claims. Begin by analysing that and feeding it back within the organisation. Make sure that after any claims, data is used to provide feedback into the product cycle to inform risks, pricing and profit.
Collate your data
Is all your data in one place? Or is it scattered in a range of sources; discrete databases, paper records and m-plain text files? Amazing as it sounds, the majority of insurers do not have ready access to their own data.
Make it accessible
When you unify your data, ensure it is formattable and accessible to different departments.
Apply insights
Once your data is in an accessible form, you can use data analytics to dive into it and discover new insights. Apply these insights to better inform product offers and development.
Focus on the customer experience (CX)
Using social data to better understand customer needs, interactions can be personalised to provide improved CX. This can be vital: according to one survey, approximately 41% of respondents have left an insurer because of poor CX. In another survey, 27% of Gen Yers and 23% of Gen Xers indicated that they want to interact with their insurer through digital self-service.
The tools and technology that may help
Artificial Intelligence (AI)
According to research led by the McKinsey Global Institute (MGI), processing data can be done better and faster with machines. Now is the time to investigate how AI tools can help you surf the data wave, reduce manual work and optimise your productivity.
Analytics
Big data analytics software will help you clean up your data, collate it, build accurate models and employ predictive machine learning models. Here’s a list of some of the best tools you can access.
Ask for help
Companies who are agile, fast-movers and early adopters are more likely to ask for help with data analytics from third parties, compared with companies that are moving more slowly. Insurers shouldn’t assume they have the internal expertise they need.
Get ready to reach out and ride the wave.