Big data opens new horizons for insurersFirms must learn to mine
gems of information buried under mountains of data
Brendan McCabe, Head of Insurance Coverage, HSBC
Big data is transforming the insurance industry, presenting
difficult management challenges but also rich opportunities to
reduce risk, optimise pricing and customise services.
Insurers have always depended on their capacity to master
statistics. Traditional datasets on demographics, longevity, crime,
natural disasters and so on may be huge but they are structured.
Big data, for the most part, is not. Social media posts, mobile
communications, online searches and real-time signals from billions
of internet-connected devices do not yield up their secrets readily
to a spreadsheet. Teasing out patterns from oceans of high-volume,
high-velocity and high-variety information is tough. Companies that
succeed will be better able to streamline claims operations and
detect fraud. Perhaps most importantly, more-detailed consumer
profiles will help them fend off competition from non-insurers by
improving customer targeting and service.
Big data is growing fast – at 40% a year in the US on one
estimatei - as business wakes up to its potential. The Organisation
for Economic Co-operation and Development reckons data-driven
innovation can increase productivity by 5% to 10%ii. McKinsey
Global Institute estimates that US health-care costs could fall by
8%iii. In the insurance sector, one firm that recently built a big
data platform says it can now support the reporting and analytical
needs of nearly as four times as many staffiv. Today’s car – a
computer on wheelsThe spread of telematics – data sent from
recorders fitted in vehicles - is a good example of how insurers
are harnessing big data. In the past, the premium would reflect a
driver’s age and gender, type of car, post code and claims history
– information too broad to predict how much risk the insurer is
really running. Now, data transmitted from the car or truck enable
insurers to distinguish a cautious driver from one who brakes and
corners hardv.
This Usage Based Insurance (UBI) gives companies a powerful tool
to match the rate charged to the actual risk incurred. An obvious
danger for insurers slow to adopt UBI is that they will be saddled
with less-safe drivers whom its rivals have been able to identify
and reject.
Change is afoot in other sectors as ever more people and things
are connected to the Internet. Biometric data sent through wearable
fitness devices will enable doctors and insurers to spot medical
problems early. Home insurers can draw on real-time information
from arrays of in-house sensors and smart meters. Big data
transmitted from machine to machine (M2M) may render some
commercial insurance business models obsolete by giving insurers
deep insight into how factories operate their equipment.
The capacity to calculate risk more accurately will help
insurers to minimise the risk-based capital they will need to hold
under the looming Solvency II directive. Collating big data should
in principle also assist them in meeting requirements under Pillar
III of the directive for more frequent, extensive reportingvi.
However, survey evidence shows many firms losing the race to
overcome the related data and IT challenges in time to implement
Solvency II by the 1 January 2016 deadlinevii.
A snooper’s charter?Access to masses of personal data plainly
raises pressing privacy issues. In a 2014 survey, only 27% of
people across 15 countries were willing to trade their privacy for
the convenience of easier access to informationviii. Alert to these
concerns, the European Union is proposing data-protection rules
that would let policyholders demand the removal of their
information from insurers’ databases. What’s more, the head of the
Association of British Insurers has warned of a regulatory
crackdown if the use of big data results in higher premiums for
riskier customers, effectively making insurance unaffordable for
some peopleix.
Data security is an equally sensitive issue following
high-profile hacking attacks and information leaks. The legal
question of who owns data is also emotive: does the information
from a car’s decoder belong to the driver, the insurer or the auto
manufacturer? And the inappropriate use of data is a serious
business risk, as companies worldwide periodically discover to
their cost.
Insurers have been criticised even within the industry for
having been slow to make use of big datax. However, firms are
picking up the pace. The Progressive Corp., a big Ohio-based car
insurer, wrote more than $2 billion in UBI-based auto insurance in
2013 out of total premiums of $17 billionxi. The technical
complexity of operating big data-based systems is daunting:
Progressive captures data from more than 9 billion driving miles at
one-second intervals. It is an open question whether smaller
insurers will have the resources to develop proprietary systems. If
not, big data may prove to be a catalyst for consolidation.
Breaking down the silosRegardless of whether it is handled
internally or outsourced to specialist boutiques, big data requires
top-notch analysts to pick out the signal from the noise in huge
data plumes. But the main challenge, according to 60% of
respondents to an EIU survey of executives, is to overturn the silo
mentality of insurers that prevents the capture and application of
big data across different business linesxii.
A failure to adapt will expose the industry to increased
competition from Internet and e-commerce companies that have been
quicker to exploit big data. The threat is not theoretical. Google
has bought a UK website that offers personal finance products
including insurancexiii and is reportedly plotting a move into auto
insurance in the USxiv. Change is uncomfortable, but doing nothing
is not an option for insurers: big data guarantees disruption come
what may. The sooner companies embrace big data, the greater their
chance of cementing brand loyalty and stealing a march on their
competitors.
i http://www.idc.com/research/viewtoc.jsp?containerId=233485ii
http://www.oecd.org/sti/inno/data-driven-innovation-interim-synthesis.pdf,
p18iii
http://www.mckinsey.com/insights/business_technology/big_data_the_next_frontier_for_innovationiv
http://www.dataiq.co.uk/news/20140820/zenith-insurance-boosts-performance-big-data-platformv
http://www.naic.org/cipr_topics/topic_usage_based_insurance.htmvi
http://www.invoke-software.com/banking-insurance/solutions/solvency2-regulatory-reporting/white-paper?gclid=CNStkKC94cMCFQkcwwodpqQAjgvii
http://www.ey.com/GL/en/Industries/Financial-Services/Insurance/EY-european-solvency-ii-survey-2014?gclid=CK_srti-4cMCFUKr2wodfUEA9Aviii
http://www.emc.com/campaign/privacy-index/index.htmix
http://www.ft.com/cms/s/0/08f6049c-a7cd-11e4-8e78-00144feab7de.html?ftcamp=crm/email/201522/nbe/CompaniesBySector/product&siteedition=uk#axz
z3QmZAW4SD x
http://www.computing.co.uk/ctg/news/2360744/insurance-industry-behind-on-harnessing-big-dataxi
http://media.corporate-ir.net/media_files/irol/81/81824/arInter/13_annual/assets/pdf/Progressive-2013-AR-BW.pdfxii
http://www.uk.capgemini.com/resources/the-deciding-factor-big-data-decision-makingxiii
http://www.insuranceage.co.uk/insurance-age/news/2032047/google-buys-uk-aggregator-beatthatquotecomxiv
http://blogs.wsj.com/digits/2015/01/08/google-wants-to-sell-you-auto-insurance/