Insurers are investing in data scientists

A few weeks ago I described a few results of a survey we have launched during the last quarter of last year around the role and importance of data in insurance. My blog post can be found here. Since then we have published a report summarizing the findings of this survey that our members can find here. An interesting trend we identified based on this survey was the need for insurers to hire more data scientists with advanced degrees and strong background in data and computer science. Indeed we think technology is not enough nowadays and insurers need to also invest in people with deep skills in this domain. I recently came across the following article from INN: Sentry Insurance Gifts $4 Million to Grow Data Science. It seems to validate our findings and I expect to see more of these kinds of initiatives going forward.

Why the insurance industry needs more data scientists

Celent will soon be publishing an update to our 2013 report Perceptions and Misconceptions of Big Data in Insurance. In this report we looked at various elements in relation to the role and perception of data in insurance to understand where the industry was in terms of adoption of data-related technologies and more particularly Big Data. To do so we used what we call our Big Data Maturity Model. This model uses seven dimensions to categorize the industry in terms of their maturity level when it comes to adopting Big Data: Figure 1 big data   We came across an interesting article recently in the Insurance Journal that said insurers needed to hire more Big Data professionals. While we agree with this statement, we have already noticed in the early results of our 2015 survey (still in progress) that insurers have now more data scientist experts as shown in the following figure: Figure 2 data tools Technology is not enough and insurers have understood that if they want to make the most of data-related technologies they need to hire highly skilled people with solid knowledge of machine learning, statistics and predictive analytics. This is an interesting early finding and we look forward to provide our members with more on our seven model dimensions soon. Stay tuned!

The world’s most connected human

I recently read about Chris Dancy, Chief Digital Officer and Senior Vice President of Healthways, Inc. and “The world’s most connected human.” In my line of business and as an avid NPR listener, I really should have heard of him earlier than now. If you haven’t heard of him and you are reading this blog, you should know about him, too. Chris utilizes up to 700 sensors, devices, applications, and services to track, analyze, and optimize his very existence every minute of every day. I listened to a few of his interviews (I am a curious person!) and found that he has been doing this self-tracking for nearly six years. You can really say he was on the cutting edge of this idea of a quantified self before most people even heard of the FitBit. According to Chris, this quantification enables him to see the connections of otherwise invisible data. As a result, he has experienced dramatic upgrades to his health, productivity, and quality of life. So what does he track? In a NPR interview while wearing five sensors (FitBit, Nike Field band, BodyMedia sensor, Wahoo TICKR, and his phone) Chris talked about how he has become ‘one with the data’ because he has seen the benefits of understanding his moods, heart rate, and overall health. He admitted that it’s not for everyone, but being a data junkie he said this behavior fit right into his interests. He expanded what he measured because he was interested in the data for which companies are willing to give discounts. If a company was willing to give him a $600 discount for seeing a doctor, going to the gym, and eating better, he wanted to know what data were they considering and what benefits he would derive as a result of knowing what the data said. He also said something very key: “If you can measure it, someone will and that someone should be you.” So why has he intrigued me so much? Because he said in 2013 that he believed the idea of a quantified self would be ubiquitous in five years. And it would expand beyond the fitness worlds and health care implications to the physical workplace and other industries. He saw sensors as being omnipresent in giving people feedback while they work. Examples could be environmental sensors that let someone change the lighting in their office to reflect a mood one had while on vacation or track ambient sound so that the sensor notifies you to reflect on the tone of voice used in a conversation. The goal, of course, is to have a more productive work environment. Chris Dancy’s rationale for wanting to know more about the data companies use to give discounts intrigues me the most. Many health insurance companies give discounts for proving that you lead an active lifestyle and for years now, consumers have been able to send driving data to auto insurance firms who offer reduced rates for good driving via a dongle that is plugged into their car’s onboard diagnostics port. Recently this practice moved into the realm of life insurance. John Hancock has become the first life insurer to offer ratepayers a discount when they use Fitbit wristbands that enable exercise tracking. John Hancock policyholders who wear a Fitbit and connect it to the internet can get discounts of up to 15% on their life insurance policy as part of Hancock’s partnership with Vitality, a service provider that integrates wellness benefits with life insurance. I already consider myself a quantified being because I track my fitness daily through my FitBit, and use that data to push myself to walk more and be more active. I am not sure I believe that my work environment needs sensors to make me more productive at work, but maybe they would. I don’t share my FitBit data with anyone yet, but I would be willing in the right circumstances. My insurers are not asking for my data which to me means that many insurers are not ready to accept the data. As mentioned above, John Hancock is the first life insurer; maybe others are soon to follow.  Will it happen in the next three years?  My gut instinct says no but I hope to be proven wrong. IBM’s Watson Health Cloud suggests that the medical industry is looking more deeply into how to capture, analyze, and use the multitude of medical data that is created every year, some of which is from fitness trackers and other sensors.  Maybe Watson’s analysis and cloud availability of data will yield better methods of underwriting for insurance. Yet, going back to Chris Dancy . . . during one of the NPR broadcasts Robert Wachter, author of the Digital Doctor, said that today very little of the extraordinary amount of data Chris was capturing is truly useful to doctors or insurers. I guess if that changes, Chris will be ready.

A four year old’s employment records (or how not to handle a data breach)

Yesterday one of my four year old twins received a letter from a major health insurance carrier (we’ll leave them nameless, tempting as it is). The letter states that the carrier had a data breach and that his information may have been included. The list was pretty extensive, including name, address, telephone number, email address, date of birth, social security number and employment history. That’s a pretty big list and everything you need to steal an identity. They assure me that no health information was shared, but I think they have their priorities wrong. I don’t care if the thief knows I have high cholesterol, but I do care that they have my social security number. I admit I am curious about the employment history of my four year old – I think he has been holding out on me. I wondered how he had so many Legos. The challenge? We’ve never had a policy from this particular carrier. Their FAQ site (a whopping four pages of minimal information) says it could have been because they process for other carriers, but nope, none of them either. So I set out to find out more information, particularly whether others in the family were affected, since we are all on the same policy (Mom, Dad and five small kids). I started my quest at 8:45am, on the website, and then the phone center opened at 9am. What a frustrating two hours. After talking to 11 people, from 4 different companies, do I know the answer to any of the questions? Not a single one. It all started with the vendor that the problem was outsourced too. I feel for those phone clerks, as they were provided almost no information. I then found a way to the carrier (a blog post in its own right), who didn’t know any more, but managed to transfer me to two other insurance companies, neither of whom had a clue why as they didn’t have a breach and I was never their customer. My concern is that this means they don’t even know what was stolen, where it was stolen, who’s information was stolen and more. If they don’t know that about me, what about you? I honestly don’t know how you protect yourself. You can’t really go off the grid. I could do without credit cards, and go to cash, but I can’t do without utilities or health insurance. I also understand that identity theft is big business, but the protections taken by major companies feel so lax. This is the FIFTH major breach of our family in less than 18 months. My credit card, from a major bank, has been replaced three times (only one breach was their own). So to the point of the post, for those still with me: If I was responsible for data security at any of these firms, I’d fire myself. There are solid, dependable companies doing security work. If you r company has not hired one to test your security, do it. Do it today. You should be doing penetration tests, at least annually. You should have solid company policies on data access, and that access should be extremely limited. People need information to do their jobs, but they don’t need all the information. Does your company have a data governance policy? If not, start today. We all know that IT budgets are limited and that our user communities, including our customers, want more and broader access. I just caution that you move with speed, but not without safeguards. Everything can be breached. Your firewalls, your apps, your website and even, as in the case of one breach, your cash registers. More important than all of this, though, is how you handle the breach when it occurs. Even with the most amazing safeguards, some pretty smart people, and governments, are hacking into private data. When it occurs, it should not be a shock to your company. You shouldn’t mobilize a task force after it happens. You should never consider this an IT problem – it is a major problem for the most senior levels of your company, and your reputation. Your company probably has, I hope, an IT Disaster Recovery (DR) plan. Does it include a data breach? Many don’t. They worry about floods, power outages, even pandemics, but not a data breach. Even if your DR plan does include data breach, are the actions your company will take fully laid out? If you are going to use a vendor, have they been chosen and briefed and is the conduit of key information already prepped? Is the spokesman for your company prepared and ready to speak publicly immediately? In my case, the time between the public announcement of the breach and the time we received the letter was over three months. Three months! Hopefully this post will cause at least one reader to start asking questions in their company and that those questions will be well received. You don’t want to be the next company in the news, do you?

Does Google Know Your Religion?

An industry contact recently told me that her phone popped up the following “creepy” message one Sunday morning: “9 minutes’ drive time to St. XXXX Church.” This, of course, was a predictable result of Google Now keeping track of where and when she regularly went, and nobly trying to help her get to her regular 8:30 a.m. Mass on time. What makes it creepy is that most people put religion (and politics? and sex?) on a mental “Off Limits” list. Deeply personal issues like these are risky fodder at cocktail parties, and equally risky subjects for automated, data-driven insights. An application like Google Now doesn’t understand the issue unless its human coders are prescient enough to realize there are some connections we humans simply don’t want our devices to make on our behalf. But now the string of unintended consequences has begun. My contact told her story in a room full of people. The consensus reaction around the table was, “Google keeps track of where you drive, and even figures out what’s there? That’s over the top.” What else, we wondered, would the app notice about us and dispassionately reflect back via pop-up message? Some bland examples: “You’re almost out of gin, and there’s a liquor store nearby…” “Your wife won’t be home for another hour, and Melrose Place is on channel 7 right now…” “You’ve been getting a lot of emails from XXXXX—perhaps you should ask him/her out?” The conversation took us quickly from ambivalence to unease. Others in that group probably mentioned their unease to their spouses and colleagues and friends. As I write, an ever-widening circle of people is developing reasons to be suspicious and uncomfortable about Google Now—which, by the way, is a perfectly excellent and useful app about 99 percent of the time! It’s a grand example of the problem we’ll face as we try to harness the power of big data. Years ago, when desktop publishing first became a technical reality for business users, a friend of mine who was a professional designer put a sign on her wall: “Power Ability.” She was telling people that just because they could publish their own office newsletters, crammed with cutsie clip-art, didn’t mean that they should. Someone might want to give Google the same sort of advice, especially when church locations are involved.

Experimenting with external data: What's the real motivation?

Recently, we held a UK CIO Roundtable on the myths versus realities of ‘Big Data’.  The roundtable was made up of a mix of insurers from both the P&C and Life industry. I guess that it will come as no surprise that finding new ways to extract insight from data is a hot topic.  In a low interest rate environment, improving results from the core disciplines of sales, underwriting and expense management take on a greater level of importance within many firms, and data is at the heart of this. For many firms, the business drivers for investment in data and analytics have not really changed in recent years, i.e.
  • Risk mitigation – Minimising the frequency and impact of losses, through fraud detection and loss scenarios.
  • Growth / Maintaining market position – Identifying profitable segments and understanding the propensity to take a particular course of action.
  • Improving service – Identifying ways to improve the quality of service through ensuring that the right data is in the right place at the right time to aid decision making.
What has changed, however, is the increased focus on using data to make a difference to business performance coupled with a growing interest in the search for and use of new alternative external sources of data to augment with existing internal data, such as social profiles, health app data, public records, etc. Unsurprisingly, many of the firms represented stated that they were active in experimentation with new external data sources, albeit on a small scale.  What I personally found fascinating was when one insurer, having been questioned about the motivation for experimentation in new sources of data, gave the following refreshingly honest answers:
  • Competitive threat – Concern that a competitor or start-up could use data in a way that threatens their current position, rendering the way they underwrite and service customers obsolete.  Telematics in auto-insurance and use of public health records & personal health tracker apps for enhanced annuities being two examples.
  • Regulatory threat – Concern that local market regulators may extend anti-discrimination laws to prevent the use of existing rating factors used for pricing risk, such as age.  The origins for this concern were triggered by the European Union’s decision to implement the Gender Directive that came into law at the end of last year. (For non-EU citizens, the Gender Directive prevents insurers from using gender for pricing within Europe.  Since its implementation, there has been an increased interest in telematics across the region as insurers look to use the data generated to predict driving behaviour).
When I reflect on this further, the insurer’s stance makes perfect sense to me. For years, established players have relied on their scale and the history captured within their systems to provide them with a competitive edge.  All of the time that market rules of engagement remain unchanged, the insurer is better off prioritising investment towards making use of the data it already has, rather than place a big bet on a risky source of new data or a new data led propositions for unproven markets. However, in maintaining a capability to search for and then experiment with these new sources of data, this insurer has enabled themselves to be better placed to respond when the threat finally occurs.  This is a fast follower strategy (or ‘Fast Second’ to use a phrase coined by Constantinos C. Markides in his book of the same name).  For incumbent players in service industries, this strategy can often prove to be highly successful as a response to a disruptive play by a new entrant. Perhaps today more than in recent history, it could be argued that new entrants and small agile competitors using alternative sources of data to gain a competitive advantage are becoming a real threat to established insurers.  We have seen early examples within the industry already, such as with the Climate Corporation’s entry into the crop insurance market (and then its subsequent acquisition by Monsanto for an eye watering ~$1B earlier this month), the positioning of personal health apps to assist in managing risk for health insurance, and the growth of telematics for ‘pay how you drive’ insurance models. Once these new non-traditional data sources and data-led propositions start to gain real market traction, the key to success for industry incumbents may not lie in being first but instead in first being aware followed by being the fastest to execute.

New Year, New Tools to Service Insurance Customers

For those interested in how new data techniques and availability are changing business models, I can recommend the article Smarter Information, Smarter Consumers in the latest edition of Harvard Business Review. The central premise is that legislation in the U.S. and U.K. now requires government agencies to make public data available and consumable in electronic form. This enables new techniques that leverage this information and provides increased value by making the purchasing process more intelligent. The authors offer their concept of “Choice Engines” – on line tools that guide consumers to make better purchasing decisions more efficiently using public information. At some point in the future, they also predict that private data will be added to the mix and allow the engines to work at a personal, individual level. Most of the use cases are consumer product-oriented, but as this blog has described previously, customer service expectations in other industries will influence insurance purchasing. The person who benefits because their cell phone company suggests ways to lower her bill (the authors’ example) will also want the same service from her insurance agent/company. Consumers and businesses will expect to be contacted by their agent/insurer when their risk profile changes. For example, if an addition is added to a house, insureds will expect that their insurance will be monitoring building permits and will want to be contacted proactively so their insurance can be adjusted appropriately. Two questions specifically related to insurance deal with timing and distribution models. Which insurance company will be the first to employ a choice engine for its insureds and prospects? Can an insurer with a mediocre data infrastructure and skill base compete with those which invest early and heavily in data techniques? Will independent agents embrace choice engines as an enabler, or reject them as a threat to their value proposition? Would an insurer be willing to offer such a tool to a distribution force that they don’t control? There is no question that managing risk will move from a point-in-time (usually renewal) event to a more continuous process. What is to be seen is which company changes the insurance purchasing model and transforms the buying process by using a tool like a choice engine.

Toronto Peer Networking Event

Celent held its 10th Peer Networking Event in Toronto last week and met with Canadian-based carriers to discuss insurance innovation and trends in big data. The purpose of the discussions was to identify practical ways to make progress on these topics in the face of increasing demands from the day-to-day business. The group reviewed the Celent Innovation Model and then broke into subgroups for a hands-on exercise. They used the model to analyze a set of strategic projects. The group agreed that using a model to analyze the disruptive potential of projects lends a valuable perspective to the standard project review process. There was also a broad recognition that funding and implementing truly innovative projects requires significant alignment between IT and business. One participant observed that senior leadership’s challenge when sponsoring innovative and disruptive initiatives is to be ready for quick failures and recalibration. The host CIO then led a discussion on current challenges and opportunities in the Canadian market. A consensus view was that past allegiance of a policyholder to their company that existed five years ago is largely gone today. Insurers cannot take for granted their customers will renew. Thus, it is becoming increasingly important to engage and reach the customer in various ways and with new products. For example, telematics is receiving a lot of attention in the Canadian market and usage based insurance is likely to be standard very shortly. In the afternoon, Ben Moreland presented a framework in which to understand Big Data and also gave the group insight into the vendor market serving this area. His prediction for the market development of big data products is that they will follow a similar path to portals. Initially, large system providers such as IBM and BEA delivered platforms on which insurers could build portals. As the market matured, purpose-built solutions emerged that contain deep functionality built in. The final session was led by a business architect from one of the leading banks in Canada. He shared an analysis tool developed by his company that used customer experience mapping techniques to trace the journey that customer data takes through the various systems used in their insurance area. As a result of this approach, insights and opportunities for consolidation were discovered that would have remained hidden using traditional data analysis techniques. As with the all of other Peer Networking Events, the feedback from participants was that this format allows for open and active exchange of ideas between insurance company technology professionals. If you are interested in attending or hosting an event, please contact for more information.