Tag Archives: John Johansen

Data Science: Methods Matter (Part 4)

Putting a data science solution into production after weeks or months of hard work is undoubtedly the most fun and satisfying part. Models do not exist for their own sakes; they exist to make a positive change in the business. Models that are not in production have not realized their true value. Putting models into production involves not only testing and implementation, but also a plan for monitoring and updating the analytics as time goes on. We’ll walk through these in a moment and see how the methods we employ will allow us to get the maximum benefit from our investment of time and effort.

First, let’s review briefly where we’ve been. In Part 1 of our series on Data Science Methods, we discussed CRISP-DM, a data project methodology that is now in common use across industries. We looked at the reasons insurers pursue data science at the first step, project design. In Part 2, we looked at building a data set and exploratory data analysis. In Part 3, we covered what is involved in building a solution, including setting up the data in the right way to validate the solution.

Now, we are ready for the launch phase. Just like NASA, data scientists need green lights across the board, only launching when they are perfectly ready and when they have addressed virtually every concern.

See also: The Science (and Art) of Data, Part 2  

Test and Implement

Once an analytic model has been built and shown to perform well in the lab, it’s time to deploy it into the wild: a real live production environment. Many companies are hesitant to simply flip a switch to move their business processes from one approach to a new one. They prefer to take a more cautious approach and implement a solution in steps or phases. Often, they choose to use either an A/B test and control approach or a phased geographic deployment. In an A/B test approach, the business results of the new analytic solution are compared with the solution that has been used in the past. For example, 50% of the leads in a marketing campaign are allocated to the new approach while 50% are allocated to the old approach, randomly. If the results from the new solution are superior, then it is fully implemented and the old solution removed. Or, if results in one region of the country look promising, then the solution can be rolled out nationwide.

Depending on the computing platform, the code base of the analytic solution may be automatically dropped into existing business processes. Scores may be generated live or in batch, depending on the need. Marketing, for instance, would be a good candidate to receive batch processed results. The data project may have been designed to pre-select good candidates for insurance who are also likely respondents. The results would return an entire prospect group within the data pool.

Live results meet a completely different set of objectives. Giving a broker a real-time indication of our appetite to quote a particular piece of business would be a common use of real-time scoring.

Sometimes, to move a model to production, there’s some coding that needs to happen. This occurs when a model is built and proven in R, but the deployed version of the model has to be implemented in C for performance or platform considerations. The code has to be translated into the new language. Checks must be performed to confirm that variables, final scores and the passing of correct values to end-users are all correct.

 Monitor and Update

Some data projects are “one time only.” Once the data has appeared to answer the question, then business strategies can be addressed that will support that answer. Others, however, are designed for long-term use and re-use. These can be very valuable over their periods of use, but special considerations must be taken into account when the plan is to reuse the analytic components of a data project. If a model starts to change over time, you want to manage that change as it happens. Monitoring and updating will help the project hold its business value, as opposed to letting its value decrease as variables and circumstances change. Effective monitoring is insurance for data science models.

For example, a model designed to repeatedly identify “good” candidates for a particular life product may give excellent results at the outset. As the economy changes, or demographics change, credit scoring may exclude good candidates. As health data exchanges improve, new data streams may be better indicators of overall health. Algorithms or data sets may need to be adapted. Minor tweaks may be needed or a whole new project may prove to be the best option if business conditions have drastically changed. Monitoring the intended business results compared with results at the outset and results over time will allow insurers to identify analysis features that no longer provide the most valid results.

See also: Competing in an Age of Data Symmetry

Monitoring is important enough that it goes beyond running periodic reports and having hunches that the models have not lost effectiveness. Monitoring needs its own plan. How often will report(s) run? What are the criteria we can use to validate that the model is still working? Which indicators will tell us that the model is beginning to fail? These criteria are identified by both the data scientists and the business users who are in touch with the business strategy. Depending on the project and previous experience, data scientists may even know intuitively which components within the method are likely to slide out of balance. They can create criteria to monitor those areas more closely.

Updating the model breathes new life into the original work. Depending on what may be happening to the overall solution, the data scientist will know whether a small tweak to a formula is called for or an entirely new solution needs to be built based on new data models. An update saves as much of the original time investment as possible without jeopardizing the results.

Though the methodology may seem complicated, and there seem to be many steps, the results are what matter. Insurance data science continually fuels the business with answers of competitive and operational value. It captures accurate images of reality and allows users to make the best decisions. As data streams grow in availability and use, insurance data science will be poised to make the most of them.

Competing in an Age of Data Symmetry (Pt. 3)

The Internet is a mirror of sorts — a data mirror. Right now, it is a sort of fuzzy data mirror, but the pictures grow clearer as the available data grows. Soon, the image of an insurers’ customer service, pricing and claims experiences will grow crisp. How will it happen? How will insurers respond and remain competitive?

In Part 1 and Part 2 of our series, we discussed data symmetry — the leveling of the playing field that is currently happening because insurers are gaining access to many of the same streams of data. The trend runs in contrast to data asymmetry, which allowed insurers to comfortably differentiate themselves by being good at the analysis of their own in-house data. As insurers use more and more of the same data and some of the same analytics tools and methodologies, they will find themselves in a pool of sameness. Differentiation by price and service will be less about introspective analysis and more about finding and delivering on real brand promises.

So, in today’s blog we are crossing a bridge of sorts. We are going to look at how the consumer will achieve data symmetry by gaining a clear view of the real insurer.

See also: Data Science: Methods Matter

Changes in scrutiny are causing data symmetry

Insurers are the subjects of constant scrutiny. The NAIC, the Federal Insurance Office, the Department of Labor, every state and every consumer protection organization have an interest in watching insurers. Yet all of that scrutiny may pale in comparison to the impact of the coming wave of individual consumer scrutiny.

Consumers are using ratings, stars, comments and shopping patterns to give instant feedback to all service providers. Feedback (real experience) is a sales tool for aggregators and retailers. It is a reason for consumers to choose particular channels or pipelines. Amazon and eBay don’t have to build trust for any one product. They only have to facilitate feedback and let the products, services and suppliers speak for themselves.

These outside views are the result of symmetrical data availability. Prospects are now able to compare any product or service, including insurance, with greater real data, including both sources that are verifiable and those that contain unstructured data. Consumers may look at an insurer through the lens of an insurance aggregator, such as Insure.com or The Zebra, or through simple search terms such as “worst auto claims experience in my entire life.” They may also witness an insurance interaction through their relationships with friends on social media.

Reputation analysis will hold tremendous power to validate or invalidate brand promises. Does the insurer make it simple to file a claim? Does it have a poor track record in paying claims? Are renewal rates much higher or lower than competitors’? These bits of information weren’t as public in the past. Today, they are common and easy to find.

See also: What Comes After Big Data?

Data symmetry’s effect on the insurer will operate much like a looking glass. The insurer will begin to see itself, not as it has attempted to portray its brand, but as it is perceived during real interactions. This will lead some insurers to make course corrections.

The good news is that data symmetry will supply healthy doses of reality. Insurers will know and understand their competition. They will have an unprecedented, timely idea about what customers really want and how well they are supplying it. If they are prepared for the coming levels of data symmetry, insurers will also be able to make agile shifts and meaningful steps toward selling insurance through many different channels. Many of these details are still food for our insurance visions. One thing is certain, however. Data and analytics will continue to unlock the secrets of market positioning to keep insurers competitive. Data’s relevance to business decisions will always grow.

The Insurance Renaissance (Part 1)

It was in 14th century Florence that an epic awakening happened. It was all-pervasive. It wasn’t just art that began to thrive. Philosophy, economics, culture and science began rapid change, too.

Education, technology and literature were thrown into a cauldron of modernization, and world-shaking disruption and advancements spread rapidly. Fast-forward to today, and the comparison is striking. As we enter a new era of disruption and change underpinned by new technologies, business models and more (see Future Trends: A Seismic Shift Underway), the past offers an opportunity to guide and inform our future. The insights may help us see opportunities from a new perspective.

We’ve identified dozens of parallels and lessons from the Renaissance that can give insurers and technology experts food for thought as they prepare for this journey. Over the coming months, we will be taking a look at the renaissance unfolding in insurance and reflecting on when the world was shifting from the “dark ages” to a future of opportunities, possibilities and enlightenment.

See Also: The Five Charts on Insurance Disruption

As a preview, consider the original Renaissance and these modern parallels to today’s:

Focus on People

In the dark ages, individuals didn’t matter on the level they did during the Renaissance, when individual thought was cultivated and education encouraged. People gained freedoms to act and create in ways they hadn’t thought of before.

Today, technology and connectivity have brought a new level of individualism to insurance. We are moving from mass standardization to hyper-personalization for everything from marketing to product pricing. The individual voice, rather than groups, matters more than ever.

Universal Access

The Renaissance changed communication. Moveable type and printing allowed communication to be more widely disseminated to the larger population.

In today’s world, we can reach nearly everyone, any time, anywhere and in any way.  Transactions take place on mobile devices. Social media has made it so customer thoughts and decisions are not hampered by distance or hours of operation. The enlightened and prepared insurer has nothing standing between it and its customer. “Trade” has blossomed in the city of the internet.

Reality and Empiricism

Art and science in the Renaissance shared a trait — the drive for a “real view.” Artists approached painting and sculpture from the standpoint of realism, while science began to revisit the idea of research and empirical evidence.

In our era, at least for the last several decades, insurers have also attempted to operate from a standpoint of mathematical certainty — pricing products based on historical data trends. Yet the digital era is bringing with it an entirely new set of real-time data streams and, with those, real-time, personalized analysis, pricing and decision-making. Our new perspectives will soon allow us to see into individual lives and habits with striking clarity. (For more on this, see John Johansen’s blog series on Data Symmetry.) We will enjoy enhanced levels of insight to engage and service customers as never before.  And all of this reality will be brought to us with dramatic speed. Insurers that are prepared with the agility to consume and analyze data in real time will have the advantage over their competitors to better serve their customers.

Money-Driven Innovation

The Renaissance didn’t happen overnight. It was spurred by a convergence of factors, the greatest of which was increased wealth. Trade in Florence had produced a new class of financier who was willing to fund artistic and scientific endeavors. Wealth created ease; ease allowed time for thought and innovation.

Our modern businesses are also the beneficiaries of affluence. Population and economic growth have created a culture where even many of the economically disadvantaged have access to digital and mobile technologies. Those technologies provide online access to insurance to protect them against a growing array of risks.

Likewise, investments in the 20th century helped insurers become more efficient and more accessible, fueling an improved product and service landscape within the traditional insurance business model.

Today’s renaissance, however, is moving well beyond the traditional model. Significant capital investment in new insurance greenfield or start-up companies is fueling massive innovation in products, services and business models. For reference, simply consult the CB Insights Periodic Table of Insurance TechCB Insights has indicated that Q1 2016 has already topped the record for most early-stage insurance tech deal activity (Seed/Series A). This includes two start-ups: a peer-to-peer insurance company, Lemonade, and a small business insurance start-up, Next Insurance. Interest and investment is also expanding beyond venture investors to carriers and reinsurers such as Guardian Life’s GIS Strategic Ventures investment in health benefits startup, Maxwell Health and many more. For the 130-plus start-ups and private companies in the insurance tech space, CB Insights indicates that more than $3.5 billion in aggregate funding has been raised.  Money is the seed and the fuel for the massive innovation taking hold in insurance.

In the coming weeks, we will dig deeper into the details of the insurance renaissance. We will uncover some of the philosophy behind modernization, while also thinking about the practical aspects of improved operations, digital capabilities and customer service. In each case, we’ll be keeping our focus on agility, innovation and speed so that we won’t just be learning about the renaissance, but we’ll be living out its lessons within our organizations.

Competing in an Age of Data Symmetry: Part 2

In 1983, Microsoft Word was introduced. It wasn’t the first word processor, and it isn’t the only word processor, but it quickly became a standard — a “given.” From a productivity standpoint, the first adopters of word processing certainly had advantages over the alternatives (typewriters and ball point pens). Today, however, we all use Word and Excel and Outlook. Your only advantage may be in how quickly you can type or speak into your device.

This is not unlike what is happening in the world of data. Data availability has become ubiquitous. Not only has data become freely available, but data analysis through tools, consolidators and rating companies has become freely available, as well. When everyone has access to the same information at the same time, that’s data symmetry.

In Part 1 of our data symmetry blog, we followed the quickly shifting trends from asymmetrical data availability to a market filled with data symmetry. We illustrated how data symmetry is rapidly changing the idea of competitive advantage. If everyone has access to the same data and if digital technologies are increasing the number of data sources, an organization’s proprietary data will lose the ability to keep the company ahead.

Data symmetry will then throw established insurers into a mid-life crisis, with everyone from marketing to underwriting to claims asking, “What makes our insurance actually different?”

Once insurers are operating from the same data, and the prediction of symmetrically available data has become a full-blown reality, then data will no longer be a differentiator, and something else will be.  But what?

The good news is that there will always be a way to create advantage if insurers remain active in fostering their uniqueness. From a data standpoint, here are three differentiators for your organization to consider:

Moving from individual histories to virtualized views

The data that is contained in today’s individual history will pale in comparison with tomorrow’s virtual record. In the very near future, everyone will take advantage of virtualized views of complete individuals or commercial accounts.

These will includ/.e every facet of someone’s lifestyle, health history, safety records, common travel patterns, activity levels and even purchase histories. Virtual individuals will be known and understood in ways that real individuals may not even know themselves. We already see this happening in online sales of music, books, movies, coffee, auto parts and tennis shoes. Where there is a purchase, there is a preference. Purchase patterns are allowing digital retailers to accurately predict which marketing messages will work with hyper-targeted methods. Modern insurers will use these same automations and data analysis to improve timing, not only for marketing but also for claims prevention. As virtual individual data interacts with external sources, such as geographic and weather data, the insurers who have been practicing their data science will become predictive pros. Predictive analytics will still allow some competitive asymmetry to exist.

Think of data streams as colors in a box of paints — the more colors one finds in the box, the clearer the picture that an insurer will be able to be paint. Data analytics experience will be the art classes that will make some insurers capable of predictive masterpieces. The old colors will still be in the box. Claims histories and proprietary risk models will still be available, but they will sustain their value when they are supplemented with fresh colors and new data perspectives.

Innovating around products and services

Predicting results and preventing claims will support business in the current realm of insurance. Both are still subject, however, to data symmetry. Data symmetry will, in turn, push insurers to innovate. What will be striking to see is how often these front-end innovations of all types will enhance back-end data capabilities.

Early in 2016, for example, Liberty Mutual and Subaru announced a partnership that will bring usage-based insurance into Subaru’s connected car platform. Usage-based insurance is one of the clearest examples of innovative products, fueled by data that will also improve data analytics. This involves a new measure of innovation — how quickly data can move from collection to analysis. The quicker an insurer can transfer data gathering into meaningful action, the more valuable the innovation. Companies will be asking what levels of automation can be employed to turn prediction into prevention.

They will also be looking for formulas that make innovative products or services attractive to consumers. Data innovations aren’t instantly palatable to people. In-car telematics devices are a great example. The initial innovation was somewhat offset by the expense of installation and the perception that an installed device invaded consumer privacy.

Most efforts at product innovation will make consumer incentives part of the formula. As insurers turn “free data” into better ratios between pricing and risk, both the insurer and the consumer will need to see the clear benefits. Residential insurance is an example of an area ripe for innovation. Home insurance premiums are most often paid within the house payment. Most homeowners would be thrilled to have their house payment go down $100 to $200 per month. Property insurers that can take advantage of home sensor data and Internet of Things data could make that happen.

In exchange for the savings, many homeowners would sign off on the idea that their insurer now has monitoring capability. Property insurers would then be adding home data to their available data streams. This could give carriers a competitive difference. Lender/insurer partnerships (additional product innovation) may also arise with greater frequency if lenders can find corollary trends between home monitor data and clients with the fewest incidents.

This same data/pricing correlation will apply to commercial insurance. If the use of drones, security system monitoring and environmental system monitoring will result in lower insurance costs, most companies will see the value in an insurer that is looking out for their bottom line.

Insurers will find some of their differentiation in data-driven, value-added services. Anywhere that data can point to a better practice, an insurer will want to promote that to customers. Whether that means suggesting alternative travel routes for trucking companies or promoting add-on products for specialized risk, the influence of data symmetry can be overcome with creativity and innovative thinking.

Focusing on the stars

When we discuss data, our mindset traditionally envisions incoming data. Customer experience data, however, is much more of a two-way data street. Consumers are painting a new world of service with their ratings and stars. These outside views are also subject to data symmetry. Prospects are now able to efficiently compare insurers with real service data, including both sources that are verifiable and those that contain unstructured, conjectural data.

In Competing in an Age of Data Symmetry, Part 3, we’ll look at what an insurer should be doing to prepare itself for greater customer scrutiny and how reputation analysis will validate or invalidate an organization’s brand promises.

data symmetry

Competing in an Age of Data Symmetry

For centuries, people have lived in a world where data was largely proprietary, creating asymmetry. Some had it. Others did not. Information was a currency. Some organizations held it, and profited from it. We are now entering an era of tremendous data balance — a period of data symmetry that will rewrite how companies differentiate themselves.

The factors that move the world toward data symmetry are time, markets, investment and disruption.

Consider maps and the data they contained. Not long ago, paper maps, travel books and documentaries offered the very best views of geographic locations. Today, Google allows us to cruise nearly any street in America and get a 360° view of homes, businesses and scenery. Electronic devices guide us along the roadways and calculate our ETA. A long-established map company such as Rand McNally now has to compete with GPS up-and-comers, selling “simple apps” with the same information. They all have access to the same data. When it comes to the symmetry of geographic data, the Earth is once again flat.

Data symmetry is rewriting business rules across industries and markets every day. Insurance is just one industry where it is on the rise. For insurers to overcome the new equality of data access, they will need to understand both how data is becoming symmetrical and how they can re-envision their uniqueness in the market.

It will be helpful to first understand how data is moving from asymmetrical to symmetrical.

Let’s use claims as an example. Until now, the insurer’s best claims data was found in its own stockpile of claims history and demographics. An insurer that was adept at managing this data and applied actuarial science would find itself in a better position to assess risk. Competitively, it could rise to the top of the pack by pricing appropriately and acquiring appropriately.

Today, all of that information is still very relevant. However, in the absence of that information, an insurer could also rely upon a flood of data streams coming from other sources. Risk assessment is no longer confined to historical data, nor is it confined to answers to questions and personal reports. Risk data can be found in areas as simple as cell phone location data — an example of digital exhaust.

Digital exhaust as a source of symmetry

Digital exhaust is the data trail that all of us leave on the digital landscape. Recently, the New York City Housing Authority wished to determine if the “named” renter was the one actually living in a rent-controlled apartment. A search of cell phone tower location records, cross-referenced to a renter’s information, was able to establish the validity of renter occupation. That is just one example of digital exhaust data being used as a verification tool.

Another example can be found in Google’s Waze app. Because I use Waze, Google now holds my complete driving history — a telematics treasure trove of locations, habits, schedules and preferences. The permissions language allows Waze to access my calendars and contacts. With all of this, in conjunction with other Google data sets, Google can create a fairly complete picture of me. This, too, is digital exhaust. As auto insurers are proving each day, cell phone data may be more informative to proper pricing than previous claims history. How long is it until auto insurers begin to look at location risk, such as too much time spent in a bar or frequent driving through high-crime ZIP codes? If ZIP codes matter for where a car is parked each night, why wouldn’t they matter for where it spends the day?

Data aggregators as a source of symmetry

In addition to digital exhaust, data aggregators and scoring are also flattening the market and bringing data symmetry to markets. Mortgage lenders are a good example from outside the industry. Most mortgage lenders pay far more attention to comprehensive credit scores than an individual’s performance within their own lending operation. The outside data matters more than the inside data, because the outside data gives a more complete picture of the risk, compiled from a greater number of sources.

Within insurance, we can find a dozen or more ways that data acquisition, consolidation and scoring is bringing data symmetry to the industry. Quest Diagnostics supplies scored medical histories and pharmaceutical data to life insurers — any of whom wish to pay for it. RMS, AIR Worldwide, EQECAT and others turn meteorological and geographical data into shared risk models for P&C insurers.

That kind of data transformation can happen in nearly any stream of data. Motor vehicle records are scored by several agencies. Health data streams could also be scored for life and health insurers. Combined scores could be automatically evaluated and placed into overall scores. Insurers could simply dial up or dial down their acceptance based on their risk tolerance and pricing. Data doesn’t seem to stay hidden. It has value. It wants to be collected, sold and used.

Consider all the data sources I will soon be able to tap into without asking any questions. (This assumes I have permissions, and barring changes in regulation.)

  • Real-time driving behavior.
  • Travel information.
  • Retail purchases and preferences.
  • Mobile statistics.
  • Exercise or motion metrics.
  • Household or company (internal) data coming from connected devices.
  • Household or company (external) data coming from geographic databases.

These data doors, once opened, will be opened for all. They are opening on personal lines first, but they will open on commercial lines, as well.

Now that we have established that data symmetry is real, and we see how it will place pressure upon insurers, it makes sense to look at how insurers will use data and other devices to differentiate themselves. In Part 2 of this blog, we’ll look at how this shift in data symmetry is forcing insurers to ask new questions. Are there ways they can expand their use of current data? Are there additional data streams that may be untapped? What does the organization have or do that is unique? The goal is for insurers to innovate around areas of differentiation. This will help them rise above the symmetry, embracing data’s availability to re-envision their uniqueness.