Hyland TL August 2017
Follow Us
ITA MEMBERSHIP

DATA

DATA

Data Governance Breaks It All Down for Insurers

Robert Regis Hyle | April 20, 2015

There is great value to be gained from a data, as Millers Mutual Group discovered, but as the insurer learned, attempts to leverage their data always came back to the issue of granularity. Without it, trying to do predictive analytics was just too difficult.

Today, Millers nears the completion of new core systems, so the company stepped back and decided to take a look at its data in a new light. As part of that process, the Pennsylvania-based carrier implemented a data governance process that will eventually lead to an upgrade of their data warehouse so more can be accomplished with analytics and modeling.

“We learned a lot in the last six years about what is working and what isn't, so now we're making some changes,” says Larry Fortin, CIO at Millers Mutual.

Good work was done initially with dashboards, but business users knew that when they ran a report from one system and then a report from another system, the same data elements often had different values.

“We did the research on loss ratio leveraging governance to make those definitions widely known and public,” says Fortin. “We worked with the business so they understood what that meant and then documented it so when people looked at something they could go back to the definitions and learn why the reports were different.”

The challenge with data governance that many companies struggle with is to understand it is specific to the culture of the company, not the industry. Data governance is not a set of tools, but rather a discipline that companies have to implement. And it doesn’t have to look or act the same as any other governance plan, explains Svetlana Sicular, a research analyst with Gartner.

“If you make a generic framework and frame it around a company that is centralized vs. one that is de-centralized, governance is totally different,” she says. “Some companies might have a culture of making quick decisions; another would be more consensus-based and wouldn't move forward without that consensus. This defines a fashion in which you conduct data governance. In my mind, it's not so much about the industry as it is about a company culture.”

Craig Beattie, a senior analyst with Celent, believes most insurers have a data governance program in place and credits that to the fact there is more focus on data today than ever before. There also are more data products being sold to insurers and that gives some carriers the thought that if brokers, distributors, reinsurers and others parties can package data this well, maybe insurance carriers can do the same.

“So the correct answer is all insurers are doing something with data governance,” says Beattie. “A different answer, also correct, is insurers have a way to go to get it right and, to be honest, the data is holding them back.”

Business Units Lead

It was not difficult to get the business side involved with the data governance initiative at Millers, according to Fortin. The carrier went through a strategic planning exercise to look at the next three years and felt some of the data the users wanted to understand needed updates.

“The business said they need a better understanding,” says Fortin. “So now we have the core systems coming online, and over next three years a large part of our effort will be around data, governance, re-establishing the data, and leveraging it for the future. It's been a pull by the business because they want to leverage the data and they recognize some of the past inefficiencies.”

Millers has a data governance committee with membership supplied from across the organization. The vice president of operations heads the committee and, as CIO, Fortin lends an IT perspective, but it is the business side that drives the direction.

“IT plays a key role as far as process procedure, documenting points, and establishing a framework, but the business knows where they want to go and we have some good members that know how they can leverage the data,” says Fortin. “It's been good that IT is participating in the discussions, but not shouldering the entire plan.”

Most successful companies have data governance with the business units with an IT representative as the enabler. Sicular fells this is where the emerging role of a chief data officer (CDO) takes precedence, something she is seeing more of in the insurance industry.

“Ideally, governance would be the office of the CDO and that person would work with various stakeholders who represent different views of data and decide what to govern, how to govern, and how to enable the governance program based on the input and needs of the stakeholders. It all depends on what people govern,” she says.

Data governance is also distributed throughout the organization, particularly in companies that are organized by lines of business, according to Beattie.

“They own the data in their lines,” he says. “There are a few organizations that have the data centralized, but it usually comes down to the organization and how they demand control, efficiency, centralized processes, agility, and dedicating it. There is no correlation between good and bad, it's just different and comes down to the values of the organization.”

Beattie believes insurance CIOs should be chief innovation officers. As long as all the systems are running without incident, the conversation within IT should be more about innovation, which applies to data as well.

“There is a role within IT to educate the rest of the organization on what data can do,” he says. “What you tend to find is the people good with stats tend to be in the actuarial department. There is a need for a cross-functional role, particularly leveraging data. The business people are calling out to IT to help them understand what they can do with advanced analytics. There is a role in IT to help bring some of this technology into the organization.”

Big Data, Bigger Problems

Millers has the mindset that the company needs to learn to walk with data before starting to run. Fortin maintains the company can gain more insight from the data thanks to the new core systems, which will enhance the operations.

“We definitely know where we can leverage Big Data in the future. The business as well as IT recognizes we have to clean up some of what’s in place and re-establish the base platform before we can take advantage of Big Data,” says Fortin. “We can leverage data from our different systems and where there are a lot of notes we can pull into various algorithms to study the patterns. It's there, but we have to continue to re-establish the baseline.”

Big Data introduces new challenges to governance and democratizes perspectives on the data, such as the ability to access the data by more people. Some data stewards are ungoverned and the reason is new technology, which is under the radar of big enterprises, according to Sicular.

“When we hear of successes that are achieved on their ungoverned data, we say the new technology is much better than relational technology,” says Sicular. “In some cases it can get you to the answer faster, but often, because it is ungoverned, people can't get to all the data. In the brave new world of digital enterprise the question is what to govern and what not to govern. Not to govern should be a conscious decision because of what was implemented by shadow IT groups.”

There also are systems of innovation that make companies stand out and outpace the competition. There has to be a different approach of governance with these companies and it resonates with virtually all of Gartner’s clients.

“Sometimes when you speak of data governance, many organizations don't want to use the term ‘governance.’ They prefer to use a newer team, which has gotten a lot of success with our customers: ‘data advocacy.’ How do you take care of that data to be used to a wider extent throughout the enterprise?” asks Sicular.

Another issue with Big Data is the shift from truth to trust. Traditionally governance programs start with data quality. That's important, points out Sicular, because companies on the low level of maturity don't understand the value of data quality.

Advanced companies go with metadata management, continues Sicular. Those initiatives are under the information governance purview and usually start with a particular program, compliance for example, and then move on.

“It depends on the company, but the struggle is to be recognized. I tell clients not to call it governance because people get bored when they hear that word. Call it the guiding principles, which should be something found on the first two pages of your annual report, such as expanding to new markets,” she says. “Then I recommend companies create appropriate policies in a fashion that involves the users and make sure this includes data stewards. Approach it with a product in mind and the definition of the product needs to be satisfied.”

Governance has to satisfy the wants of key stakeholders because it may uncover greater needs for the company. On the higher levels of maturity, organizations need to appreciate what governance means and what happens is a shift from control to trust in terms of people having more trusting relationships with each other rather than subordination.

“When companies speak of big data in terms of trust, people are saying master data management is the single source of the truth,” says Sicular. We have this convention that it is, but it's not always right; it's just our convention. Truth is different for lines of business. In order not to step on others toes you need to get to this convention.

Many new data sources come from different places, including external data sources, so the questions become: Can you trust Big Data and who is responsible for it? Even if you have a reputable third-party provider, where does this provider get the data?

“Acquired data is often of abysmal quality,” says Sicular. “I can go on eBay and buy 10,000 Twitter followers for $10. If you analyze social media and look for influencers you will see I have 10,000 followers so you may decide to target me differently rather than as someone who spent 10 bucks for 10,000 followers.”

Governance Issues

Insurers typically look at data governance as a way to understand the data they have. There are universal definitions across the organization, they are reporting in the same way so the CEO doesn't get one report from actuaries and another from the CFO with different formulas on how it was accounted for, according to Jeff Goldberg, vice president of research and development for Novarica.

“I think data security tends to be solely the responsibility of IT and not necessarily a part of data governance,” says Goldberg. “Data governance is an understanding of how data is used, with a partnership with IT. The assumption is whatever data they bring in and use is going to be secure and that's ITs responsibility. The organization could probably do better with data security, but I think it is often pushed off to the security group within the IT organization.”

There are different approaches on how to make use of the data and that includes making sure it is managed properly. Goldberg doesn't believe there is a single approach yet and the best practices are not clear.

“In many ways our approach to data and data analytics is similar to the approach to IT 15-20 years ago where each business unit has their own data people—a federated approach—or sometimes a centralized approach with a team of data people and possibly a chief data officer,” says Goldberg. “It's not clear yet what the right approach is. Like IT I think we eventually will settle on a centralized model and organizations will have a CDO responsible for the use of the data. That won't make it easier to secure the data, but it will make one person responsible, which means if there is a data breach, they are the person whose head is going to roll so that person will have a responsibility to keep that from happening.”

A business unit aligned data approach is not necessarily the wrong approach, according to Goldberg. What is important is that best practices be shared among everyone.

“The fact is, business units aren't really clear what value they are going to get from data governance and they need a tighter integration to make that happen,” he says.

Different market forces have driven governance in recent years. For global insurers active in Europe, Solvency II pushed data to the forefront and had a significant effect on insurers, explains Beattie.

“Without that push they wouldn’t have found it can be beneficial to gain quality data that is more useful for them,” he says.

For other carriers, though, many have already been working on investments in governance programs, particularly insurance brokers

Sicular agrees that Big Data has been a driver for the implementation of data governance. To some extent, Big Data is the definition of the importance of data to the enterprise, not just the technology itself.

“Technology is often disparate and immature,” she says “The fact companies can make sense of data is a starting point to how they manage data, look at data as an asset, and look at what's important. Many companies in the financial industry, for example, have been doing governance for some time.”

Companies in regulated industries came to data governance with regulatory compliance in mind to see how they are exposed to risk. They look at the data on which they depend to make sure they mitigate or reduce risks or do something around compliance, such as anti-money laundering legislation, Basil II, and Sarbanes-Oxley.

There also is a maturity curve where companies progress from compliance toward delivering value, explains Sicular. Once they start taking care of the data they begin to gain a certain level of understanding of additional opportunities to deliver value to the organization, such as merging information siloes, doing additional analytics, creating new business models, or engaging new people.

“Governance is a decision-making framework,” she says. “You have principles, policies, standards, and best practices. There are people on the business side that are supposed to follow those best practices and contribute to them to make them actionable in a particular situation.”

Size of Company

Some smaller companies have demonstrated more success with data governance than larger carriers because it is easier for smaller carriers to implement a program. They also have a greater need for governance, explains Sicular.

“I worked with a company with 500 employees and they were successful with governance because they knew what the needed, where they were moving, and understood their goals,” she says. “Large companies have many moving parts and lines of business and feel they are OK without governance because it is often viewed as an artificial barrier. That's one reason why I suggest not using the term governance. I think it depends on how a company understands the role of data in the company, not the size of the company.”

The convergence of cloud, mobile, and Big Data created new questions as well, such as how to govern data that is outside the enterprise. Cloud allows companies to share data with others and sometimes that needs governance. Privacy and ethics are other important issues.

“Among my clients, there are different governance programs in place and when IT gets a call to start data governance, I tell them to look around and if you have something that resembles what you are planning to do, join your forces with them; don't do anything separate,” Sicular says. “Become a working group or an extension to extend responsibilities to include data and information governance.”

The leaders in the insurance field have an organized structure and are considering how new projects bring new data to the organization, what they should be doing about that, and the people are managing data as an asset rather than a by-product of a change program, according to Beattie.

“Ownership tends to be shared between the business and the IT department,” he says. All too often, we find many people consider data to be an IT problem, but actually data isn't just an IT problem, it's a business asset and requires business sponsorship. Leaders in this field have senior level sponsorship of data, leverage it as an asset, and are enriching it with other sources of data. That level of visibility within the org is probably the key difference.”

With midtier and small carriers that haven't given data much focus to governance, Beattie believes you tend to find pockets of data ownership.

“There are places within those organizations where data has done well, but it is not replicated throughout the enterprise. Senior level sponsorship and consistency across the organization is what makes leaders.”

The Future

Discussions are already taking place about multi-speed IT and rigorous IT within the insurance industry, explains Beattie. Rigorous IT delivers regulatory changes and changes in the core system. Fast IT involves keeping up with digital trends, responding to the customer, and being more agile.

“That's something we'll see playing out, says Beattie. “There are things you absolutely have to get right and there are a lot of stakeholders to ensure that everything is correct, but that slows things down. Stakeholders need to agree on the numbers in the reports.”

Even with good business data, some questions, such as why one business unit is performing better than another in the same state, can take several months to adequately answer.

“There is a bucket of data that can be queried quickly and this gives the business incredible agility and allows for data decisions as opposed to anecdotal or gut feelings,” he says. “This is a game changer. You have data that has to be absolutely right and data that is good enough to answer questions and provide fast answers. That will lead to a change in agility while still remaining compliant.”

Beattie explains that even with data that companies know is not absolutely correct, it can still provide useful information. Carriers just need to be able to deal with it differently.

“It is easy to say people don't understand the data and keep questioning why answers are different, but we are moving past that to a better understanding of where it comes from, how accurate it is, and what kinds of things it can be used for,” says Beattie. “It's a challenge for any organization. They are still grappling with data governance and data quality so they have to get slow data right and the regulatory issues correct. Once they get past that they are getting to a space where they can start thinking about fast data, answers to questions, and enable business to be as agile as they can be.”

The data governance committee at Millers Mutual was established well ahead of when the new systems will be ready in mid-2015. At that point, the committee will be more active.

“There is pent-up demand in regards to corporate and operational reports, more dashboards, and more scorecards,” says Fortin. “Once we get the core systems producing the data we can see how we can leverage it. In the past we may not have cared about some elements, but we are looking at having the data produced and captured right away because we know we are going to need it down the road.”

 

 


Featured articles

MT TL

Guidewire MR

ELECTRONIC CHAT

The Email Chat is a regular feature of the ITA Pro magazine and website. We send a series of questions to an insurance IT leader in search of thought-provoking responses on important issues facing the insurance industry.

WEB EVENTS

ITA is pleased to present the 2014 Webinar Series. We have many topics for you to choose from and attendance is open to all ITA members. The webinar topics are current and exciting — ranging from predictive analytics to telematics and will focus on the direction insurance carriers need to follow for the future. All webinars are presented by insurance IT professionals along with some of the leading analysts and consultants in the field. There is no cost to attend an ITA webinar. For more information and to register for the webinar, click the “title” of the webinar below.

BLOGS AND COLUMNS

only online

Only Online Archive

ITA Pro Buyers' Guide

Vendor Views

Partner News