After decades of gathering and warehousing data on insurance purchases, exposures and claims from their commercial clients, more insurance brokers are offering to share the power of their “big data” sets with their middle-market clients.

Whether it's to compare their insurance purchases with that of their peers or to determine the likelihood and effect of a hack attack on vulnerable information technology systems, insurance brokers' data analytics tools are becoming more sophisticated in response to their middle-market clients' risk management needs.

While certain analytics, such as benchmarking tools used to compare insurance purchases of organizations in a particular industry, typically are available at no additional charge to most middle-market clients, some brokers do charge additional fees to conduct more sophisticated analyses, such as predictive modeling.

“Analytics can run the gamut from what we might term "benchmarking' — information about the types of deals and insurance being purchased in the market — to loss projection modeling, looking at a company's historical loss information and projecting what might happen in the future,” said Claude Yoder, head of global analytics at Marsh L.L.C. in New York.

“If you keep going, we have things that are more stochastic, understanding the nature of possible outcomes and creating that distribution. What we've been doing more and more is predictive analytics — looking at a profile of a company, a profile of a situation — and predicting what the outcome might be.”

For example, to help clients get a better handle on their cyber risks, Marsh recently introduced a model that predicts the likelihood and severity of a data breach.

“There are a lot of companies across different types of industries that are concerned about data breaches. This information can be used to determine whether and how much cyber risk coverage to purchase,” Mr. Yoder said.

“All clients are looking for more analytics to factor into their decision-making process,” said Sandy Crystal, executive vice president of Crystal & Company in New York. “Our client base — whether it's a risk manager or a company without a risk manager — ultimately they need to justify to senior executives why they're making the decisions they're making.”

Using data derived from sources such as U.S. Securities and Exchange Commission filings, historical merger and acquisition activity, and publicly available claims data, Crystal has developed a directors and officers liability analytics tool to help publicly traded organizations determine their likely exposure to securities claims.

Crystal also recently introduced a workers compensation predictive modeling tool to help midsize companies identify potentially costly occupational injury claims so they can be reviewed, managed and closed sooner.

But Julie Zimmer, vice president of sales and middle-market segment leader at Hub International Ltd. in Chicago, said most of the broker's middle-market clients are interested more in benchmarking capabilities than predictive models, though Hub provides both.

“I don't have a lot of middle-market clients wanting to know the day of the week they get most injuries. They're not as granular as the bigger risk management organizations that have big retentions, but they still need information to learn and to become better,” she said.

“Middle-market clients are more interested in market insight. They want benchmarking that compares them to their peers,” said Rob Stein, New York-based chief broking officer of U.S. retail operations at Aon Risk Solutions, whose Global Risk Insight Platform has been used for nearly a decade to help clients perform benchmarking exercises. Aon also has predictive modeling capabilities that perform more sophisticated analytics.

For example, some middle-market clients may be interested in the outcome of a recent analytics project Aon performed on how the aging workforce affects workplace injury claims, Mr. Stein said.

“It's using technology and modeling to be specific about an issue that's facing clients today,” Mr. Stein said. “For example, workers in the 45 years-and-older category have increased 49% and now make up 44% of the available workforce. This will require engineers to re-engineer their work environment and change their safety and ergonomics programs.”

Victoria Nolan, risk and benefits manager at Clean Water Services in Hillsboro, Oregon, said she has not yet requested data analytics from her broker “in part because we are the keepers of our own data. Therefore, it would require more work on our part to transfer or input the data that at the present time we do not have the resources for. The alternative of getting analytics of data from similar entities has not been very useful in the past.”

However, she plans to include questions in her next broker request for proposal about the type of analytics the water utility company can get.

“I figure that this is such a rapidly developing area that there should be some better tools available in the very near future,” Ms. Nolan said.

Aon's Mr. Stein said, “Broker analytics is probably more advanced than it's ever been, but it likely hasn't even scratched the surface of what it might become due to the rapid pace of the technology that's available.”