The news on insurance rates for construction companies just keeps getting worse. A recent article in ENR summarizes the situation pretty well: Prices Rise Sharply as Clouds Darken Construction Insurance Market for Buyers.
In the article itself, ENR quotes the brokerage, Marsh, as stating that base pricing has gone up 25% since 2018, and umbrella insurance renewals — policies that kick in after the limits of other policies have been reached — have as much as doubled. More specifically, according to the global insurance broker Willis Towers Watson, general liability coverage renewals are up as much as 25%, builders' risk is up as much as 20% and subcontractor default insurance up to 10% more. These increases are particularly hard on the construction industry, since insurance can run as high as 4-6% of a total project’s costs. So, it’s in the interest of any construction company to lower those premiums.
There is a bit of a silver lining in the data. Firms that have a history of smaller losses can expect lower hikes that range from 5% to 10%. The trick, of course, is reducing those losses to get lower premiums. Worker and third-party injuries are always listed among the biggest risks for any construction company. And thanks to predictive-based safety (PBS), there’s a proven means of substantially reducing safety risk and, as a result, reducing losses.
Using AI and Predictive Analytics to Change Safety Behavior on Job Sites
PBS uses artificial intelligence (AI) and analytics technologies to analyze a wide array of project and observational safety data to predict accidents before they happen and then provide recommendations for how to prevent them.
What kind of data? If it has anything to do with a project, it probably has value, but there are three primary categories. First, there are safety observations, which can be conducted by AI or human observers. In the case of human observers, Newmetrix (formerly Smartvid.io) can make that process super simple and effective. And as for AI, Vinnie (our nickname for the Newmetrix AI), can analyze site images from essentially any format to create observations of a wide array of hazards, from work at height and heavy equipment hazards, to improper body positioning and lack of PPE.
Safety incidents are the second big bucket, and the definition of what qualifies should be broad: recordable incidents, near misses, property damage … the more data, the better.
Finally, almost any general project data that the contractor can collect, such as weather data and staffing levels, can all help the AI make more accurate predictions.
Because that’s what it’s all about, ultimately: providing the AI with plenty of accurate data that it can use to predict where the next safety incident is likely to occur and provide prescriptive recommendations for preventing them. But like any other computer program, the predictions of an AI are only as accurate as the data it is fed. To get a high level of accuracy, data observations, in particular, shouldn’t be used to “nail” people, because if that takes place, people will be less willing to report safety hazards when they see them, which fouls the data and skews predictions.
Once the AI makes its predictions and prescriptions, the organization needs a process to act on them. If these insights go unused, they’ll have zero impact on the recordable incident rate.
And PBS has an enormous impact on safety. Adopters of PBS have seen predictions with up to an 86% accuracy rate and up to 60% reductions in the annual rate of recordable incidents. Those two figures have caught the attention of major insurers.
Insurance Implications of PBS
Insurers traditionally determine a construction company’s rate by looking at the standard rate, and then factoring in that company’s previous losses, its experience modification rating (EMR) and its safety practices. It’s a fairly blunt instrument, but it’s been effective at mitigating insurers’ risk so they don’t lose money too often on their policies.
PBS, however, changes the game, because it can enable personalized premiums. All that PBS data enables insurers to predict and base pricing on the individual company risk, instead of using the traditional EMR model. PBS also enables insurers to more accurately predict future results in their underwriting, provide more personalized customer service and streamline the claims process. (If you’d like to go deeper into this topic, check out this article that Rose Hall, Vice President, Head of Construction Innovation at the insurer AXA XL co-wrote with Newmetrix’s Tim Gattie.)
But don’t take our word for it. Listen to Bret Bush, Risk Engineering Strategy and Execution Lead at the IoT Innovation Lab for The Hartford, one of the largest underwriters of workers compensation insurance in the world:
"A better run business is going to result in less losses and there should be rewards for that,” Bush said, “through underwriting, through terms and conditions, and the opportunities to purchase insurance …."
Bush had a lot more to say in a webinar we conducted with him — Construction Tech with an Insurance Lens — where he also discussed how construction companies can get The Hartford to pay for the installation of Newmetrix on their job sites.
PBS is making construction jobsites safer, and because it’s reducing risk, it’s also lowering insurance premiums. I think that’s a pretty solid ROI for everyone.
Want to get started with PBS at your company? Download our Data Driven Guide!