Now there’s a risky business: Using Employee Perception to Identify Poor Insurance Underwriting Risks
A recent OSHA fine announcement led me to check the Aniline scores for the company in question, Sugar Creek Packing Company, a packaged food manufacturer founded in 1966.
What I saw looked troublesome, even without the knowledge that OSHA had previously cited Sugar Creek for safety violations. Here is Aniline’s Integrity score over the last two years shown in relation to the average Integrity score for the Food Manufacturing industry, which is a leading indicator of workplace injury performance and associated workers’ compensation claims:
The Aniline Integrity score measures employee expressions of the company mission along with its values, honesty, ethics, compliance, and commitment.
As illustrated above, over the past two years Sugar Creek’s highest score was just 34 - consistently below the Food Manufacturing industry average of 45 - and stands at 32 as of October 2023.
In November 2021, the low point was 22, lower than 99% of all manufacturing companies in our study of OSHA injury rates among US employers with high exposure to workplace injuries. It’s no surprise that OSHA cited the company in 2019 and 2022 for other violations.
This exemplifies why our partners who underwrite liability insurance have learned to harness the power of our employee perception metrics; in short, culture drives risk.
The Aniline Employee Perception Dashboard reveals weakness in several other areas, suggesting that insurance underwriters - as well as creditors and suppliers - might want to take a moment to consider the relevance of employee perception to their bottom line in areas such as property, product liability, and management errors & omissions (E&O):
These values are derived using Aniline’s “Natural Language Processing” model, or NLP, a form of artificial intelligence.
The Aniline NLP model identifies the topics contained in the unstructured narrative of online employee reviews and then assigns a sentiment score based on the relative levels of negative or positive expression.
Having processed nearly 1 billion employee perceptions, Aniline scores are comparable across companies and are used by insurers to enhance risk selection and pricing decisions.
As examples of what drives these scores, here is a small sample of relevant commentary shared publicly by Sugar Creek’s employees as summarized in the Aniline user platform:
“Washington Court House Location: - OSHA Violations that go unaddressed: missing guards, covers, floor tiles, etc. - Terrible Safety protocols - Improperly maintained/repaired equipment.”
“Equipment never works and I have never seen so many fires at a place I've worked.”
“Cold, hostile, unhealthy, unsafe work environment.”
“They lie to you about everything they offer on the orientation.
“There's a lot of favoritism and nepotism from very top management, which creates a lot of incompetency throughout the company.”
With the advent of online media platforms over the last 15 years or so, the workforce is increasingly comfortable with sharing information. Topics such as pay along with the internal workings of their employers, which in generations past were considered either personal or ill-mannered, are now widely shared online. Perhaps in an altruistic way, this reflects a trend towards transparency - and possibly a desire to give greater leverage to those in entry-level jobs.
While companies have long focused on customer commentary, we recommend that businesses pay attention to workforce signals as an equally important source of meaningful business intelligence.
By James Marple
james@aniline.ai