Overview
Reviewing past losses is one of the most practical ways to anticipate future risk, but useful insight requires a structured approach. Comparing absolute loss dollars alone can be misleading; convert losses to a common "exposure unit" to measure relative risk over time.
An exposure unit can be payroll, square feet, units produced, or dollars of sales — whatever best reflects the activity that creates the risk. Tracking losses alongside that exposure lets you see whether claims are proportionate to growth or indicate an underlying control problem.
Key takeaways
- Normalize losses to an exposure unit (payroll, sales, square feet, etc.) to compare risk fairly.
- Chart losses and key business metrics together to spot parallel trends and anomalies.
- Use forecasts and hindcasts to test how well historical relationships predict claims.
- Investigate unexplained shifts and fix process or staffing gaps to reduce future claims.
How it works
Begin by collecting consistent data: insurable losses (claims paid or incurred) and the business metric that best represents exposure for that line of risk. Plot both series on the same time axis so you can compare trajectories and rates rather than raw totals.
Calculate a loss rate (for example, claims dollars per $1,000 of sales or per employee) and examine whether that rate moves with changes in operations, seasonality, or staffing. If a pattern emerges, you can model it and project expected claims under different business scenarios.
For industry-specific perspectives and risk program examples, you can review resources like Loss of Reservoir Insurance and curated briefs such as Insurance & Risk Management Briefs to compare methods and benchmarks.
What it may cover (and what it may not)
This approach helps explain loss trends tied to measurable business activity — staffing levels, payroll, production volume, or property size. It supports budgeting for premium increases and prioritizing loss control investments.
It won't by itself identify root causes that are non-quantitative, such as cultural issues, poor supervision, or isolated bad hires. Those require targeted investigation and corrective action beyond the numbers.
Common mistakes to avoid
Do not compare absolute losses year-to-year without adjusting for exposure; growth in sales or headcount will naturally raise claim totals even if the risk per unit stays constant. Avoid treating one anomalous event as a trend; verify whether it is repeatable or an outlier caused by a singular issue.
Also, don’t ignore simple fixes. If a single poor-performing employee caused multiple claims, that indicates a screening or training gap that should be closed promptly to prevent recurrence.
Questions to ask an agent
Ask about common exposure units they use for your industry and whether your historical loss-to-exposure ratios look typical. Request benchmarks or case studies that show how changes in staffing, production, or space affect loss rates.
Discuss whether loss control services can help diagnose non-quantitative causes and what specific steps will reduce the frequency or severity of the most common claims.
Next steps
Start a simple charting process: collect yearly or quarterly loss dollars and the corresponding exposure metric, calculate loss rates, and plot both series to look for correlation. Use hindcasting (apply your model to past data) to test accuracy and refine your assumptions.
If you want examples of specialized programs or different risk approaches, review material such as Oceanography Risk Insurance to see how exposure definitions vary by activity. When you have results to discuss or need assistance implementing controls, consider taking the next step and talk to an agent about targeted loss-control options and coverage alignment.
Frequently Asked Questions
How do I choose the right exposure unit?
Pick the metric most closely tied to how losses occur — payroll for workplace injury risk, sales for product liability, or square footage for property-related incidents.
How many years of data do I need to see a reliable trend?
Three to five years is a common minimum to smooth seasonality and single-year outliers, but more data improves confidence if your operations change slowly.
What if my losses don’t correlate with any business metric?
Investigate qualitative factors such as training, supervision, or product quality checks; engage a loss control specialist to perform a site review.
Will forecasting claims also predict insurance premiums accurately?
Loss forecasts give a useful indicator of underwriting exposure, but premiums also depend on market conditions, deductibles, and policy limits, so discuss the estimate with your agent.