People eat more ice cream in the summertime. As statements go, that one isn’t likely to make an anthology of insightful wisdom anytime soon. It turns out, however, that it actually has significant implications for how we assess general liability (GL) risk.
The risk of most GL perils like slip, trip, and fall (ST&F) is directly related to the number of people who are exposed to the peril. The more people that walk over a patch of ice, the greater the chance that someone will slip on it. Despite the fact that customer visits is a key driver for GL exposure, however, insurers almost never consider it in pricing models.
The reason for this is pretty simple – it’s not cost effective to sit outside someone’s business with a notepad and count how many people go in every day. As an industry, therefore, insurers rely on revenue as a proxy for customer visits. This works pretty well. There’s a strong correlation between the two, and revenue is way easier to get. There are several problems with it, however.
Let’s consider an ice cream parlor and a soup shop. Both are classified with NAICS codes as “Snack and Nonalcoholic Beverage Bars”, and both have comparable revenue. The timing of that revenue is radically different, however. Anywhere north of the Mason-Dixon line ice cream parlors do the bulk of their business in the summer, while soup shops have higher volume in the winter. The result is that soup shops have far higher exposure to ST&F due to icy conditions. An annual revenue number will never show that, however. Pricing a GL policy simply on revenue and NAICS codes will generate the same premium for both businesses, when the actual risk may be different by a factor of three or more.
Another issue with timing is lag. Since revenue is a lagging indicator, relying on it as a proxy for GL exposure means the data is always out of date. In a steady-state business this may not be a huge factor, but as we are attempting to understand the long-term effects of COVID on our economy, working on old data is like trying to drive a car by looking in the rearview mirror.
In addition to timing, bias is also another problem with revenue data. Unless a business is publicly traded, insurers are relying on the business to provide them with their own revenue numbers. This is the same business that is attempting to minimize their insurance premium. Insurers can mitigate this conflict of interest by invoking audit clauses, but that brings us to the final problem with revenue – ease of access.
Insurers use revenue because it has been easier to get than visit data. But easier does not equal easy. Actually verifying revenue data requires a human being to manually review tax returns, receipts, or other documents. This, in turn, undermines any attempts at straight-through risk assessment for small commercial policies. When it comes to revenue data insurers can have easy or we can have trustworthy, but they can’t have both.
All that said - there is a better way for insurers to use data to model GL and ST&F. Anonymized high-precision building footprints, and detailed points of interest provide insurers with a source of truth for how people are actually interacting with the physical places they are insuring.
Collectively, these datasets give insurers a new option for estimating customer visits. We know what and where businesses are, and how consumers move in and out of those businesses. This gives insurers a new metric that is highly correlated to total visits and addresses the key problems with using revenue highlighted above.
SafeGraph’s sole focus is curating geospatial data that businesses can depend on. With high quality POI and geometry data, insurers can have confidence in their GL and ST&F analysis.
SafeGraph data enables insurers to:
