The Insurance Line Audit That Recovers 15% Overnight

Insurance is the P&L line nobody reads. It shows up as a single number on the fixed-cost schedule, gets renewed annually with a rubber stamp, and quietly grows 5 to 8 percent a year because of inflation and because nobody is watching. On a multi-unit operation it is easily $80K to $200K a year. About 15 percent of that is money you are paying for coverage you do not need, on exposures you no longer have, at classifications that were entered wrong five years ago.

The audit that recovers it takes about a day. I did it two years ago on a Michelin-recognized Bay Area group and moved the annual premium from $118K to $99K without touching a single coverage limit. Same protection. $19K back on the bottom line. Here is the playbook.

The six lines that make up 90 percent of the spend

Every operating insurance program is basically six policies stacked together. Learn the six and you understand your whole schedule.

1. General liability. The slip-and-fall coverage. Usually the second-largest single line, $15K to $40K on a five-unit operation.
2. Property. Building (if you own), tenant improvements, contents, business interruption. $15K to $35K.
3. Workers compensation. Legally required in almost every state. Priced off your payroll and your experience modification. $20K to $60K. Usually the biggest line.
4. Cyber liability. Data breach, ransomware. Growing 15 to 25 percent a year. $3K to $12K.
5. Employment practices liability (EPLI). Wrongful termination, harassment, discrimination claims. $5K to $15K.
6. Umbrella / excess liability. Sits on top of GL and auto, kicks in when a claim blows through the underlying limits. $4K to $10K.

You may also have liquor liability, hired and non-owned auto, crime coverage, employee dishonesty, spoilage, and food contamination. Those are smaller lines. The six above are where the money is, and where the audit finds the leaks.

Leak #1: general liability class codes

Every insurance policy prices your exposure using a class code. The wrong class code will overprice or underprice you by 20 to 40 percent, and once a wrong class code is on your policy it will sit there forever unless someone pulls it out. Nobody at the carrier is going to call you and offer a lower classification.

Common miscodes on a restaurant or hospitality group:

• A fast-casual concept coded as full-service. Higher premium.
• A takeout-heavy location coded with a bar-and-lounge factor it does not have.
• A catering-only kitchen coded as if it had a dining room.
• A pop-up or ghost kitchen coded as a full brick-and-mortar restaurant.

The fix is a written request to your broker asking for the current class code on every location and the class code you should be on. Then produce two years of receipts, a floor plan, and a menu that supports the lower classification. Repricing is retroactive to the current policy period in most carriers, meaning you get a refund on the current year, not just a lower rate next year.

Leak #2: property coverage on units that no longer exist

Every multi-unit operator I have audited has property coverage on a location that closed. Sometimes the location closed two years ago. The premium is small ($800 to $3,000 per location per year), which is why it never gets flagged. Read the schedule of locations on your declarations page. If a location is listed and the doors are shut, get it removed and get the prorated refund.

Property audits also miss:

• Sprinkler system credit. If you have a monitored sprinkler system and your policy is not giving you a 5 to 15 percent credit for it, ask. Provide the inspection certificate.
• Central station alarm credit. Same story. 3 to 8 percent credit if you have monitored intrusion or fire alarm.
• Coinsurance mistakes. If your building or contents are insured to less than 80 or 90 percent of replacement cost, the coinsurance clause will haircut any claim proportionally. Most operators are either underinsured (and will get haircut at claim time) or overinsured (and are paying premium on coverage they will never use).
• Business interruption limits. Priced off your gross earnings. If those numbers were entered before your last two openings, the coverage is out of sync with the exposure.

Leak #3: workers comp and the experience modification factor

Workers comp is priced by class code, payroll, and something called the experience modification factor (EMR or MOD). The EMR is a multiplier: 1.00 is average, below 1.00 means better-than-average claims history, above 1.00 means worse. A 0.90 EMR versus a 1.10 EMR is a 22 percent swing on your workers comp premium.

The EMR is calculated by the National Council on Compensation Insurance (NCCI) or your state's rating bureau, based on the data your carrier reports to them. Two things go wrong:

1. Your carrier reports claims that never became payouts. A first-report-of-injury with zero actual medical cost still shows up as a claim if the carrier does not close it clean. That claim raises your EMR for three years.
2. Your payroll classifications are wrong. Bartenders coded as cooks, or vice versa, will move the EMR by several points because the base rates differ.

Pull your EMR worksheet from the rating bureau. It is a free document. Your broker has it. Audit every claim listed. Any claim that closed at zero, or that you paid out of pocket, or that was reversed, should be scrubbed. This is a 60-day process and it can move your EMR by 5 to 15 points, which on a $40K workers comp premium is $2K to $6K a year.

The EMR worksheet is the single most overlooked document in the commercial insurance stack. It sits at the rating bureau. It is free. Nobody looks at it. Most of the workers comp savings in any real audit come out of that one page.

Leak #4: cyber, EPLI, and umbrella creeping up

The "just in case" lines are the fastest-growing part of the modern insurance schedule and the least audited. Three specific traps:

Cyber liability

Cyber has been the runaway premium line for four years running. Carriers are asking for detailed IT questionnaires and pricing off them. If you have multifactor authentication on all admin accounts, endpoint detection, and offline backups, say so on the questionnaire. Most operators leave the questionnaire vague and get the worst-case rate. A completed, accurate questionnaire can save 20 to 40 percent on the cyber line.

EPLI

Employment practices coverage is priced off headcount. If your headcount dropped and you did not update the carrier, you are overpaying. If it grew, you are underinsured. Also: the deductible on EPLI is usually $10K or $25K per claim. Moving it to $50K per claim usually cuts the premium 15 percent and you can self-insure the difference from a small reserve.

Umbrella

Umbrella limits get set once and never revisited. A $5M umbrella on a five-unit group may have been right at inception. On a group with $30M in revenue and enterprise catering exposure, it might not be. A $10M umbrella on a group that closed two units and shed revenue is oversized. Either way, worth a look.

The audit, step by step

Here is the sequence I run. It takes about a day for a five to ten unit group.

1. Pull every declarations page. All six policies. Every location on the schedule. Every limit, every deductible, every classification.
2. Reconcile the schedule of locations against reality. Anything on the schedule that is closed, get removed. Anything that opened and is not on the schedule, get added.
3. Verify every class code. One page per location, comparing the actual operation to the code description in the ISO or NCCI manual.
4. Pull the EMR worksheet. Audit every open and closed claim.
5. Fill out the cyber questionnaire accurately and completely. This one is worth an hour by itself.
6. Ask the broker for a credit review. Sprinkler, alarm, safety program, driver training. All documented, all applied.
7. Get one competing quote from a different broker. Not five. One. Enough to know if your incumbent is close to market.

The rebid cycle: every three years, not annually

Annual rebidding is what brokers want you to do because it lets them shop you around every year and take a commission on whichever carrier lands. It also destroys the loss-history discount you build up by staying with a carrier for three or more years. Most carriers give a persistency credit of 3 to 8 percent starting in year three. Rebid annually and you never earn it.

My rule: rebid every three years. In years one and two, do the audit above and negotiate with the incumbent. In year three, take it to market with a real RFP against three carriers. That cadence saves more money over five years than annual shopping does, and it takes about a quarter of the effort.

Insurance brokers make more money when you switch carriers. You make more money when you stay long enough to earn the loyalty credit. Their interest and yours are not aligned on renewal timing.

The real example: $118K to $99K

The Bay Area multi-unit group I keep referencing had five locations, 215 employees, and a total insurance program of $118K. The audit turned up seven specific fixes:

• Two locations coded as full-service that were actually fast-casual with limited seating. Reclassified. Saved $6,200.
• Property coverage on a location that had closed 14 months earlier. Removed. Refunded $2,100 for the current year, saved $2,400 going forward.
• Sprinkler credit missed on three of five locations. Applied. Saved $1,800.
• Two zero-payout workers comp claims still showing on the EMR worksheet. Scrubbed. EMR moved from 1.04 to 0.97, saving $3,900 on the workers comp premium.
• Cyber questionnaire completed properly with MFA and endpoint detection documented. Saved $1,600 on the cyber line.
• EPLI deductible moved from $10K to $50K per claim, backed by a $50K reserve on the balance sheet. Saved $1,800.
• Consolidated GL, property, umbrella, and cyber under one carrier for a package discount. Saved $1,300.

Total: $19,100 in year-one savings on a $118K program. About 16 percent. Zero reduction in actual coverage. The broker helped us do it, and made less commission that year, and is still our broker three years later because we asked them to earn the relationship instead of pretending we did not know the market.

What I got wrong the first time

Three things I would do differently:

1. I trusted the broker's audit. I asked our incumbent broker to audit the policy on their own. They found $4K. When I did it myself I found $19K. Brokers audit against carrier interests, not yours.
2. I underestimated the class code fight. Reclassification is the biggest single lever and it is the one carriers push back on hardest. Bring receipts, floor plans, and menu documentation to the first conversation, not the second.
3. I did not build the audit into the annual operating calendar. A one-time audit saves money once. A recurring audit every 24 months holds the number over time. Schedule it before you leave it.

The point

Insurance is the fixed cost that pretends to be a rounding line. On a $30M operation, 15 percent of a $118K premium is $18K. Every year. That is 30 basis points of margin sitting inside a line the operator never touches. Read the declarations page. Verify the class codes. Audit the EMR. Ask for the credits you are entitled to. The whole thing takes a day and it pays for a decade.

The pattern shows up in every fixed cost line, not just insurance. Somebody set the number once. Nobody has looked at it since. The audit is boring, the savings are quiet, and the compounding is real. On a five-year horizon, the operator who audits fixed costs every 24 months keeps five to seven figures more of the money the operation is making than the operator who does not.

Cadence beats charisma, and audits beat renewals. Pull your policies this month.