General liability limit checking when the show accepts two different limit structures

A rule was written into the workflow last spring: general liability each occurrence must be at least one million dollars and the general aggregate must be at least two million. It runs against every certificate, it takes no time at all, and 44 exhibitors on a 341 exhibitor show failed it.

Somebody went through the 44 by hand in December. Four of them had a genuine shortfall. The rest were compliant under a structure the show accepts and the rule had never heard of. That is what general liability limit checking looks like when it is written as one comparison instead of as a decision across branches.

What the show actually requires

Start from the published requirement, in full, including the word or.

NAMM's exhibitor insurance requirements for its 2027 show, published in 2026, ask for comprehensive general liability with either minimum limits of $1,000,000 each occurrence of bodily injury and death and $1,000,000 each occurrence for property damage, or a minimum combined single limit of $2,000,000 each occurrence for bodily injury, death and property damage. Two structures, either of which satisfies the show.

Other shows publish a different shape again. IAAPA's requirements for its 2026 Expo ask exhibitors running certain activities for US$3 million general aggregate with US$2 million per occurrence. The American Urological Association's 2025 requirements set $1 million for injuries to any one person in any one occurrence with a $2 million aggregate, and add a $4 million umbrella requirement for higher risk activity.

Three shows, three structures, and no way to write one comparison that handles all of them. Whatever your show publishes, the checking logic has to mirror it exactly, including the alternatives.

Where the form and the requirement disagree

The ACORD 25 has a commercial general liability block with fixed boxes: each occurrence, damage to rented premises, medical expenses for any one person, personal and advertising injury, general aggregate, and products with completed operations aggregate.

There is no combined single limit box in that block. Combined single limit appears further down the form, in the automobile liability section, where it sits next to the per accident language.

So a show accepting a $2,000,000 combined single limit for general liability is accepting something the standard form has nowhere to print. Brokers handle it by typing it into the description of operations box, or by showing the equivalent figures across the occurrence and aggregate boxes, or by attaching an ACORD 101 remarks schedule.

The practical consequence for anyone building a check is direct. Searching the page for the phrase combined single limit will find the automobile row, and a rule reading that number is testing a vehicle policy against a general liability requirement. Extract by position and section rather than by phrase, and treat the general liability block and the automobile block as separate objects that happen to share a page.

Evaluating both branches

Back to the 341 certificates and the 44 failures.

Of the 44, thirty one carried $1,000,000 each occurrence with a $1,000,000 general aggregate. Those satisfy the first branch of a NAMM style requirement, which asks for a per occurrence figure and says nothing that a one million aggregate breaches. They failed only because the rule demanded a two million aggregate that the requirement never asked for.

Nine more carried $1,000,000 each occurrence with excess or umbrella coverage shown on a separate row, taking the total available above the threshold. Whether those pass depends on a policy decision the show has to make, covered below.

Four had general liability each occurrence below $1,000,000 with nothing above it. Those are the real failures, and 4 out of 341 is 1.2 per cent of the file.

The cost of the difference is worth stating plainly. Forty exhibitors were told their insurance was inadequate when it was not. Each of those conversations costs an email from a coordinator, a call to a broker, a reissued document, and a re-check, and it burns credibility with an exhibitor who was right all along. Run that across a portfolio of eight shows and it is a meaningful share of an operations team's autumn.

The logic that avoids it is a disjunction. Evaluate branch one, evaluate branch two, and fail the certificate only when every branch fails. Record which branch passed, because that record is what lets somebody later ask how many of your exhibitors are on each structure.

That last count turns out to be useful in a way nobody expects. If 89 per cent of your floor arrives on the same structure, the minority structure is the one your request wording is failing to describe, and it is usually the smaller exhibitors who sit in it. Requirements written by someone thinking about a large corporate exhibitor tend to assume a large corporate policy, and the sole trader with a show specific policy from a specialist provider gets caught by wording that was never aimed at them.

Does the aggregate apply to your show alone?

This is the question underneath the numbers, and the form gives you a partial answer that almost nobody reads.

The general liability block carries a set of checkboxes headed with how the general aggregate limit applies: per policy, per project, per location, or something else specified. A two million aggregate applying per policy is shared across everything that exhibitor does for the whole policy year. If they have a bad summer, the aggregate available to your show in January could be much smaller than the number printed on your file.

An aggregate applying per location or per project is narrower in scope and stronger for you, since it ring fences a limit for the work in question.

Extract that checkbox alongside the numbers. It costs nothing at extraction time and it lets a risk conversation happen on evidence. Whether to require a per project aggregate is a decision above an operations team's pay grade, and having the data is the precondition for anybody making it.

Can umbrella coverage make up the difference?

Sometimes, and the show has to decide once and publish the decision.

Umbrella and excess policies sit above an underlying policy and extend the total available for a covered claim. An exhibitor with $1,000,000 primary and a $4,000,000 umbrella has more available cover than one with $2,000,000 primary and nothing above it, which is the argument for counting it.

Against that, the umbrella has its own attachment point, its own terms, and its own list of underlying policies it sits over, and none of that is visible on a certificate. The umbrella row on an ACORD 25 gives you a limit and a policy number. It does not tell you whether the general liability policy in question is scheduled underneath it.

My own preference is to accept umbrella coverage towards the total, require the primary to meet a stated floor on its own, and say so in the requirement document. That gives smaller exhibitors a route to compliance without letting a certificate with a $100,000 primary and a large umbrella through on arithmetic alone. Whatever you decide, the failure mode to avoid is deciding case by case at a service desk, because that is how two exhibitors in the same aisle get different answers.

Writing the rule so it fails safe

Three properties make the difference between a limit check that helps and one that generates work.

• Extract, then evaluate. Pull each limit into its own field with the section it came from attached. A single boolean computed at extraction time cannot be re-evaluated when your requirement changes, and requirements change whenever a venue updates its hire agreement.
• Fail with a reason. A rejection saying limits inadequate makes a person reopen the document. A rejection saying general liability each occurrence is $500,000 against a required $1,000,000 can be forwarded to a broker as it stands.
• Route ambiguity to a human, and keep the queue small. A missing figure, an unreadable scan, or a structure nobody has seen before goes to a person. A figure that clears a branch goes straight through. The six fields that decide access are all amenable to the same split.

Where this stops

A limit that clears your requirement tells you what the policy was written at. It says nothing about what remains available, and the ACORD 25 itself notes that limits shown may have been reduced by paid claims.

The deeper limit is that adequacy is a judgement your requirement already made, and limit checking only enforces it. Whether $1,000,000 is the right figure for a hall where exhibitors build two storey stands is a question for a risk manager and an insurance broker, and it should be revisited when your show changes character. An operations team can enforce a requirement precisely and still be enforcing a number that was set in 2012.

The third is that limits interact with the other checks. A certificate with perfect limits, the wrong named insured and a policy expiring before teardown fails on both of the other counts, and the party scoring in additional insured verification plus the date tests in expiration tracking have to run alongside this one for any of them to mean much.

Take your current insurance rule this week and read it next to your published requirement, line by line. If your requirement contains the word or and your rule contains a single comparison, you already know how many of last edition's rejections were your own onboarding logic rather than an exhibitor's coverage.