Before funding another AI rollout, keep premium, commission, staff capacity and profit in separate rows. Premium measures the insurance business placed. Your MGA’s return depends on the revenue it keeps and the costs it incurs, which means adding all four numbers can count the same improvement more than once.
In one insurance underwriting group we helped, finance independently measured $1.96M in added net recurring profit. We connected intake and carrier reporting through one submission record, reducing repeated preparation around the underwriter’s decision. Reporting fell from 40 to 10 staff-hours per program each month. The group also recorded $16M more bound premium and $2.4M in additional commission as separate results. The case doesn’t supply a common financial period or a component-by-component reconciliation, so those figures don’t form one total.
That separation is the part to carry into your next funding request. Let’s say a program produces $20,000 more commission in a month and needs $6,000 in extra operating costs. That leaves $14,000 of additional contribution before other changes and one-time costs. If the same team also frees 50 hours while payroll stays unchanged, those hours give it room to work. They don’t add another cash saving to the $14,000.
Here’s how to build a benefit ledger that keeps those distinctions intact, shows what changed because of the workflow, and gives finance a usable decision about expansion.
Start with the revenue your MGA keeps
A larger premium figure can hide a smaller financial gain because the MGA doesn’t keep the whole premium as its own revenue. Use your actual commission and fee arrangements to establish the amount your business earns. Check whether the figure already excludes payments to distributing brokers and whether cancellations, adjustments or contingent payments change the amount finance can recognize.
Ryan Specialty’s published revenue policy illustrates why the distinction matters. It describes commission and fee revenue separately from premiums held for others, and it explains how cancellations and contingent commissions affect recognition. Your finance owner should apply your contracts and accounting policy to the ledger.
For the worked month, suppose the additional bound premium is $200,000 and the MGA retains 10% as commission after broker sharing. Assume that $20,000 is the revenue finance recognizes for the month after applicable adjustments. The premium explains where the commission came from, but it isn’t a second revenue benefit. Put both figures on the record and mark the premium row “context only.”
Now give the comparison a date and a population. Name the program, the submissions included, and the month being measured. Compare actual revenue with what that same program would reasonably have earned without the change. A prior month can be a starting point, but a change in incoming submissions, premium rates, risk mix or carrier appetite can change the result even if intake work stays the same.
Follow the changed work before crediting AI
The source case gives us a concrete operating mechanism. Intake and carrier reporting reused much of the same information, but teams had been entering and assembling it more than once. We built a shared record and tools around it, so the reporting team could use information already captured for the policy. Underwriters kept the insurance decisions. The reduction in report work and the finance-measured profit are different pieces of evidence about that engagement.
For your ledger, connect each claimed benefit to an equivalent change you can inspect. If faster intake is meant to produce more bound business, trace the submission from receipt to a complete file, then to the underwriter’s decision and the recognized commission. That shows where work changed and which business passed through the changed workflow. A faster preparation time alone doesn’t establish additional commission.
In our example, the $20,000 increase is observed in the hypothetical finance records. Crediting it to the AI workflow still needs a comparison. Have the underwriting and finance owners check a comparable untreated group or a documented estimate of what the program would have produced without the change. Match the type of business, demand and period as closely as the records allow. Record the remaining differences so an estimate doesn’t become a measured causal result just because it appears in a spreadsheet.
The calculation below assumes the full $20,000 is incremental for this teaching example. If the owners can support only part of it, the accepted benefit must fall with that evidence. That is how this record helps you fund growth: it makes the supported gain visible before the next program inherits an optimistic assumption. A wider growth decision still starts with the business goal and the bottleneck most limiting it.
If you’re an MGA leader and want to grow your business without adding programs that keep your team busy but leave profit flat, we’ll identify the biggest bottleneck standing in your way, build the plan to remove it, and implement it in eight weeks* or less - guaranteed or we continue to work for free and you don’t pay.
Find my bottleneck* Eight-week delivery, or twelve weeks when substantial data foundations are required, as agreed before work begins. If a covered stage fails its agreed benchmarks, we provide up to four additional weeks of remedy work at no charge, then waive or refund that stage’s fee if it still fails. Successfully delivered stages remain payable. Agreed client-caused delays pause the clock. Delivery guarantee terms.
Work one month through the ledger
Here’s what that looks like in practice. The hypothetical program incurs $2,000 for the new software and model usage, $1,000 for support and checking, and $3,000 in extra servicing costs. Those costs total $6,000 for the month. Assume these are all the changed recurring costs within this example’s program scope, with each cost included once.
| Measure | Worked month | Treatment in the decision |
|---|---|---|
| Additional bound premium | $200,000 | Business volume only. Its related commission appears below. |
| Additional retained commission | $20,000 | Recognized revenue for the month. Attribution to the workflow is an explicit example assumption. |
| New software and model usage | $2,000 | Subtract the incremental monthly cost. |
| Support and checking | $1,000 | Subtract the incremental monthly cost. |
| Additional servicing | $3,000 | Subtract the cost of handling the added business. |
| Released staff time | 50 hours | Keep as capacity because payroll is unchanged. Include no separate dollar benefit. |
| Incremental contribution before transition costs | $14,000 | $20,000 minus $6,000. This is a subtotal, not another benefit to add. |
| One-time transition expense | $8,000 | In this example, finance expenses it fully in this month. |
| Contribution after that transition expense | $6,000 | $14,000 minus $8,000. This remains a program contribution measure, not firm-wide net profit. |
The $14,000 helps you judge the ongoing operating change. The $6,000 shows what remains in this month after the assumed transition treatment. Keep both, with their labels, because a rollout can improve ongoing contribution while taking time to recover its setup costs. Use the treatment finance actually applies, and track payment dates separately if you’re making a cash-funding decision.
The 50 hours can still matter. If the team uses that time to handle the business producing the additional commission, the resulting revenue is already in the commission row. Adding hours multiplied by a salary rate would count another benefit even though payroll didn’t fall. If a contractor invoice actually falls, record the reduction against that invoice. If a planned hire becomes unnecessary, record the avoided future cost as an estimate against the staffing plan, with its start date and assumptions.
Give related rows one overlap label, such as “program-A-growth.” The premium, related commission and final contribution belong to that group. Finance can then see the detail without summing the detail and its subtotal. Use a separate group for a genuinely separate expense reduction, and check that it hasn’t already been included in the cost comparison.
Now change one condition. Suppose finance can support only $10,000 of the additional commission as attributable to the workflow, while all $6,000 of the added recurring costs remain necessary. The contribution becomes $4,000 before transition costs and negative $4,000 after the $8,000 transition expense. The rollout’s costs haven’t changed, but the supported benefit has. That changes the case for expansion, which is why the attribution assumption belongs beside the calculation.
Copy one record for each claimed benefit
The finance owner and the operating owner should complete the record together. Finance checks the amounts and treatment. Operations checks what changed in the work and whether the comparison describes the same job. That combination prevents an invoice total from becoming an AI result without a supported operating explanation.
Copy this blank structure for each benefit row. Keep the related costs and overlap references attached, even when another team owns them.
| Field | Fill in for your program |
|---|---|
| Decision and scope | The rollout or renewal being considered, program, workflow and included submissions. |
| Measure and units | Premium, retained commission, expense, staff-hours, contribution or profit. Name currency where needed. |
| Matched period | Actual period, comparison period and any adjustment needed to make them comparable. |
| Baseline and actual | The two amounts, their units and the records that support them. |
| Without-change estimate | What would have happened without the workflow change, how it was estimated and what could explain the difference. |
| Changed work | The action that changed, its owner and the route from that action to this measure. |
| Source and status | Links to permitted records. Mark the amount observed, calculated, estimated or unmeasured. |
| Incremental costs | Software, checking, support, servicing and transition costs, with period and accounting treatment. |
| Overlap group | Related rows and totals that already include this benefit. State which row enters the calculation. |
| Accepted amount | The supported financial benefit or capacity measure, without mixing the units. |
| Accountable owners | Finance and operating owners, acceptance date and unresolved question. |
| Funding decision | Expand, retain, narrow or hold. Record the next evidence check and its owner and date. |
For a completed illustrative record, take the commission row above. Its scope is one program’s intake workflow for one month. The comparison revenue is $100,000 and actual recognized retained commission is $120,000. The difference is $20,000. The sources would be the monthly commission ledger, policy records and submission timestamps, checked by the finance lead and underwriting operations lead. The claimed mechanism is more eligible files reaching an underwriter in time to be written.
The worked record marks the revenue difference as observed within the hypothetical and the full workflow attribution as an assumption. It attaches the $6,000 recurring costs and $8,000 transition expense, uses “program-A-growth” as the overlap group, and excludes the premium and released-hours rows from the dollar total. Its decision is to hold expansion until the owners support the without-change estimate. Finance owns that check at the next monthly close. Replacing that assumption with evidence is what turns the calculation into a funding basis.
A spreadsheet is enough to begin. Use exports from the policy, commission and cost records you already have. If AI helps extract or match those records, require source references, check the matches and have it leave missing amounts unresolved. Let a fixed calculation produce the totals and let finance accept them. A generated explanation shouldn’t silently supply a commission rate, missing invoice or causal share.
Use the accepted benefit to make one funding decision
Before the next expansion meeting, take one benefit from the AI proposal and complete this record with its two owners. You’ll have an amount finance can accept, a clearly marked estimate, or a specific missing record to obtain. Each outcome tells you more about the next commitment than a total made from premium, profit and hours.
For another program, repeat the comparison with that program’s commission arrangements, incoming work, servicing demands and rollout costs. The first program’s result doesn’t establish those inputs. If existing software configuration or a change in work allocation can produce the same accepted output, compare that route too. The MGA intake route comparison helps when you’re choosing between AI, offshore support and your current team. Once the economics support a route, the new-program readiness test checks whether it works from intake through carrier reporting.
Your bigger goal may be more profitable growth over the next 3–18 months. This ledger helps you decide whether one operating change contributes to that goal. Choosing the next business priority requires comparing it with the other demands on your team, time and money. Our core-project discussion starts with that goal and whether an engagement fits.
If you’re an MGA leader and want to grow your business without adding programs that keep your team busy but leave profit flat, we’ll identify the biggest bottleneck standing in your way, build the plan to remove it, and implement it in eight weeks* or less - guaranteed or we continue to work for free and you don’t pay.
Find my bottleneck* Eight-week delivery, or twelve weeks when substantial data foundations are required, as agreed before work begins. If a covered stage fails its agreed benchmarks, we provide up to four additional weeks of remedy work at no charge, then waive or refund that stage’s fee if it still fails. Successfully delivered stages remain payable. Agreed client-caused delays pause the clock. Delivery guarantee terms.