Insurance commission reconciliation means matching every policy your agency wrote to the commission your carriers or IMO paid on it, then working through whatever doesn't match. The method that holds up is monthly: log each submission as pending, match it to the issued report, separate existing policies from new business, compare what you were paid with what the rate schedule says, and clear a short queue of problems before you close the month.
Say you spend $4,000 a month on leads for three agents. At month end you get a stack of statements, a policy export from your IMO and a CRM full of contacts. Can you say which of those leads turned into policies, and what each policy is paying you? Most owners can't. The data exists, but it sits in four places that nobody has connected. The steps below connect them, and the checklist at the end is the version you run every month.
What commission reconciliation needs: four kinds of data
Start by knowing which file holds what. Most agencies treat their IMO export as "the book" and assume the money is in there. It usually isn't. An agency principal we work with put it this way in one of our sessions: the data the IMO sends tells you annual premium, but it doesn't tell you commission on anything.
You need four layers to know what you earned, what you're owed and where it came from:
- Submissions: every application your agents send in, logged on the day with the client, carrier, product, lead source and the split on the turn-in sheet.
- Policy or book data: the issued-policy report from your IMO or carrier: policy number, carrier, product, premium, split, writing agent and issue date. No dollars.
- Rate schedules: the commission schedule for each carrier, product and contract level, with the dates each rate applied. Rates change, and agents move up levels, so a policy issued in 2022 needs the 2022 rate at that agent's 2022 level.
- Commission statements: what the carrier or IMO paid, per agent, usually weekly, including splits, cancellations and negative lines.
The submission layer is the one agencies forget. It's also the only thing that links a CRM contact (where the lead came from) to a policy (where the money comes from).
Why your CRM can't be the commission system
It's tempting to type commissions into your CRM next to each contact. It works until you ask who should see whose numbers. An agent should see their own book and their downline. A manager sees their team. The principal sees everything. A CRM built for contacts and campaigns isn't set up for that, so keep commission data in its own place and connect the two through the submission record.
Step by step: commission reconciliation from submission to paid
The sequence, in order:
- Log every submission as a pending record.
- Match the issued-policy report to your pending records.
- Mark existing policies so they don't count as new business.
- Estimate the commission each policy should pay.
- Match statement lines to clients and compare with the estimate.
- Put every mismatch in the reconciliation queue.
- Close the month with a human review.
Step 1: Log every submission as pending
When an agent submits an application, create a record right away. Give it a status of pending, and include the client name, date of birth, carrier, product, the lead source from the CRM and the commission split percentages from the turn-in sheet.
"That means that the record I ingest is going to be quote-unquote pending, like a draft. It's not going to be final issued business because it still needs to be approved." — Kevin Beeftink, founder of RevEngine
The pending record carries the lead source forward. Weeks later, when the carrier issues the policy, you won't be able to work out where that client came from unless you wrote it down at submission.
Get the split right on the turn-in sheet. Your IMO reports commission based on the split it receives, so whoever is on the sheet is who gets credited. If an inside-sales agent worked the lead and a senior agent only ran the appointment, the sheet has to say so, or you will never see lead-spend return per agent. We cover this in more depth in how to attribute commission splits.
Make it one ledger. If turn-in sheets live in separate files per agent, export them into one sheet and add an automation that appends each new turn-in as it comes in.
Step 2: Match the issued-policy report to pending records
When the IMO or carrier report arrives, match each issued policy to a pending record. Use the policy number or application ID where both sides have one. Fall back to name plus date of birth. Name alone is weak: spellings differ, couples share a surname, and the CRM often holds a nickname. A pending record built from the turn-in sheet gives you a far better match than a name search across two systems.
Each pending record ends up in one of these states:
- Issued and matched: move it to issued and attach the policy number.
- Still pending: leave it. Check again next month.
- Declined or withdrawn: close it, and keep the lead source for your conversion numbers.
Any issued policy with no pending record goes to the queue in Step 6.
Step 3: Separate existing policies from new business
This step is where agencies overstate growth without noticing. If a client changes plans, gets a rewrite or shows up again on this month's export, your report will count them as new business unless you check against the prior book.
"The churn analysis made me reconcile policies already known in the system so they're not identified as new business being written. If I already logged that policy, it shouldn't be recognized as a new policy written this month." — Kevin Beeftink, founder of RevEngine
The check is simple. Compare this month's book with last month's. A policy that was there before is existing. A policy that is new on the report but belongs to an existing client gets flagged so someone decides whether it's a replacement, a plan change or a second product. Only policies with no prior match count as new.
Carriers often pay these cases differently too. For Medicare Advantage, CMS compensation rules have plans pay the renewal commission rate, which is lower than the initial rate, when a member moves to a like plan type. Count that as new business and your projected commission is wrong twice: once in the growth number and again in the dollars.
The same comparison gives you churn. A policy in last month's book that is missing this month has lapsed or cancelled. You can't measure churn until existing policies are reconciled, and you can't measure growth either.
Step 4: Estimate what each policy should pay
The issued report gives you premium. To get dollars, apply the rate schedule for that carrier, product, policy year and the agent's contract level on the issue date.
Example: a Medicare supplement policy with $1,800 annual premium, where the agent's first-year rate is 22% (a made-up rate for illustration). Expected first-year commission is $1,800 × 22% = $396, or $33 a month if the carrier pays monthly.
Do this for every issued policy and you have an expected figure to hold each statement line against. Without it, you're reading statements and trusting them.
Step 5: Match statement lines to clients
Your commission statements list the client name, policy number, amount and often the split. Match each line to an issued policy from Step 2. When a line matches, compare the amount paid with your estimate from Step 4.
Ask for statements as CSV files. PDFs are slow to process and easy to misread, and you'll be doing this every week.
Continuing the example: the statement shows $27.50 for that policy, against an expected $33. That's $5.50 a month short. Over twelve months it's $66 on one policy, and short payments like that tend to come from one cause repeated across many policies, such as a wrong contract level or a rate from the wrong year.
Step 6: Work the reconciliation queue
Everything that doesn't match goes in one queue. Treat the queue as a normal part of the month. A visible list of problems is better than a report that hides them.
"Then, we've got reconciliation, so gaps, things that don't quote unquote make sense from whatever data you provided. So, something came up short, wasn't paid, was over. Unpricable because we're missing certain types of data." — Kevin Beeftink, founder of RevEngine
| Queue type | What it looks like | Usual cause | First thing to check |
|---|---|---|---|
| Short | Paid less than expected | Wrong contract level, old rate, split mismatch | Agent level on the issue date |
| Unpaid | Issued policy, no statement line | Carrier delay, wrong writing agent, missing appointment | Writing agent on the carrier record |
| Over | Paid more than expected | Duplicate line, split shares above 100%, rate change you didn't load | Duplicates and split totals |
| Unpriceable | Can't estimate at all | Missing rate schedule, unknown product, missing agent level | Which input is missing |
Two checks catch a lot of the "over" lines. First, look for the same policy paid twice in one period. Second, add up split shares per policy. If one agent shows 70% and another shows 50%, the line totals 120%, and something on the turn-in sheet or the IMO record is wrong.
Unpriceable lines need a different answer. They mean you can't give an answer yet because the data is incomplete, and the report should say so. Don't guess a number. When you fix the cause, such as loading a missing rate schedule, fix it once in the system. Asking every agent to re-upload their own spreadsheets each month moves the problem around without fixing it.
The queue also feeds your list of missing commissions to raise with the carrier or IMO. Group short and unpaid lines by cause before you call, so one correction request covers twenty policies instead of one.
Step 7: Close the month with a human review
Software can do the matching. Someone still has to sit down at month end, work the queue and answer questions from agents and managers. The bigger the book, the longer that takes, so put it on the calendar.
Before you send any report, check it against the numbers people already know. Say a manager knows their team wrote about 30 policies and your report says 19. They will stop trusting every other figure in it.
"So, to what extent does what I'm putting together in this report reflect what they expected to see? Because if it's off, that will immediately make them question everything else, right?" — Kevin Beeftink, founder of RevEngine
Premium and commission totals pulled straight from the IMO data rarely get pushback, because they're the IMO's own numbers. Attribution, meaning which lead source or campaign gets credit, is where questions come up. Be honest about it. You can usually show what a campaign contributed before you can prove it caused the sale.
Month-end commission reconciliation checklist
Run this in the last week of each month. The first month takes the longest, because that's when you find the missing rate schedules and the agents with no turn-in sheets.
- Confirm every submission this month has a pending record with lead source and split.
- Load the latest issued-policy report and match it to pending records.
- Compare this month's book with last month's. Mark existing policies, then list new business and lapsed or cancelled policies.
- Check that rate schedules and agent levels are loaded for every carrier that issued this month.
- Load this month's commission statements as CSV and match every line to a policy.
- Review the queue: short, unpaid, over and unpriceable. Assign each item an owner.
- Check split totals per policy and look for duplicate lines.
- Send correction requests to carriers or the IMO, grouped by cause.
- Spot-check the report against what managers expect before you send it.
- Write down what's still open and carry it into next month's queue.
What a reconciled book shows you
Once policies, rates and statements line up, your book of business starts answering questions you couldn't ask before. A good book dashboard opens on four numbers: estimated annual commission, annualized premium, clients in the book, and a breakdown by product category and carrier. From there you filter by status, category or agent.
Some things become visible once that's in place:
- Premium increases: Medicare supplement clients whose premium rose 10% or more year over year are worth a review call.
- Issue dates: policies near the end of their renewal-paying years show up by filtering on issue date. Any change you suggest has to be in the client's interest, and your compliance team keeps the final say.
- Product mix: clients with one product are cross-sell candidates. See how to analyze your book for cross-sell and churn.
- Lead-spend return: with splits recorded correctly at submission, you can see what each lead source paid back per agent.
For what happens to commission when a policy cancels early, read how a commission chargeback moves down your hierarchy.
What to do this week
- Pull last month's commission statements and one issued-policy report, and match ten policies by hand. Note how many matched on policy number and how many needed a name.
- Find out where your turn-in sheets live. If there's more than one file, start a single ledger.
- Ask your IMO for statements as CSV, and ask for the rate schedules and agent levels that applied over the last three years.
- Put a two-hour block on the calendar for the last week of the month and run the checklist above.
If you'd rather not do the matching by hand, RevEngine reports your agency's earnings, checks whether carriers paid what was owed and finds opportunities in your existing book. You can see how the product works and what it costs, or book a RevEngine demo and bring last month's statements.
