Tracking Commissions and Chargebacks in a CRM

An advanced commission is a loan. The carrier pays you first-year commission up front — commonly nine or twelve months of it — against premium the client has not yet paid. If the policy lapses inside the advance period, the carrier takes back the unearned portion.

That is unremarkable until you notice what it means for cash flow: an agent writing thirty policies a month is carrying a rolling balance of unearned advances that can run well into five figures, against which every lapse is a withdrawal. Agents who track this see a smooth year. Agents who do not have a good March, spend it, and get a February statement with a negative balance.

The difference is entirely record-keeping. This page is what those records have to contain.

The data model

A CRM that handles this properly has a policy object separate from the contact, and the policy object needs these fields. If your system does not have them, you are tracking this in your head or in a sheet.

The two bolded fields are what nearly every generic CRM lacks, and they are the reason an insurance-specific object model matters.

The window arithmetic

Chargeback terms vary by carrier and contract. The common structures:

Full chargeback inside the window. Policy lapses in month 4 of a 9-month advance: the carrier reclaims the unearned 5 months. Some carriers reclaim the entire advance regardless of months paid, particularly in the first 90 days.

Pro-rated. You keep commission on premium actually paid, return the rest. More common on 12-month advances.

First-year cliff. Some products charge back 100% if the policy lapses in the first six months, then pro-rate to twelve.

You cannot infer which applies. It is in your carrier contract, and it needs to be recorded per carrier in your system, not remembered. General mechanics are set out well in this agent-side guide to insurance chargebacks.

The derived field to build: chargeback_window_end = effective_date + window_months. Then a single filter — policies where status = active AND today < chargeback_window_end — gives you your at-risk book. That one view is most of the value of tracking any of this.

The reserve calculation

Here is the number to run monthly.

1. At-risk advance. Sum the unearned advance across every policy still inside its window. For a 9-month advance on a policy in month 3, unearned is 6/9 of the advance.
2. Expected lapse rate. Use your own 13-month persistency if you have twelve months of history. If not, final expense persistency commonly runs in the 65–80% band, and treat anything above that as a claim requiring evidence.
3. Reserve = at-risk advance × expected lapse rate.

Worked example. Sixty in-force policies inside their windows, average unearned advance $310, expected lapse 25%:

• At-risk advance: 60 × $310 = $18,600
• Reserve: $18,600 × 0.25 = $4,650

That $4,650 is not income. It is a liability sitting inside your bank balance. An agent who treats it as spendable has a systematic overstatement of earnings that compounds with volume — which is exactly why the agents who blow up are usually the ones having a good year.

The persistency half of this — how to move the lapse rate rather than just reserve against it — is in persistency and chargebacks.

The reports your system has to produce

1. At-risk book. Active policies inside their chargeback window, with days remaining and unearned advance. Sorted by window end ascending.

2. Draft failure alerts. A missed draft is the single earliest reliable signal of a lapse, and it arrives weeks before the policy actually terminates. A system that surfaces failed drafts within 48 hours converts a chargeback into a save call. This is the highest-ROI alert in the entire category, and it is worth checking whether a platform does it before checking anything else.

3. Chargeback history by lead source. This is the report that changes buying decisions. Two lead sources with identical cost per issued policy can have 15-point different persistency, and only this report shows it. Feed it into the 90-day vendor audit.

4. Commission reconciliation. Expected commission per issued policy versus what the carrier statement actually paid. Discrepancies are common — wrong contract level applied, product misclassified, override misrouted — and they are found by comparison or not at all.

5. Persistency by agent, for agencies. A downline producer with high volume and poor persistency is generating overrides you will pay back. See recruiting agents and downline.

Doing this without a CRM

Possible, and worth doing properly if you are below the buying threshold in do you actually need a CRM.

Build one sheet, one row per policy, with the fields in the table above. Add a formula column for window end date, and a conditional format that flags anything within 30 days of it. Reconcile against carrier statements monthly — not quarterly, because a misapplied contract level found in month one is a phone call and found in month six is an argument.

The limits show up around a hundred in-force policies, or the moment a second person needs the same view. That is the same threshold everything else in this category hits, and the migration field map exists so the sheet you build today imports cleanly when you cross it.

Nothing here is tax or accounting advice. It is the record-keeping your accountant will ask you for.