Telecom commission reconciliation: a process that actually catches shortfalls

Commission reconciliation is comparing what a supplier owed you against what the supplier actually paid, line by line, and chasing the difference. Almost every agency says it reconciles. Most are really doing a total check: the deposit looks about right, so the statement gets filed.

A total check catches a missing statement. It does not catch a single service that quietly stopped paying eleven months ago. This post is the line-level version: the process, the variances worth hunting, and the point where doing it by hand stops being viable.

What reconciliation actually requires

Three things, and the first is the one most agencies do not have.

An expectation. A record of what you believe you are owed this month, per service, per client, per supplier, derived from your own contract and rate data rather than from the statement. Without this you have nothing to compare against, and you are not reconciling, you are reading.

The actual. The supplier statement, parsed to the line level rather than the total.

A comparison you can act on. Expected against actual, per line, with the differences sorted so a human can work them in priority order.

If you only take one thing from this post: the expectation is the hard part and the whole point. Building it is most of the work, and once you have it, reconciliation becomes arithmetic.

The monthly process

A version that works, whether you run it in a spreadsheet or a platform.

1. Refresh the expectation. Before the statements arrive, update your ledger for anything that changed: new services installed, disconnects, upgrades, rate changes at renewal. This is the step that gets skipped under time pressure and the step that makes everything downstream wrong.

2. Import the statements. Every supplier, every line. Supplier formats differ in layout, field naming, and how they identify a service, which is why this step resists automation more than it should.

3. Match lines to services. Join each statement line to the service in your ledger. Identifiers rarely match cleanly across systems, so expect to maintain a mapping between the supplier’s reference and yours.

4. Compute the variance. Expected minus actual, per line. Sort by absolute value.

5. Triage. Not every variance is an error. Separate them into: explained by something you know about, timing differences that will resolve next month, and genuine problems.

6. Work the real ones. Open a dispute or query with the supplier or TSD, record the reference, and track it to resolution. This is where recovery actually happens, and where most processes die because nobody owns the follow-up.

7. Close the period. Record what was resolved, what remains open, and what you wrote off. Carry the open items into next month.

The most common failure is not step four. It is that steps six and seven have no owner, so variances get found and never chased.

The variances worth hunting

Sorted roughly by how much money they hide.

Missing lines. A service that is live, billing, and simply not on the statement. The highest-value category and the hardest to see, because you are looking for the absence of something. This is what an expectation ledger exists to catch, and nothing else will.

Rate errors. The line is present but the percentage applied is wrong. Often a rate that was correct under an old agreement and never updated after a renewal or a supplier programme change. Small per line, large in aggregate, and it persists for months because nothing looks broken.

Quantity and seat drift. The client added twenty seats; your commission is still calculated on the original count. Common on UCaaS and anything else with a per-seat component.

Services that stopped paying. A line that appeared for fourteen months and then did not. The client is still being billed. Nobody notices, because the total only moved by a little.

Unapplied spiffs. A qualifying bonus that never arrived. Easy to miss because spiffs land on different schedules and are irregular by nature, so a missing one looks like a month without one.

Clawbacks you cannot verify. A deduction appeared. Was it valid, correctly calculated, and applied to the right service? Without your own contract-term data you cannot tell, so most agencies accept these unchecked.

True-ups in your favour that never came. When a supplier corrects a billing error for a client, your commission should be corrected too. It is not always.

What it is worth

Be sceptical of anyone quoting you a percentage of commissions recovered as an industry figure, including us. The honest framing is arithmetic you can do yourself.

Take your annual commission income. Estimate the share you would not notice going missing: one mid-sized service, or a rate a couple of points low across a supplier. For most agencies that number is uncomfortably large relative to the effort of checking, which is the actual argument for reconciliation. Not a headline recovery rate, but that the losses are invisible by construction and therefore unbounded until you look.

When to automate

There is a real threshold, and it is not about sophistication.

A spreadsheet is fine when you have a handful of suppliers, a book you can hold in your head, and one person who owns the process and enjoys it. Plenty of profitable agencies run this way for years. If that is you, take the free reconciliation template and start there rather than buying software.

A spreadsheet stops being fine when any of these are true:

• More than a few suppliers, each with its own statement format
• Enough services that a missing line cannot be spotted by looking
• More than one person needs to work the process, or the owner takes holiday
• You need history: what did we expect in March, before the true-up rewrote it
• Disputes are being found but not tracked to resolution

The tell is usually the fourth one. The moment you need to answer “what did we think we were owed at the time,” a spreadsheet that gets overwritten every month has already lost the information.

We build commission management software that does this, so discount our view on where the threshold sits. What we would defend regardless of what you buy: build the expectation ledger first. It is the asset. The comparison is easy once it exists, and no tool can reconcile for you if you have never written down what you are owed.

Start here

If you reconcile nothing today, do not begin with your whole book. Pick your largest supplier, build the expected figure for one month for that supplier only, and compare it to the statement line by line. It will take an afternoon and you will learn two things: roughly what your variance rate looks like, and how bad your service data is.

Both are worth knowing before you decide whether this is a spreadsheet problem or a software one.