Usage-Based Billing: How to Check Each Invoice

How do you check a usage-based invoice?

You take the usage the customer actually had in the billing period, rate it at the price their agreement sets for that date, and compare the result with what the invoice billed, line by line. If you close the month as the accountant or finance manager, this is how you know usage revenue is complete and right before anyone else asks. Usage-based billing, also called metered billing, charges a customer for what they used, such as active users, messages, transactions or hours, often on top of a fixed platform fee. Because the amount changes every month, nobody notices when it is wrong, and the errors that slip through mostly bill less than the customer owes.

What makes usage-based billing harder to check than a subscription?

A subscription bills the same amount every month, so a wrong invoice stands out against the last one. A usage invoice is different every month by design, so a low figure looks like a quiet month rather than an error. It also depends on three things that live in different places: the meter that counts usage, the agreement that prices it, and the billing system that turns one into an invoice. Each handoff between them is a place where a number can be lost, cut short or priced at last year’s rate.

How does metered usage become an invoice?

Usage moves through four steps, and each one can be checked against the step before it. The meter records each event as it happens. At the end of the period, a usage report totals the events for each customer and each metric. Each total is then rated: the included allowance is taken off, and what remains is priced at the agreed rate or tier. Last, the rated lines go onto the invoice next to any fixed fees.

Where do metered billing errors come from?

Most metered billing errors come from a handoff, not a calculation, and most of them under-bill. These are the ones worth looking for every month.

• A usage report cut short: the report is pulled before the period ends, or misses a time zone, so the last days of usage never reach the invoice.
• A stale rate: an amendment or renewal changes the price per unit from a start date, and the billing system keeps the old rate.
• The wrong allowance: the included usage is taken off twice, or the allowance from a larger plan is applied after the customer moved to a smaller one.
• A tier boundary: usage that crosses into a higher-priced tier is priced at the lower tier throughout, or the tiers are applied to the whole quantity when the agreement prices each band separately.
• A metric with no price: a new product or add-on is metered, but nobody added it to the billing system, so it is counted and never billed.
• A missed commitment: the agreement sets a minimum spend for the year, and the shortfall, the amount the customer still owes to reach it, is never invoiced at the true-up date.
• A unit mismatch: the meter counts in one unit, such as minutes or gigabytes, and the price list charges in another, such as hours or terabytes, and the conversion is wrong.

Over-billing happens too, usually through duplicate events or an allowance left off, and a customer who is over-billed tends to tell you. Under-billing is silent, which is why it needs its own check.

What do you compare on each usage invoice?

You compare five things, each against the document that sets it.

• Quantity: each usage line against the usage report for the full period, and the usage report against the meter’s own totals.
• Rate: the price per unit against the agreement, as it stood on the dates the usage happened, after any amendment.
• Allowance and tiers: the included usage taken off once, at the level the agreement sets, and each band priced as the agreement prices it.
• Coverage: every metric the agreement lets you charge for appears on the invoice, even when its usage is small.
• Arithmetic: each line, the subtotal, any credit and the total, worked out again from the quantity and the rate.

Worked example: a month of metered usage at Quillfield

Quillfield Scheduling is an invented company, one of the Demo companies Ressura uses to show its checks, and the customer and every figure below are made up for the example. Quillfield sells workforce scheduling software, and one field-service customer, Pinecrest Home Services, pays on a usage-based order form.

The order form sets a platform fee of $2,000.00 a month, which includes 250 active workers and 10,000 text reminders. Each worker above that is billed at $4.00 a month, and each text above the allowance at $0.03. A signed amendment raises the worker rate to $4.50 from 1 September, when Pinecrest added a second region.

In September, the meter counts 640 active workers and 31,800 texts. The usage report the billing team pulled, though, ran on 27 September, three days before the month ended, and shows 27,400 texts. The September invoice bills the platform fee, 390 extra workers at $4.00 and 17,400 extra texts at $0.03, for a total of $4,082.00.

Now rate the month from the meter and the agreement. The worker count is right, 640 less the 250 included, so 390 extra workers, but the amendment sets $4.50 for September usage, not $4.00. That line should be $1,755.00, not $1,560.00, so it is $195.00 under. The texts should be 31,800 less the 10,000 included, so 21,800 extra texts at $0.03, which is $654.00 rather than the $522.00 billed, another $132.00 under. The invoice should total $4,409.00, and it is $327.00 under for the month.

Neither error looks wrong on its own. The rate is the one Pinecrest paid for months, and a lower text count looks like a quieter month. Both also repeat: until someone updates the rate, every month from September on is $195.00 under on workers alone, and a report pulled early cuts off usage every time it happens.

How does usage billing affect revenue at month end?

Usage you delivered in a period is, in most usage contracts, revenue earned in that period, whether or not you have invoiced it yet. The FASB’s revenue standard, ASC 606, treats an amount that can vary with what the customer does as variable consideration (paragraph 606-10-32-6), and a variable amount that relates only to one period’s service can be allocated to that period (606-10-32-40). Where the amount you can bill corresponds directly to the value of what you delivered, such as a fixed price for each unit used, the standard’s practical expedient lets you recognize revenue at the amount you have a right to invoice (606-10-55-18).

That makes the usage report a revenue record, not just a billing input. If it is cut short, revenue and the receivable are understated together. In the Quillfield example, the three days of texts the report missed were delivered in September, so they belong in September’s revenue even if they are billed later. A right to payment that needs only the passage of time is a receivable (606-10-45-4), so the amount you have earned but not yet billed is still money the customer owes.

What do you do when a usage invoice under-bills?

You confirm the gap against the agreement, then bill it, in the way the agreement allows. Recompute the period from the meter and the price on the date, keep the arithmetic with the invoice, and check whether the agreement sets a limit on billing for earlier periods, since some contracts restrict back-billing after a set number of days. Send the customer a short note with a supplementary invoice that names the line, the rate, the quantity and the document behind each, rather than adding an unexplained amount to next month’s bill.

Then fix the cause, not just the invoice. Update the rate from the amendment’s start date, move the usage report to run after the period closes, or add the missing metric to the billing system, so the same gap does not appear again next month. Record any line billed above the agreement as its own figure too, and never net it against the amount under-billed: an over-billed customer may be owed a credit, while an under-billed one owes you money, and the two answer different questions.

How often should you check usage invoices?

Check every usage invoice before it is sent, or at the latest during the month-end close, because a usage gap found before the invoice goes out costs nothing to fix. Found a quarter later, it means a supplementary invoice, a conversation with the customer about the delay and an adjustment to revenue already reported. Checking every invoice, rather than a sample, is what lets you hand your boss or your auditor a record of what was checked and how. For the subscription side of the same question, see why ARR is not revenue, and for revenue that sits on the balance sheet before it is earned, see what the auditor asks about deferred revenue.

How can Ressura check your customer invoices?

Ressura’s Order-to-Cash Module checks every invoice you send against what the customer agreed to pay. It checks each price against the price the customer agreed for that item on that date, applies an amendment from its start date on every line it covers, checks the quantity you invoice against what the customer ordered or what you delivered, and adds every line, subtotal, credit and total up again. It also raises a fee the agreement lets you invoice for a period when no invoice covers that period, and flags any price below the agreed price that has no clause or recorded approval behind it.

It works from what you already have: the agreement or price reference the customer accepted, amendments and price notices, orders and delivery records, the invoices you sent and your credit notes. Any gap is shown in dollars with the arithmetic beside it, and if everything matches, the result says so and lists what was checked. There is nothing to connect to start, and later you can connect QuickBooks Online or Xero so every invoice is checked as you send it.

Sources

1. ASC 606