Slow AP costs you twice: the discounts you miss and the late fees you pay

Most AP cost analysis stops at processing cost per invoice. That is the visible number and it is not the largest one.

The larger cost is on both sides of the payment date: discounts you were offered and could not take, and late fees you paid on invoices that were approved after the due date.

The discount arithmetic is better than it looks

The standard term is 2/10 net 30. Two percent off if you pay within ten days, otherwise the full amount at thirty.

Read as an interest rate: you are paying 2% to hold the money an extra twenty days. That is 2% for 20 days, or roughly 36% annualized.

There is essentially no use of working capital in property management returning 36%. Which means an offered 2/10 discount that goes untaken is not a rounding error, it is a deliberate choice to borrow at 36%, made by default rather than by decision.

Even 1/10 net 30 works out to roughly 18% annualized. Still better than any alternative use of the cash.

The reason nobody takes it is cycle time, not policy

Nobody decides to skip discounts. They become unreachable because of arithmetic in the process.

A ten-day window starts at the invoice date, which is usually before you receive the invoice. Two to four days lost in the mail or in a vendor's email queue. A day or two before somebody opens it and enters it. Then approval, which is where the time actually goes.

By the time an invoice with a 2/10 term reaches a payable state, day ten has passed. The discount was never available at your current cycle time regardless of intent.

Which reframes cycle-time improvement. Cutting three days off approval is not just an efficiency story about staff hours. It is what makes a 36% annualized return reachable at all.

Late fees are the same problem with the sign flipped

The mirror image, and easier to quantify because you already paid it.

Utility invoices carry late fees, commonly 1.5% per month on the unpaid balance. Municipal utilities frequently add a flat penalty and, in some jurisdictions, a shutoff path that generates a reconnection fee on top.

Vendors with contracted terms charge interest on past due amounts. Most operators never look at whether that line is being applied.

The number worth pulling: total late fees and finance charges paid across the portfolio last year, from the GL rather than from memory. It is a single query against the accounts you post those to, and it is usually larger than anyone expects because it arrives $40 at a time across hundreds of invoices.

Trusted vendor bypass is the specific lever for utilities

Utility invoices are high volume, low variance, recurring, and the largest source of late fees. They also do not need the approval chain a capital invoice does.

Routing routine utility bills and trusted vendors around the approval chain removes the stage that was consuming the days, for exactly the invoice population where approval was adding no information. A regional approving a $340 electric bill for a property they manage is a signature, not a control.

Two guardrails worth having. Bypass rules should not apply to purchase orders, because a PO-backed invoice needs its match checked regardless of vendor. And bypass should be revocable per vendor rather than permanent, since a trusted vendor whose invoices start varying is no longer a routine one.

Whether to chase discounts is a real decision, not a given

Being fair to the other side of this, because "always take the discount" is too simple.

Taking a 2/10 discount means paying twenty days earlier, which pulls cash forward. For an operator managing to a tight cash position, or drawing on reserves with a lender approval step, twenty days is not free.

The honest framing is that a discount is a 36% annualized return available if you have the cash. If you do not, declining it is a liquidity decision, and it should be recorded as one rather than happening because approval was slow.

The distinction matters because those two situations call for different fixes. One is a treasury conversation and the other is a workflow problem.

Measure three things

Discount capture rate. Of invoices offering a discount, the share where you took it. Most portfolios cannot compute this because discount terms are not captured as structured data at intake, which is itself the finding.

Late fees and finance charges paid, from the GL, for the trailing twelve months.

Speed to submit, from receipt to payable, with the distribution rather than the average. The average hides that most invoices clear fast and a tail sits for weeks.

Those three together put a dollar figure on cycle time, which is what turns a process improvement into something a CFO funds.

The context on our own numbers

ResProp eliminates $150K or more in annual AP costs with BillRoute, and 40 hours a month of staff time. Being precise about what that measures: it is processing cost and labor, not discount capture.

Discount capture and late fee avoidance are additional and we do not have a clean portfolio-wide figure for them, because most customers were not measuring discount capture before the change and so there is no baseline to compare against. Rather than model a number and present it as measured, the honest version is that the mechanism is visible in the cycle time data and the dollar effect depends on your vendor terms.

What we do not do

We do not manage your cash position or decide when to pay. Payment scheduling is a treasury decision and BillRoute Pay executes what you approve rather than optimizing your timing.

We do not negotiate vendor terms. If your vendors do not offer discounts, none of the first half of this post applies to you, and the late fee half still does.

Run one query

Pull total late fees and finance charges from the GL for the last twelve months, across the portfolio.

That number is real, already spent, and directly attributable to cycle time. It is also the easiest business case you will ever build.

What was it?