The cost of a rejected check, end to end

The bank fee for an item that falls out of automated processing is the part of this cost that is written down. It is rarely the largest part, and treating it as the whole number leads organizations to accept a reject rate that is quietly expensive.

What actually happens to a rejected item

An item a reader/sorter cannot read leaves straight-through processing and is handled by a person at the bank. The bank charges for that handling at its own fee schedule, and banks publish an acceptable reject rate above which the charges apply.

That is the first cost. The rest of them happen inside your organization, and none of them are billed to you in a way that makes them easy to count.

1. The bank exception fee. Charged per item above the acceptable rate, at the bank's published schedule. The only line that appears on a statement.
2. Reconciliation time. Somebody has to work out which items were affected, match them against the run that produced them, and confirm what cleared and what did not.
3. Reissue. Where an item has to be produced again, that is a new print run, new stock, new approval, and new handling, for a payment that was already made once.
4. Explanation. In payroll, an item that does not clear means a person was not paid. That conversation consumes time in payroll and in the affected department, and it is the cost nobody schedules.
5. Investigation. When a rate rises, someone has to determine whether the cause is the printer, the toner, the fonts, the stock, or a layout change, which is a project rather than a task.

Putting a rate into item counts

Typical MICR toner in this industry rejects at 2% to 4%. Rates are abstract, so it helps to convert them into the unit your operation actually deals in, which is items.

Those are counts of items that left the automated path. Multiply by your bank's fee for the direct cost, and then multiply by whatever your own reconciliation and reissue work is worth, which is the number most organizations have never calculated.

Why the rate is worth attacking rather than budgeting

Reject rates are frequently treated as a cost of doing business, which is reasonable if the rate is a fixed property of printing checks. It is not. It is an outcome of specific, controllable inputs.

The inputs are the ones we keep returning to. Magnetic toner formulated for the engine it runs in. Calibration verified on a MICR verifier rather than assumed. Certified fonts tuned to that engine and toner. Stock that conforms to ANSI X9.18, including its requirement that the paper contain no ferromagnetic material, stored in its sealed wrapping so it stays at the specified moisture content. Maintenance performed on published page counts rather than when something looks wrong.

Each of those is a small, boring discipline. Together they determine which column of the table above your operation lives in, and in payroll they determine how many times a year somebody has to explain to an employee why a check did not clear.

What to measure

Ask your bank for the exception item count and the volume it was measured against, rather than only the charge. A count and a volume give you a rate. A rate gives you a trend. A trend tells you whether a maintenance kit is overdue or a supplier change had consequences.

Then verify your own output before it leaves the building, because the alternative feedback loop runs through the mail, the bank, and a statement, and it is measured in weeks.

The cost that is hardest to argue about

Every organization that prints checks has a worst case, and it is rarely financial. In payroll it is an employee who was not paid, for a reason nobody in the room can explain on the spot.

That conversation costs the payroll team time, costs the employee's manager time, and costs something less countable in how the payment process is regarded afterwards. It is also the cost most likely to force a rushed decision, because the pressure to fix it immediately is entirely legitimate.

The equivalent in an insurance claims or accounts payable operation is a payee who has to call and ask. The mechanics differ. The pattern does not: an item that did not clear turns into a person waiting, and then into somebody inside your organization doing unplanned work.

What a reduction is worth

Because the range for typical MICR toner is 2% to 4%, the arithmetic of improvement is easy to sketch even without knowing your bank's fee. On 10,000 items, moving from the upper end of that range to the lower end is the difference between 400 items handled by a person at the bank and 200.

Those 200 items are not only 200 fees. They are 200 items removed from reconciliation, from reissue, and from the queue of things somebody has to explain. That is the part of the calculation that never makes it into a comparison of toner prices.

What the achievable end of that range looks like

The 2% to 4% range describes typical MICR toner. It is not a floor. Our ProMark™ MICR toner is independently tested at a 0.1% reject rate or lower, on an IBM 3890 high-speed reader/sorter.

Run that through the same arithmetic. On 20,000 items a month, 2% is 400 items handled by a person at the bank and 0.1% is 20. The difference is 380 items a month that never enter an exception queue, never generate a fee, and never become a phone call from somebody who was not paid.

We give the equipment and the tester because a reject rate without a source is not a specification. Ask any vendor quoting a number for the same two details.

The same arithmetic runs the other way when an operation drifts. A printer moving toward the edge of tolerance does not announce itself, and the first evidence is usually a charge that is larger than last period for reasons nobody has investigated. By the time it is large enough to prompt a question, several runs have already been mailed.

Fee schedules and acceptable reject rates are set by each bank. The 2% to 4% range describes typical MICR toner in this industry, not a guarantee about any particular account or printer.