Card Authorization Holds for Service Work

An authorization hold puts money aside on a customer's card without taking it. The bank confirms the funds and reduces the available balance, but nothing moves until the business captures the payment. In service work that gap is the whole point: the customer approves the job today, the crew shows up Thursday, and the charge only lands after the work is done.

Authorization hold in one paragraph

When a card is authorized, the issuing bank sets the amount aside and returns an approval code. The merchant then has a limited window to capture that approval and turn it into a real charge. If the merchant never captures, the hold falls off and the money returns to the available balance on its own. Nothing was ever charged, so there is nothing to refund. That distinction matters, and it is the single thing most customers get wrong when they see a pending line in their banking app.

The seven day clock nobody can extend

Card networks do not let an authorization sit open forever. For card-not-present transactions the practical ceiling is about seven days, and some card types are shorter. Nobody can extend it: not the processor, not the business, not the platform. If the work slips past the window, the hold expires and the job has to be re-authorized from scratch.

• A hold is not a charge. Expiring costs the customer nothing.
• The window is roughly seven days, and it starts at authorization, not at scheduling.
• A capture after the window fails, so the safe practice is to capture with hours to spare.
• If the hold expires, the customer authorizes again when the new date is confirmed.

Why not just charge upfront

Charging upfront is simple and it is right for some work: parts you have to order, deposits on large projects, anything where the business is out of pocket before the truck rolls. The cost is trust. A first time customer paying in full for a Thursday cleaning has no leverage if Thursday goes badly, and the business absorbs the dispute risk when they ask for the money back. Refunds are also slower and messier than a hold that simply never became a charge.

Why not just invoice after the job

Invoicing after the fact protects the customer and exposes the business. The work is done, the crew is paid, and collection becomes a phone call. For small residential jobs the collection cost can exceed the margin on the job. An authorization taken before the visit removes that risk without taking the money early.

The middle path: authorize, complete, review, capture

Pinstripe Fleet supports a service-gated hold. The customer authorizes the amount when they approve the job. The business records a date the work will be complete by. When the job is marked complete, the customer gets an email with the details and any photos, and a review window opens. If they say nothing, the hold is captured and a receipt goes out. If they raise an issue, the charge is paused while the business sorts it out.

• Authorization placed: the customer sees a pending amount and gets an email that explains it is not a charge.
• Work completed: the business marks the job done, with proof attached.
• Review window: 24 hours for the customer to object, set at the platform level so it is the same everywhere.
• Capture or release: silence means the charge goes through; an objection pauses it.

Why the review window is 24 hours for everyone

A fixed window is a feature, not a limitation. If every business set its own, customers would face 24 hours with one vendor and 72 with the next, and nobody would know what the pending line on their statement means. A single platform rule also keeps the capture safely inside the seven day authorization window, which is the constraint that actually breaks payments when it is ignored.

What happens when the job slips

Weather, parts, and sick crews happen. When a hold is approaching its limit the business gets a warning email while there is still time to finish and capture. If the work still does not happen, the hold is released and the customer gets an email saying so in plain terms: nothing was charged, the pending amount clears on the bank's own schedule, usually within 5 to 10 business days, and a new authorization is needed if they still want the work.

"The customer never has to ask what the pending charge is. They get told before it appears, when the work is done, and again if it goes away."

Honest tradeoffs

• Payout timing shifts. Money arrives after capture, not after approval, so cash lands a few days later than charging upfront.
• Jobs scheduled more than about five days out do not fit a hold. Those should be charged at completion or deposited upfront.
• Some banks display holds badly, so customer communication has to be explicit and repeated.
• Someone has to mark the job complete accurately. The whole model rests on that timestamp being real.

Who this is right for

Recurring residential work, one-off cleanings, lawn and snow visits, small repairs, and any job booked within a few days of the visit. It is the strongest fit where the customer is new, the ticket is a few hundred dollars, and both sides want a reason to trust each other on the first job.

How to explain it to a customer in one line

Say this: we are placing a hold, not a charge; you will see it as pending; after the work is done you get 24 hours to tell us if something is wrong; if we do not hear from you, the hold becomes the payment and we send your receipt. Customers rarely push back on that sentence, because it is the deal they wanted anyway.