Taking Payments in the Field: Mobile Processing for Service Businesses

Why does "invoice later" quietly kill service businesses?

The traditional field-service billing loop looks like this: finish the job, leave a paper invoice or email one that night, wait. The customer means to pay. The invoice migrates to a kitchen drawer. Two weeks later the office sends a reminder; week four brings the awkward phone call; somewhere around day 35–45 the check arrives — for work whose payroll, parts, and fuel you funded a month ago.

Multiply that across a season and the damage compounds. A crew doing $60,000/month with a 40-day average collection cycle is permanently floating about $80,000 of completed work — capital that comes from the owner's line of credit, at the owner's interest rate, with the owner's stress. Add the percentage that never arrives at all: industry surveys of trades businesses consistently find that a meaningful slice of invoiced work — often cited in the low single digits of revenue — is eventually written off or collected at steep cost.

Now compare the paid-on-site loop: tech finishes, presents the total on a phone or handheld, customer taps their card, receipt texts itself, money settles next morning (or the same day with same-day funding). Collection cycle: about ninety seconds. The psychology matters as much as the mechanics — the moment of highest willingness to pay is the moment the fixed AC starts blowing cold air. Every day between service and settlement erodes both goodwill and collection odds. Field payments exist to close that gap to zero.

What are your options, ranked?

Ranked from lightest to heaviest — most crews end up running two of these side by side:

1. Tap-to-pay on the phone itself. Modern iPhones and Android phones can accept contactless cards and wallets (Apple Pay, Google Pay) with no extra hardware — the processing app turns the phone's NFC into the terminal. Zero hardware cost, nothing extra to charge or lose, card-present rates on every tap. Limits: no chip slot for cards that won't tap, and no magstripe fallback. As tap adoption keeps climbing (the large majority of newly issued US cards are contactless), this covers more of the wallet every year.
2. Bluetooth readers — Clover Go, BBPOS and similar. A pocket-size reader ($50–$150) pairs with the tech's phone and adds chip dip plus tap, with swipe fallback. This is the workhorse tier: every physical card works, every transaction qualifies card-present, and a lost reader costs $100 to replace instead of $600. Clover Go's advantage for growing companies: it lives in the same Clover account and reporting as the countertop and handheld devices you may add later.
3. Handheld smart terminals. All-in-one Android devices with built-in card reader, receipt printer, camera, and 4G — the tech carries the whole checkout in one unit, no phone pairing to fuss with. Cost runs roughly $300–$700 per device. Best for higher-volume crews, businesses that want printed receipts on the spot, and owners who prefer dedicated hardware over apps on personal phones.
4. Text-to-pay and pay links. No hardware at all: the tech or the office sends a secure link by SMS or email for the exact amount; the customer pays on their own phone. Technically card-not-present (so pricing is a bit higher), but with the customer typing their own card and full AVS data, it prices far better than a keyed-in card over the phone — and it's the right tool for deposits before the truck rolls, unattended completions, and chasing yesterday's invoice.

The practical fleet standard we recommend: a Bluetooth reader in every truck, tap-to-pay as backup on every phone, text-to-pay from the office. Full hardware options are on our mobile swipers and readers page.

What happens with spotty coverage? (Offline mode, honestly)

Rural routes, new-construction sites with no service, metal buildings, basements: coverage gaps are a real feature of field work, and payment vendors' marketing tends to gloss over what "offline mode" actually is.

Offline (store-and-forward) mode encrypts and queues a card transaction on the device, then submits it for authorization when connectivity returns. The honest fine print:

• The transaction is not approved yet. The customer sees "accepted," but no bank has said yes. If the card is declined, over limit, or stolen when the queue processes hours later, that's your loss — the customer left with a working AC either way.
• Set offline limits deliberately. Most platforms let you cap per-transaction and total offline amounts. A sane field policy: enable offline mode with a modest cap (many businesses land somewhere around $100–$250), and for big-ticket jobs in dead zones, walk to the driveway, use the phone's hotspot, or take a deposit by pay link before dispatch.
• Settle the same day. Queued transactions typically must be submitted within a platform-defined window (commonly 24–72 hours) or they expire. Make "reader synced" part of the end-of-day routine, like fueling the truck.

Also know your fallback ladder: if the reader won't connect, tap-to-pay on the phone may still work (different radio, different path); if there's no connectivity at all, a text-to-pay link fires the moment the truck regains bars. The failure mode to design out is the one that ends in "just mail us a check."

How much does typing the card in really cost?

Here's the quiet leak in most field operations: the tech finishes the job, calls the office, and someone types the customer's card number into a virtual terminal. It works — and it's the most expensive way to accept a card, twice over.

First, the rate. Interchange — the card networks' wholesale cost, explained in our interchange guide — prices by how the card is presented. A tapped or dipped card proves the physical card was there; a keyed number proves nothing. Keyed transactions price roughly 0.5–1.0 percentage point higher than the same card presented in person, and keyed entries missing AVS data downgrade to categories that are worse still. Second, the risk: card-not-present transactions carry weaker chargeback protection — in a "that's not my charge" dispute, a keyed transaction with no signature and no chip data is nearly indefensible, while an EMV dip generally shifts counterfeit liability off you entirely (the same logic behind the EMV liability shift).

The math that makes readers free: a crew keying $15,000/month pays roughly $75–$150/month in avoidable interchange penalty — call it $1,200 a year at midpoint. A Clover Go costs about $100 once. The reader pays for itself in the first four to eight weeks, then keeps paying you every month after. If your statement shows meaningful keyed volume and there's a truck at the customer's address anyway, you're donating margin for no reason.

How do tipping and receipts work in the field?

Two small configuration decisions that shape both revenue and professionalism:

Tipping. Mobile checkout flows can present tip prompts exactly like a restaurant tablet — percentages, custom amount, or none. Whether to enable it is a business-culture call: common in mobile detailing, junk removal, moving, and appliance delivery; less customary in licensed trades like HVAC and plumbing, where some owners prefer a clean invoice (and some customers tip in cash regardless). Points to get right if you enable it: configure sane preset amounts for big tickets (20% of a $9,000 install is not a tip prompt anyone wants), decide how tips flow through payroll, and remember tips are taxable wages. Many trades shops split the difference — tip prompts on for small-ticket service calls, off for installs and quotes.

Receipts. Digital-first is the field standard: text or email the receipt from the device, with the business name, tech, line items, and — critically — a billing descriptor that matches your company name so the charge is recognized on the statement. Handhelds with printers cover the customers who want paper (property managers and commercial clients often do, for their own files). Every receipt should also state your warranty/guarantee line — it's free marketing on a document the customer actually keeps, and it's evidence that cuts both ways in any later dispute.

How does card-on-file work for service agreements?

Maintenance memberships — seasonal HVAC tune-ups, quarterly pest control, monthly pool service, annual backflow testing — are the best revenue a service business has: predictable, pre-sold, and route-dense. Billing them by invoice reintroduces exactly the chase-the-check problem memberships were supposed to solve. The right mechanism is tokenized card-on-file: the customer's card is stored in the payment gateway's secure vault as a token, and your systems charge the token on schedule without ever holding the card number.

Run it properly:

• Get written authorization once. A signed or click-through stored-credential agreement stating the amount, frequency, and cancellation terms. It's a card-brand requirement, and it's the document that wins the rare "I never agreed" dispute.
• Let the vault do the billing. The gateway's recurring engine charges each cycle, retries soft declines, and emails receipts automatically. Enrollment happens in the driveway: the tech dips the card once at signup and checks the "save for membership" box.
• Turn on account updater. When stored cards expire or get reissued, network updater services refresh the token automatically — the difference between memberships that quietly lapse and memberships that renew for years.
• Consider ACH for commercial agreements. Property managers and commercial accounts on monthly contracts are often happier on bank draft, and a $1-ish ACH beats 3% of a $2,500 monthly contract by a wide margin — the full comparison is in ACH vs. credit cards for invoicing.

What do real crews actually run? (Texas examples)

Three composite setups that mirror what we deploy across Dallas–Fort Worth and the surrounding counties:

HVAC company, 6 trucks, Plano. Clover Go in every truck paired to company phones; tap-to-pay enabled as backup. Service calls collect on-site; installs take a 50% deposit by text-to-pay at scheduling and the balance by dip at completion. Maintenance memberships bill monthly on card-on-file with account updater. Office keeps a virtual terminal for the rare phone-in payment, with AVS required. Their August reality — 14-hour days, full routes — is exactly when on-site collection matters most: nobody has office time to chase invoices during peak season.

Plumbing outfit, 3 trucks, Fort Worth. Handheld smart terminals with built-in printers — their commercial and property-management clients want paper. Offline mode capped at $200 for dead-zone basements and mechanical rooms; end-of-day sync is on the closeout checklist. Commercial monthly accounts moved to ACH draft, which cut their card fees on contract revenue to nearly nothing. Emergency after-hours calls take a card-on-file authorization by phone-link before dispatch — ending the 2 a.m. "bill me later" losses.

Mobile detailing, 2 vans, Frisco. Pure phone setup: tap-to-pay on each operator's phone, no extra hardware, tip prompts on (15/20/25% — tips run meaningful in detailing), text-to-pay links for the recurring fleet-washing accounts and for no-contact completions at office parks. Receipts by text with a review link attached. Their entire payment stack cost $0 in hardware and set up in an afternoon.

Different trades, same pattern: match the heaviest tool to the highest-volume moment, keep a zero-hardware fallback, and move recurring revenue to card-on-file or ACH.

What does it cost per tech?

Industry-typical numbers for equipping one technician, honestly stated:

Worked example for a 5-tech crew averaging $12,000/month per tech, currently keying half its volume: readers for everyone costs roughly $500 once. Moving that keyed half ($30,000/month across the crew) to card-present saves about 0.5–1.0% of it — $150–$300 every month — before counting faster collection, fewer indefensible disputes, and the invoices that stop aging past 30 days. There is no other $500 in a service business that returns like that.

One warning from the contracts world: field hardware is exactly where non-cancellable equipment leases get pitched, because "$29/month per truck" sounds painless across a fleet. Five trucks × $29 × 48 months is $6,960 for maybe $1,500 of hardware — and the lease survives even if you switch processors. Buy the readers. They cost less than the paperwork.

Frequently asked questions