Two businesses, one phone number
Most trade businesses drift into commercial work rather than deciding on it. A residential customer manages a strip of rental units and asks whether you would look after those too. A property manager finds you through the same search results as everyone else. A general contractor needs a subcontractor at short notice. Within a year you are running commercial jobs alongside residential ones, using the same intake script, the same invoice template and the same technicians, and the numbers get quietly worse without anyone being able to say why.
As of September 2026, this is one of the most common growth traps in the trades. The work itself is often familiar - the same equipment, the same skills, sometimes the same neighbourhoods. What changes is everything around the work: who calls, who approves, who pays, when they pay, what they expect to be told afterwards, and what happens when a technician arrives and cannot get through the door. Treating that as a variation on residential service is how a business ends up with a full schedule and an empty bank account.
This guide separates the two operationally. It covers how the call itself differs and what the intake has to capture, payment terms and the cash-flow consequences of net terms, scheduling windows and site access, pricing structure, work authorisation limits, and the account reporting commercial customers expect to audit. It closes with the part nobody writes about, which is when to turn commercial work down. The platform side of it runs on Run with Jarvis, where call answering, scheduling, dispatch and invoicing sit in one operations layer rather than three.
The call is a different call
A residential service call has one person in it. They own the problem, they own the property, they approve the price and they pay at the end. Your intake needs a name, an address, a phone number, a description of the fault and a time. Five fields and a card at the door, and the job is closed.
A commercial call routinely has three or four people in it who never appear on the same phone call. The person dialling is often a tenant or a shift supervisor who noticed the problem. The person who can approve spending is a property manager or facilities coordinator, possibly at a different address, possibly for a company that owns two hundred doors. The person who pays is an accounts payable department that has never heard of your technician and will only recognise an invoice carrying a work-order number they issued. And the person who will let your van through the gate is a site contact whose phone number nobody has written down.
If your intake script does not know that, the failure is silent and expensive. The technician arrives with a name and an address, cannot reach anyone with a key, waits forty minutes, does the work anyway on a tenant's verbal say-so, and the invoice bounces six weeks later because there was no purchase order and the tenant had no authority to approve anything. Nobody did anything obviously wrong. The intake just captured the wrong five fields.
What a commercial intake has to capture, at minimum, is the site rather than the address - a building name or store number, plus the unit, suite or area within it. The work-order or purchase-order number if the caller has one, and an explicit note if they do not, because that gap is a decision somebody has to make before the technician leaves. The authorised contact for approvals with a direct phone number, not the tenant. The access arrangement, including whether an escort is required and who provides it. And the billing entity, which is frequently not the name of the business on the door.
That is roughly double the residential intake, and it has to be collected consistently at three in the morning by whoever or whatever answers the phone. An AI receptionist that asks a different set of qualifying questions once it identifies the caller as a commercial or managed-property account is doing exactly the job that a tired human at the end of a long day will not do reliably. The same discipline that stops residential price shoppers from consuming your day, covered in handling price shoppers on the phone, applies here in reverse: the goal is to spend longer on the commercial call, not shorter, because the fields you skip are the ones that block the invoice.
Where the two models actually diverge
| Dimension | Residential | Commercial (direct) | Property management account |
|---|---|---|---|
| Intake fields | Name, address, fault, time, phone | Site and unit, fault, work-order number, authorised contact, billing entity | All of commercial, plus tenant contact, escort and access rules, per-property authorisation limit |
| Who approves | The person on the phone | Facilities or operations lead, often not on site | Property manager, against a standing limit per site |
| Payment | Card or link at completion | Net terms against an invoice with a work-order reference | Net terms, batched monthly, often per-property statements |
| Scheduling | Arrival window agreed with the homeowner | Fixed access window, sometimes after business hours only | Access window plus escort availability and tenant notice |
| Pricing | Published flat rates and diagnostic fee | Negotiated rate card, after-hours multiplier, agreed travel | Rate card applied across a portfolio, annual review |
| Reporting expected | An invoice and a receipt | Job detail, parts used, time on site, photos on request | Auditable history per property, spend by site, response times |
Read that table as a description of six separate processes, not one process with variations. Each row is a place where a residential habit produces a commercial failure, and the failures compound: the intake miss creates the approval gap, the approval gap creates the payment dispute, and the payment dispute is the one your bookkeeper finds in a month with no cash in it.
Payment terms and the cash-flow consequence
The single largest difference is not price. It is the gap between finishing the work and holding the money.
Residential service is close to instantaneous. The technician completes the job, the customer pays by card or through a payment link on the spot, and the money settles in a day or two. Your working capital requirement is the cost of parts on the van.
Commercial work is a loan. You buy the parts, pay the technician's wages that Friday, and invoice against terms that may be thirty days from the invoice date, or thirty days from the end of the month in which the invoice was issued, which is a very different thing and worth reading carefully before agreeing to it. Then the clock only starts if the invoice was accepted, which requires the work-order reference, the correct billing entity and often a signed field ticket attached.
Work the arithmetic on your own numbers and treat these figures as an illustration to replace with your own. Say a commercial job bills at $1,400 with parts costing $380 and labour costing $290. You are out of pocket $670 on the day of the job and you recover it forty-five days later on average, once you account for invoices that go out late and payments that arrive after terms. Ten such jobs a month means roughly $6,700 of your money is walking around inside your customers' buildings at any moment, and that figure grows in exact proportion to how well the commercial side is going.
This is why a growing commercial book can starve a healthy business. The revenue line looks excellent. The bank balance does not, because success on net terms consumes cash rather than producing it. Three habits keep it survivable. Invoice the same day, every time, because most late payment is late invoicing wearing a disguise. Make the invoice unrejectable by attaching the work-order number, the field ticket and the parts detail at the moment the technician closes the job rather than reconstructing them later. And reconcile against the accounting ledger continuously rather than monthly, because a receivables ageing report you look at once a quarter is a history lesson rather than a management tool.
There is also a defensible middle ground worth using more than most shops do. New commercial accounts do not have to start on terms. Card at completion for the first several jobs, then terms once there is a payment history and a signed credit application, is a normal and unembarrassing thing to ask for, and the accounts that refuse it are telling you something useful.
Scheduling windows, access and escorts
Residential scheduling negotiates with one person about their day. Commercial scheduling negotiates with a building.
A retail unit may only permit noisy work before opening or after closing. A medical or food-service site may bar work during operating hours entirely. A secured facility requires a badge, an escort and a sign-in that adds twenty unbilled minutes at each end. A multi-tenant property requires notice to the tenant before anyone enters their space. Some sites have a single maintenance contact whose availability, not yours, determines when work can happen.
None of that is exotic and all of it is invisible if it lives in a free-text note on a customer record. The fix is to model access as a property of the site rather than the job. Each site on a commercial account carries its own access window, its own escort requirement, its own contact and its own after-hours rule, and the dispatch board treats those as constraints when it places work. A technician should never be routed to a site during a window in which they cannot enter it, and dispatch should not have to remember which of forty properties is which.
That is a materially harder dispatch problem than residential routing, because you are optimising against fixed windows rather than open days, and it gets harder again the moment you have several technicians and overlapping commitments. The mechanics of running that board are covered in the CRM and dispatch guide, and if the commercial accounts sit across more than one branch, the consistency problems in the multi-location operations guide arrive immediately, because two offices will otherwise invent two different sets of rules for the same portfolio.
Travel is the other quiet scheduling difference. Commercial portfolios are rarely tidy. A property manager with twelve buildings will have three of them at the awkward edge of your map, and they will expect the same response time everywhere. Decide before signing whether those outlying sites are priced for the drive or subsidised by the close ones, using the same logic as your residential boundary in the service radius and drive-time pricing guide. The difference is that on a commercial account you usually cannot decline the far site without losing the near ones, so the rate card has to carry it.
Pricing structure and who agrees it
Residential pricing is published and per-job. You have a flat rate for common work, a diagnostic fee, and a price the customer either accepts or does not at the door. Each job is its own negotiation and it closes in one conversation.
Commercial pricing is negotiated once and then applied hundreds of times. The account agrees a rate card - an hourly labour rate, sometimes tiered by technician level, a parts markup, a defined after-hours or weekend multiplier, a trip or mobilisation charge, and often a minimum billable duration. That card then governs every job for the term, which means every assumption baked into it is repeated at scale.
The estimating discipline is therefore completely different. A residential mispricing costs you one job. A commercial mispricing costs you every job on that account until the annual review, and by then the account is embedded in your schedule and hard to walk away from. Build the rate card from your own loaded cost, add explicit allowances for the things commercial work adds - escort and check-in time, work-order administration, reporting, the cost of financing forty-five days of receivables - and then decide the margin. A rate card that merely matches your residential hourly rate has silently absorbed all of those as losses.
Two clauses are worth insisting on. An annual review date, so the card is not frozen against your cost base indefinitely. And explicit definitions of after-hours and emergency, in hours rather than adjectives, because "urgent" means whatever the caller needs it to mean at eleven at night.
Work authorisation limits
This is the clause that decides whether the invoice gets paid, and most small shops never ask about it.
Almost every commercial and property-management customer operates a spending limit for site-level staff. A maintenance supervisor might approve routine work up to a set amount without escalating; anything above that requires a manager, a purchase order, or in some organisations a competitive quote. Your technician standing in a plant room at seven in the evening has no idea what that number is, and the supervisor telling them to go ahead may genuinely believe they can authorise it.
Capture three things when you open the account and put all three where the technician can see them on the job. The authorisation limit per site. The names and direct numbers of the people who can approve above it. And the required form of approval above the line - verbal will not survive a dispute, so a written confirmation, an emailed purchase order, or a signed field ticket with a printed name.
Then enforce it in the field, which means the technician's job record carries the limit and the escalation contact, and the process for exceeding it is to stop and call rather than to finish and explain. That is an uncomfortable conversation once. An unpaid five-figure invoice is an uncomfortable quarter. The evidence trail that wins a commercial billing dispute is exactly the one that wins a card dispute: what was authorised, by whom, when, and what was actually done, all captured at the time rather than reconstructed from memory two months later.
Reporting, because commercial accounts audit you
A residential customer wants an invoice and, if you are lucky, remembers your name in three years. A commercial account wants a history.
At renewal, or the first time a property manager's own boss asks why maintenance spend is up, somebody will pull your account and want to see spend by property over twelve months, the job history for a specific unit, response times against whatever was promised, recurring faults at a site, and the parts fitted on a particular date. If assembling that takes you a week of spreadsheet work, two things happen. You will do it badly, and you will do it late, and the account will start quietly shopping.
The reporting is not a favour you do for the customer. It is your renewal argument, and it is frequently your best upsell: a report that shows the same unit failing four times in a year is the case for a capital replacement that you are already positioned to quote. It is also, uncomfortably, the record that protects you when a site contact misremembers what was authorised.
None of that is possible if the account's history is distributed across a call log, a scheduling app, a separate invoicing tool and a technician's photo roll. This is the strongest practical argument for all-in-one business solutions over stitched-together point tools in a business with commercial accounts: the report exists because every call, job, part and payment was recorded against the same property record from the start, not because somebody rebuilt it afterwards.
When not to take commercial work
Growth advice rarely includes the word no, which is why so many small trades take on an account that eventually breaks them. There are four honest reasons to decline.
Decline when the terms outrun your cash. If the account would push your average collection period past what your bank balance and credit line can absorb, the work is a liability regardless of the margin. Model the receivable before signing, not after.
Decline when the compliance packet costs more than the account earns. Larger commercial and property-management customers require certificates of insurance, additional-insured endorsements, workers' compensation evidence, licence copies, sometimes bonding and lien waivers per job. That paperwork has a real annual cost in premiums and administration, and a small account that demands enterprise-grade compliance is a net loss dressed as a logo.
Decline when the rate card only works at a volume nobody has committed to. Portfolio pricing is frequently justified with a promise of future work. If the discount is real and the volume is aspirational, you have taken the discount and left the volume with the customer.
And decline, or cap, when a single account would become large enough that losing it would end the business. Concentration risk is the quiet killer in commercial service. An account at a large share of revenue eventually knows it, and the annual review stops being a negotiation.
Making one system serve both
The conclusion is not that commercial work is bad. It is steadier, it schedules further ahead, and one good property-management relationship can fill a slow season that residential demand cannot. The conclusion is that it is a different business, and the failures come from running it on residential habits.
Practically, that means one system holding two shapes of record. A residential customer with a card on file and a simple job history. A commercial account with multiple properties, named authorised contacts, an authorisation limit per site, access and escort rules, a negotiated rate card, work-order references on every job, terms-based invoicing and a job history somebody can audit without a spreadsheet. Same technicians, same calendar, same phone number - different intake, different approval path, different invoice, different report.
Get that separation into the system and the two sides stop damaging each other. Leave it in people's heads and the commercial side will keep generating invoices that cannot be collected and site visits that cannot get through the door. Compare current plans on the pricing page, and if you want a second opinion on whether a specific account's terms and rate card actually work against your job mix, get in touch.
