HVAC Software for 6 to 15 Techs: What Breaks When You Add Trucks

Four things break, and they break in a predictable order. First the schedule: the owner or office manager who was dispatching by memory can no longer hold ten techs, their skills, their locations, and the day's emergencies at once, so jobs start getting assigned by who called in last rather than who is closest and most qualified. Second the phone: call volume outgrows the person answering it, and calls start rolling to voicemail during exactly the hours you make money. Third the money: with five techs you could eyeball whether a job made money, and with twelve you cannot, so job costing quietly disappears and you find out how the quarter went when your accountant tells you in April. Fourth the consistency: three techs trained by the owner all sell the same way; twelve techs trained by whoever was available do not.

The tell that you are in this window is that revenue per technician stops rising and starts falling. A well-run residential shop generally produces somewhere between $180,000 and $250,000 per technician per year, with top-quartile operations pushing $250,000 to $400,000 and up. If you added two trucks and total revenue rose while revenue per tech dropped, you did not grow — you bought capacity and then failed to sell it. That is the single most useful number to watch through this entire transition, and most shops in this window are not tracking it monthly.

The important thing to be clear about is that only some of this is a software problem. A dispatch board does not fix a shop with nobody assigned to run it. Marketing automation does not fix a booking rate problem in the phone room. What software does at this size is make coordination possible for one person that would otherwise take three, and make the numbers visible enough that you can manage on them. Buying a platform and hoping it will supply the discipline is the most expensive mistake available in this window, and it is a common one — the shops that get burned by an enterprise rollout are almost always shops that bought a system instead of building a process.

One more structural change nobody warns you about: at this size you are no longer buying software for yourself. You are buying it for a dispatcher, a CSR, an office manager, and a dozen techs, most of whom did not choose it and some of whom will actively route around it. Adoption becomes the deciding variable. A platform your techs actually use at 90% beats a better platform they use at 50%, every time, and it is not close.

Everything in this window comes back to one metric. A technician you pay for eight to nine hours is billable for far less than that — industry measurement puts it around 5.7 billable hours out of 8.8 paid, and utilization targets typically sit in the 65% to 85% range. Wrench time, measured strictly, runs 55% to 65% in a typical shop and 75% to 82% in best-in-class operations. The gap between those two states is the entire game at your size.

Drive time is the largest single piece of that gap. Technicians spend roughly 28% of the workday driving — about two and a half hours on a nine-hour day. That number is not fixed: a tightly dispatched operation with clustered routing runs 18% to 20%, while a sprawling service area dispatched ad hoc runs 32% to 35%. The difference between those two, across ten trucks, is on the order of a full extra truck of capacity that you already own and are already paying for.

Run it in dollars so it stops being an abstraction. Recovering one billable hour per technician per day, at ten technicians and a conservative $150 effective hourly rate, over roughly 250 working days, is $375,000 a year. Good dispatch and routing discipline typically recovers one to two hours per tech per day at this size. Against that, the entire software stack in this guide — even the expensive version — is a rounding error. This is why the buying question at ten trucks is almost never "can we afford this" and almost always "will we actually run it."

To manage the number you have to see it, which means the platform has to timestamp reality: dispatched, en route, on site, complete. If your techs are not clocking those transitions on their phones, you do not have data, you have anecdotes. Start collecting those timestamps before you buy anything else, because they are what tells you whether your problem is routing, scheduling, parts availability, or a couple of technicians who are quietly running four-hour tune-ups.

And watch what you incentivize. Utilization pushed too hard produces rushed jobs and callbacks, and a callback is the most expensive hour in the business — an unbilled truck roll plus a damaged customer relationship plus, often, a review you will spend six months outrunning. Track callback rate alongside utilization or you will optimize yourself into a hole.

This is where the money is actually decided, and it is the one calculation vendors will not volunteer. Platforms in this category price three different ways, and the right answer flips depending on your ratio of office staff to field techs.

ServiceTitan prices per technician — commonly quoted in the $250 to $500 per tech per month range depending on modules, with a 12-month minimum and an implementation fee that is typically reported anywhere from $5,000 to $50,000. At twelve techs that is roughly $3,000 to $6,000 a month, plus onboarding, plus the Pro modules for marketing, phones, and pricebook that are sold separately. Service Fusion prices per company with unlimited users: Starter at $192 a month, Plus at $298, Pro at $489 — the same price whether you have four office staff or fourteen. FieldEdge sits in the middle, custom-quoted around $100 to $250 per user per month. Workiz is flat-tier at roughly $225 Lite, $295 Standard, and $495 Ultimate, with a VoIP phone system built in rather than bolted on. Housecall Pro tops out around $329 a month at MAX with per-seat charges above the included seats.

The crossover is not subtle. A shop with twelve techs and five office people paying per-user pricing is paying for seventeen seats; the same shop on Service Fusion Pro pays $489 total. That is a five-to-ten-times difference in monthly software cost, which over three years is real money — enough to fund a truck, or a dispatcher, or a year of the marketing that would fill both. If your ratio of non-field users to field users is high — a lot of office staff, seasonal helpers, part-time CSRs, an owner and a GM and a bookkeeper who all need logins — flat pricing is a structural advantage and you should weight it heavily.

The counter-argument for per-tech pricing is honest and worth stating: at ten-plus trucks, ServiceTitan's reporting, pricebook, and dispatch depth genuinely produce revenue that cheaper platforms do not. Good-better-best presentation discipline enforced at the tablet, membership tracking that does not leak, and call-to-close attribution you can actually trust are worth real percentage points on average ticket and close rate. The question is not whether it is better. It is whether the incremental revenue it produces at your size exceeds a $40,000-to-$70,000-a-year all-in cost difference — and whether you have the office discipline to use the parts you are paying for.

Two costs almost every shop in this window forgets. Implementation is not just the fee — it is 100 to 300 hours of your office staff's time during a rollout, in a business that does not pause. And card processing at roughly 2.6% to 3.5% is frequently the largest line in the entire software budget: on $3 million of card revenue that is $78,000 to $105,000 a year, dwarfing every subscription combined. Compare effective processing rates across platforms with the same energy you compare monthly prices, because a half-point difference at your volume outweighs the entire monthly plan cost.

The most common version of this question comes from an owner at nine trucks who has been told by three coaches and a mastermind group that he has outgrown Housecall Pro. Sometimes that is true. Often what he has outgrown is his own dispatch process, and moving that process into a more expensive system just makes the same mess cost more per month.

The honest triggers for enterprise-tier platforms are specific. You are running more than one profit center — service, install, and maintenance — and need them costed separately rather than blended. You have a real call center with two or more CSRs and need call recording tied to booking rate by person. You are running paid marketing at a scale where per-campaign revenue attribution changes the budget. You have a membership base in the four figures that needs automated renewal and visit scheduling. You have a GM or ops manager whose actual job is running the system. If three or more of those are true, the step up is defensible. If one is true, fix that one thing where you are.

The counter-triggers are just as specific: nobody in the building owns the system, your pricebook is not built, your technicians resist the tablet, or the reason you are shopping is that a competitor bought it. Enterprise platforms punish weak process rather than substituting for it, and a failed rollout costs you the implementation fee, a year of contract, and — the part that hurts most — a demoralized office that will resist the next change you propose.

The middle path is underrated and it is where most shops in this window belong. Service Fusion at $298 to $489 flat, Workiz at $295 to $495 with the phone system included, or FieldEdge quoted per user, all deliver competent dispatch, job costing, and maintenance agreement management at a fraction of enterprise cost. FieldEdge in particular is HVAC-native with strong agreement handling; Workiz makes the most sense if your phone system is the mess and you would rather buy one thing than integrate three. Housecall Pro at MAX remains legitimate through about twelve trucks for a clean residential service operation — the ceiling is real, but it is higher than the upgrade pitch suggests.

If you do step up, negotiate like it is a truck purchase, because the numbers are comparable. Implementation fees are discountable, module bundles are discountable, and the end of a vendor's quarter is a materially better time to sign than the middle of one. Ask for the total three-year cost in writing including implementation, per-tech growth, module add-ons, and processing rates. A vendor who will not put that on paper is telling you something useful.

At this size dispatch stops being a scheduling task and becomes a capacity allocation job. The dispatcher is deciding, all day, which technician is worth sending to which opportunity — and that decision is worth more per day than most of what happens in the office. Treat it as a skilled role, not as reception with a map.

Staffing ratios vary more than the internet suggests. A pure dispatcher handling only assignment can cover fifteen to twenty technicians; a dispatcher who is also answering inbound calls, chasing parts, and handling customer escalations saturates somewhere between six and ten. Most shops in this window discover they do not need a second dispatcher — they need to stop putting three jobs on one dispatcher. Splitting the phone from the board is usually the cheapest capacity you can buy, and it typically comes before any software purchase.

On routing, the built-in board in most mid-tier platforms is adequate at six trucks and starts creaking around ten, particularly if your service area is geographically spread. The honest options are to upgrade the platform, or to bolt on dedicated optimization: Routific runs $49 to $149 per vehicle per month and produces genuinely better multi-stop routes than entry-level dispatch boards, working alongside whatever CRM you already run. At ten trucks that is $490 to $1,490 a month against a drive-time reduction worth considerably more — but only if somebody enforces the routes. Dispatchers override optimized routes constantly, usually for good reasons and sometimes out of habit, and an override rate above about 20% means you are paying for routing you are not using.

Some practical discipline that costs nothing. Cap drive time between appointments at 30 minutes and treat exceptions as decisions rather than defaults. Zone your service area and assign techs to zones for the day instead of scattering them. Hold one or two flex slots per day for the emergency calls you know are coming, so an August afternoon does not blow up an optimized board. Batch maintenance visits geographically — they are the most schedulable work you have and the easiest to cluster, which is exactly why a healthy membership base makes routing better and not just revenue steadier.

And put parts in the picture, because parts runs are the invisible drain at this size. A tech driving to the supply house mid-day is a billable hour converted into windshield time. Truck stock discipline, a parts runner once you are past eight or nine trucks, and inventory tracking in the platform typically return more hours than another round of route optimization does.

Every call your shop answers is already paid for — you spent the marketing money to make the phone ring. What happens in the next four minutes decides whether that spend produced revenue, and at this size it is being decided by employees whose performance you are probably not measuring.

The benchmarks are stark. Booking rate on answered calls averages in the low-to-mid 40s across the trades by some measures and 65% to 75% for residential CSRs by others, with top performers above 85% and underperformers below 50% — the spread matters more than the exact number. Moving a shop from 50% to 70% on 400 monthly calls is 80 additional booked jobs a month. At a $450 average service ticket that is $36,000 a month in revenue from calls you were already paying to generate, with no additional marketing spend and no additional trucks.

Getting there is unglamorous: record calls, score them weekly against a simple rubric, coach one behavior at a time, and post booking rate by person where the team can see it. The three behaviors that move the number most are asking for the appointment directly instead of quoting a price and hanging up, offering two specific time windows rather than asking when the customer is free, and handling the price question with the diagnostic fee and the value of the visit rather than an apology. This is the single highest-return coaching in the business and it requires no software beyond call recording — which is exactly why platforms with built-in phone systems, like Workiz, or enterprise phone modules are worth more at this size than at any smaller one.

Then there is the coverage problem, which is different from the conversion problem. Shops in this window typically miss 25% to 40% of inbound calls — after hours, during the lunch overlap, and worst of all during the August and January surges when every missed call is a replacement opportunity. A homeowner with no cooling does not leave a voicemail; they call the next shop. An AI receptionist that answers overflow and after-hours calls runs $49 to $500 a month depending on volume and integration depth, and at this size the integrated options matter — Numa at roughly $249 to $379 a month and Avoca in the $300 to $800 range are built to sit on top of a real dispatch board rather than just take a message.

Do the comparison honestly against staffing. An in-house CSR is a $55,000 to $72,000 a year loaded cost and covers roughly one shift, five days. AI coverage handles nights, weekends, and the surge minute when four calls arrive at once, at a fraction of that. The right answer at ten trucks is usually both: humans on the primary line during business hours because they book better, AI behind them so that nothing rings out. What you should not do is put AI on the primary line to avoid hiring a CSR — a good CSR out-books an AI on the calls that matter most, and the replacement calls are the ones that matter most.

Job costing stops being optional. At five trucks you can feel whether the month was good; at twelve you cannot, and shops routinely discover that an entire profit center has been running at a loss for two quarters. You need labor, materials, and overhead allocated per job, and you need it visible weekly, not at tax time. This is the capability gap that most legitimately justifies a platform upgrade in this window — more than dispatch, more than routing.

Reviews become an operational system rather than a personal habit. The owner who asked every customer personally cannot ask on behalf of twelve techs. Automated requests at invoice close — available in essentially every platform here — plus a dedicated tool like Podium at around $399 a month or NiceJob at $75 to $149 covers a single-location shop well. Birdeye at roughly $299 to $1,200 a month becomes the right answer specifically when you cross into multiple locations or multiple brands and are managing several Google Business Profiles at once; below that it is more platform than a single-location shop needs. Watch review velocity by technician, because it is one of the best proxies you have for who is actually delighting customers.

Technician consistency becomes a real line item. Twelve techs with twelve different diagnostic standards produce twelve different close rates, and the spread between your best and worst tech on the same job type is usually far wider than owners expect. Measured diagnostics with MeasureQuick at $30 to $120 per tech per month turn "it is getting old" into evidence a homeowner can see, and structured training like Interplay Learning at $35 to $150 per user per month shortens the ramp on the apprentices you are now hiring continuously rather than occasionally. Both matter more at twelve trucks than at four, because at four the owner is still riding along and setting the standard in person.

Your customer database becomes an asset instead of a list. A shop in this window typically has several thousand past customers, and that database will reliably outproduce cold paid marketing per dollar — targeted campaigns to aging equipment, expired agreements, and declined estimates are the highest-yield marketing available to you. Platform-native campaign tools handle the basics; dedicated database marketing tools like Arch, or an agency like Scorpion for shops running significant paid media, make sense once the list is big enough that manual segmentation stops being practical.

Finally, integration debt starts compounding. Every tool you add is another sync to maintain, another place data goes stale, and another vendor to blame when the invoice does not match the job. The rule at this size is that a new tool must either integrate natively with your platform of record or replace something. Adding a fifth disconnected system with a Zapier bridge is how shops end up with two versions of the truth about what a job cost.

Two realistic stacks for a ten-truck shop. The lean version: a flat-priced platform at $300 to $500, AI phone coverage at $250 to $400, QuickBooks at $90 to $200, routing at $500 to $1,000, reviews at $75 to $400, and diagnostics at $300 to $600 — roughly $1,500 to $3,100 a month, about 0.6% to 1.2% of revenue for a shop doing $3 million. The enterprise version: ServiceTitan at $3,000 to $6,000 with modules, plus the same accounting, diagnostics, and training layers — $4,000 to $8,000 a month, plus a one-time implementation of $5,000 to $50,000 and several hundred hours of internal time.

Both are defensible. Neither is defensible if the recovered billable hours do not show up, which is why the sequencing below front-loads measurement and puts the expensive decision last.

Days 1 through 30, measure before you buy. Turn on the dispatch timestamps you already have, and establish four baselines: revenue per technician, billable hours per tech per day, calls answered versus calls received, and booking rate on answered calls. Record phone calls if you are not already. Do not purchase anything this month. Most shops find at least one of these numbers is materially worse than they assumed, and it is usually the phone.

Days 31 through 60, fix the free things. Split the phone from the dispatch board if one person is doing both. Zone your service area and hold flex slots for emergencies. Start weekly call scoring with your CSRs and coach one behavior at a time. Set a 30-minute drive-time cap between stops. Add AI coverage for after-hours and overflow — it is the fastest payback available and it does not require changing your platform. These changes cost a few hundred dollars a month and typically move the numbers more than a platform migration does.

Days 61 through 90, make the platform decision with data instead of anecdotes. You now know your baselines and you know which of them your current platform is genuinely blocking. Run the per-user versus flat math on your actual seat count including office staff. Get three-year total costs in writing including implementation, growth, modules, and processing rates. If you migrate, do it in the shoulder season — never in July or January — and give your office six weeks of parallel running. And whatever you decide, re-check the same four baselines 90 days later. If revenue per technician has not moved, the software was not the problem, and the next purchase will not fix it either.