How Do I Get My HVAC Techs to Sell Maintenance Agreements?

Signals you actually need this

The clearest tell is a widening gap between your repair revenue and your recurring revenue. If your service board is full every July and empty every October, you don't have a demand problem — you have a membership problem, and the membership problem lives in the truck. Shops that track it usually find agreement attachment clustered in a narrow band: two or three techs converting a meaningful share of eligible calls, and the rest converting almost none. That spread is the signal. It means the skill exists inside the building and simply isn't distributed.

A second signal is the shape of your objection log. When a service manager asks a low-converting tech why the agreement didn't get offered, the honest answer is rarely "the customer said no." It's "I didn't bring it up." Non-offer, not rejection, is the dominant failure mode. If you can't distinguish offer rate from close rate in your reporting, you're blind to which half is broken — and coaching the wrong half wastes ride-along hours. Offer rate is a behavior you can mandate. Close rate is a skill you have to build. Treat them as separate lines.

Third: check whether your dispatch and CSR desk ever mentions the plan before the truck rolls. Plenty of shops discover the tech is being asked to introduce a recurring commitment cold, at the exact moment the customer just got handed a $900 repair invoice. That's the worst possible sequencing. If the front desk never seeds it, the tech is carrying the whole lift alone, and your attachment numbers reflect that structural handicap rather than tech ability.

Fourth: look at what your techs actually get paid for. If the spiff schedule pays on repair revenue and pays a flat, small amount per agreement, you have told the crew — in the only language that matters — that agreements are a hobby. A tech optimizing rationally will chase the bigger number. Nobody is being lazy. The comp plan is doing exactly what it was designed to do, and it was designed badly.

Fifth signal, and the one owners miss most: turnover in the office. When a service manager leaves, informal accountability walks out the door with them, and attachment rates decay quietly over the following quarter because nobody is running the huddle anymore. A written, published matrix survives that transition. A manager's memory does not. That's the argument for making this a system rather than a personality.

Finally, watch for a plateau after a strong quarter. Shops that run a one-time membership push often see a spike, then a slow regression to the prior baseline within eight to twelve weeks. A spike without a scoring system is a campaign, not a change. If your last three membership pushes each produced a bump that faded, the missing ingredient is a persistent measurement layer — the thing that keeps the behavior in front of the tech after the enthusiasm wears off.

What good looks like versus what bad looks like

Bad looks like a single number. The shop tracks revenue per ticket or average invoice, ranks techs on it, and calls that a scorecard. The problem is that one number can be maxed by a tech who is superb at diagnosing and closing repairs and completely uninterested in anything recurring. He tops the board while your membership base flatlines. Single-metric measurement doesn't just fail to encourage agreement selling — it actively rewards ignoring it, because the hour spent explaining a plan is an hour not spent on the next repair.

Bad also looks like exhortation. A monthly meeting where the owner says "we need more agreements," followed by no change to what gets measured or what gets paid, produces about two weeks of lift. Techs are pragmatic. They allocate effort toward whatever their pay stub responds to.

Good looks like a matrix with eight or nine lines on it. A workable HVAC set: diagnosis quality, repair close rate, agreement offer rate, agreements sold, IAQ and accessory attachment, financing presented, average ticket, callback rate, and post-call review captured. Each line carries a weight your service manager sets. Each tech gets a 1-to-5 level on each line. The composite is the sum of weight times level across all of them. A tech at level 5 on diagnosis and repair but level 1 on agreements, IAQ, financing, and reviews scores in the middle of the pack, not the top — and the matrix makes exactly where he's losing points visible to him without a confrontation.

Good also means the offer line and the sold line are separate. Offer rate is largely a compliance behavior: did the agreement get raised on an eligible call, yes or no. You can reasonably expect that near 100% and coach it hard, because it costs nothing but a sentence. Sold rate is a genuine skill with a natural ceiling well below that. Blending them into one metric lets a tech hide a non-offer behind a "the customer wasn't interested" narrative that nobody can audit.

Good means the weights are yours and they move. Launching a new membership tier before allergy season? Raise the IAQ and agreement weights, drop repair-ticket weight slightly, and republish the matrix. The crew re-aims the next morning without a meeting. Quarterly re-weighting is typical; overnight re-weighting for a seasonal push is the whole point of running weights instead of hard-coded targets.

Good means it's published. A matrix nobody sees is a private opinion. Post it in the bay, push it to phones, review it in a five-minute huddle. The behavior change comes from a tech being able to see his own gap and the distance to the next level before his next call — not from a quarterly review three months after the misses happened.

One more marker of a good system: it accounts for call type. Not every dispatch is agreement-eligible. A warranty callback on a two-week-old install is not the moment. If your matrix penalizes a tech for not offering on calls where offering would be tone-deaf, the crew will correctly conclude the scorecard is dumb and stop respecting it. Define eligibility up front — typically a completed diagnostic or repair on a unit the customer owns, no open dispute — and score offer rate only against that denominator.

Real cost, real ROI, and where the money actually comes from

Start with what an agreement is worth, because the case for the scorecard rests on it. Residential HVAC maintenance plans commonly run in a broad band of roughly $150 to $400 per year for a single system, with multi-system households scaling up from there. The direct margin on the plan itself is thin — you're sending a tech out one or two times a year for a tune-up, and labor eats most of it. That's not where the return is.

The return is in three downstream effects. First, retention: a member calls you rather than the first search result when something breaks, so you stop re-buying that customer through lead spend every time. Second, off-season utilization: scheduled maintenance visits are the inventory you use to fill October and March, the months that otherwise force you to carry idle labor or cut hours and lose techs. Third, the visit itself is a qualified sales opportunity with zero acquisition cost — a tech standing in front of a fifteen-year-old system on a scheduled tune-up is the cheapest replacement lead your business will ever generate.

Model it honestly for your own shop rather than trusting an industry average. Take your current member count, your observed annual attrition on that base, your average member ticket over a year including repairs and replacements, and compare it to the same figures for a non-member customer. The delta per member per year is your real agreement value. Most operators who run this exercise find the plan fee is the smallest component by a wide margin.

Now the cost side of building the scoring system. The lightest version is free: a spreadsheet with KPIs down the rows, techs across the columns, weights in a header row, and a SUMPRODUCT formula rolling the composite. Real cost is your service manager's time — call it a half-day to build and twenty to thirty minutes a week to maintain. The failure mode is staleness. A sheet that stops getting updated during a busy August never restarts in September.

Above that, purpose-built scorecard and sales-performance platforms sit in a range. Gamification and leaderboard tools commonly land in the low tens of dollars per user per month. Commission and attainment tracking tools start with free tiers and climb from there. Enterprise incentive-compensation platforms — the ones that model and pay complex multi-component plans across many branches — price by quote and are genuinely priced for scale. CRM-hosted custom scorecards are viable if you're already standardized on the platform, but you build the matrix yourself; the software supplies the data and the dashboard, not the method.

Pick where the teeth live before you pick a vendor. If your problem is visibility, buy a leaderboard. If your problem is that the comp plan actively points away from agreements, fix comp first and skip the leaderboard entirely — no amount of TV-screen recognition beats a paycheck that rewards the opposite behavior. Most shops under about fifteen techs get further, faster, by fixing the spiff schedule and running the matrix on a shared sheet than by buying software.

On spiff design, a few practical trade-offs. A flat per-agreement dollar amount is simple and techs understand it instantly, but it's easy to game with low-quality sales that churn in month three. Tiered spiffs that step up past a monthly threshold drive push at the end of the period and create a lull at the start. A composite-linked bonus — the tech's monthly bonus multiplier keyed to the weighted score — is harder to explain but far harder to game, because there's no single lever to pull. A common compromise: small immediate per-agreement spiff for the dopamine hit, plus a monthly bonus scaled to the composite for the real money.

Budget for coaching time as a line item, not an afterthought. A ride-along costs you a productive half-day of a senior tech plus the junior tech's slower route. If a low-converting tech's attachment moves meaningfully after two or three ride-alongs, that's an obvious return — but only if you actually schedule them. Most matrices fail not because the math is wrong but because nobody blocked the calendar for the coaching the matrix identified.

Finally, count the clawback. Agreements sold under pressure to customers who cancel in ninety days cost you the spiff, the admin, and some goodwill. If you're paying on agreements, pay on agreements that survive a defined window — thirty to ninety days is typical — or you'll have optimized for signatures rather than members.

How this plugs into your existing workflow

The matrix doesn't replace anything. It sits on top of the field-service software you already run and reads from records the techs are creating anyway. The integration question is simply: which of the nine KPI lines can be pulled automatically, and which need a human to score?

Pull automatically what your system already stores — agreements sold, invoice totals, add-on line items, financing applications submitted, callback tickets tied to a prior visit, and review links sent. Score by hand what software can't see: diagnosis quality and, critically, offer rate. Offer rate has no natural system-of-record unless you create one. The cheapest fix is a required checkbox or dropdown on the invoice: "Membership discussed — sold / declined / not eligible." One tap. It converts an invisible behavior into a countable one, and that single field change often does more for attachment than a training day.

Sequence the workflow so the tech isn't carrying it alone. The CSR seeds it on the booking call — a single line noting that most customers on this repair end up joining the maintenance plan. Dispatch flags whether the customer is already a member so the tech doesn't pitch someone who joined last spring, which is embarrassing and erodes trust. The tech offers at the right moment: after the diagnosis is explained and before the invoice total lands, when the customer is already thinking about the cost of not maintaining the system. Then the office follows up on declines within a week, because a meaningful share of "not right now" is a timing objection, not a value objection.

Cadence matters more than tooling. A weekly five-minute huddle reviewing the board beats a monthly hour-long meeting, because the feedback loop is short enough for a tech to connect a specific behavior to a specific score. Monthly is where the bonus lands. Quarterly is where you revisit weights — unless a seasonal push or a new plan tier justifies moving them sooner.

Watch the adjacent effects, because they're real and they cut both ways. Push agreement attachment hard and you'll often see IAQ and accessory attachment rise alongside it — the same conversational muscle, the same moment in the call. You may also see average repair ticket dip slightly as techs spend minutes on the membership conversation instead of scanning for the next upsell. That's usually a good trade, but decide it deliberately rather than discovering it in a report. Callback rate is the one to watch for degradation: a tech rushing the technical work to make time for the pitch is a net loss, which is exactly why callback quality belongs on the matrix as a counterweight.

This same weighted-matrix pattern is not unique to HVAC. Plumbing and electrical shops run it with their own service-plan equivalents. Multi-unit retail and restaurant operators use the identical structure to balance a floor rep across attachment, ticket, and customer-satisfaction lines. The RevOps discipline underneath is the same in every case: define the full set of outcomes a role should produce, weight them to current strategy, measure the levels, and connect the composite to compensation. HVAC just has an unusually clean version of the problem, because the recurring-revenue line is so obviously separable from the transactional one.

Getting the crew to adopt it without a revolt

Roll it out as a measurement period first. Score for three or four weeks with no money attached and tell the crew that's what you're doing. This does two things: it surfaces the data-quality problems in your own reporting before they become pay disputes, and it lets techs see their numbers without feeling ambushed. Nearly every shop discovers something broken in week one — a field that isn't being filled in, an eligibility rule that's wrong, a tech whose numbers look terrible because his route is all warranty callbacks.

Involve the top-converting techs in setting the weights. They know which lines are realistic and which are fantasy, and their buy-in carries the room in a way an owner's mandate does not. It also converts your best sellers from quiet outliers into visible standard-setters, which is the mechanism by which the skill actually spreads.

Give the low scorers a concrete path, not just a low number. A published matrix without a route to improvement reads as a punishment board. Pair each weak line with one specific action: a two-sentence framing to practice, a ride-along with a strong tech, or a rehearsal of the three objections that actually come up. Most techs who are structurally low on offer rate move substantially within two to three months once they've said the sentence out loud a dozen times.

Do not let the board become a public humiliation device. Rank on composite, absolutely, but review individual gaps privately. The point is a nudge toward the whole visit, not a weekly ritual of naming the worst performer. Shops that get this wrong lose techs — and in a tight labor market, losing a competent tech to gain a few agreements is a terrible trade.

Last, protect the technical work. State plainly that a botched repair outranks any agreement, and back it up by keeping diagnosis quality and callback rate weighted heavily on the same matrix. Techs need to hear that you are asking them to sell the whole visit, not to become salespeople who happen to carry gauges. That framing — complete the visit, don't just complete the repair — is the one that tends to land.

Related questions

What's a realistic maintenance agreement attachment rate?

It varies widely by market, call mix, and plan price, so benchmark against your own trailing twelve months rather than a published figure. Set the first target as your current top tech's rate applied to the whole crew — that's proven achievable in your market with your plan.

Should CSRs get a spiff on agreements too?

Yes, if they're seeding or selling them on the phone. A front desk that raises the plan on inbound calls materially lowers the lift on the tech. Score the desk on mention rate the same way you score techs on offer rate.

How do I handle a tech who refuses to sell anything?

Score honestly, coach twice, then decide. Some excellent diagnosticians genuinely won't sell. You can route them to callbacks and complex diagnostics and staff agreement-heavy routes with others — but pay them on the role they actually fill, not the one you wish they filled.

Does this work for commercial HVAC service?

The structure transfers, but the KPI lines change. Commercial contracts are longer-cycle, often negotiated above the tech, so weight the tech's matrix toward site-condition reporting, scope-expansion flags, and deficiency documentation that feeds the contract renewal instead of a point-of-sale close.

FAQ

What if my techs just ignore the scorecard?

They ignore it when nothing depends on it. Tie the composite to the spiff and the monthly bonus, publish the board where the crew sees it, and run a short weekly huddle on where each tech stands. Visibility plus money is what makes it stick — a matrix that only the owner looks at changes nothing, no matter how well-designed the weights are.

How do I set the right weights for maintenance agreements?

Start by weighting agreements near the top of your set, just below diagnosis quality and repair execution — you never want the scorecard to imply that selling outranks fixing. Run those weights for a month, look at whether the composite ranking matches your own gut ranking of who's a complete tech, and adjust from there. If a launch or a slow season demands a push, raise the agreement weight temporarily and republish.

Will this work for a small crew of three to five techs?

It works better on small crews, because you can review every scorecard in a genuine one-on-one and adjust weights quickly. The trade-off is statistical noise: with few calls per tech per week, a single unusual route can swing a rate badly. Review on a rolling four-week window rather than week-to-week so one bad stretch doesn't distort the picture.

What if a tech is great at repairs but terrible at selling agreements?

That's precisely the gap the composite is built to surface. He'll score high on the technical lines and low on the commercial ones, and the number will land in the middle. Pair him with a strong seller for two or three ride-alongs and give him one specific sentence to practice rather than a full script. Most techs improve meaningfully within a quarter once the framing stops feeling like a pitch.

How often should I update the scorecard weights?

Quarterly is a sane default, with overnight changes reserved for a real strategic shift — a new membership tier, an IAQ push before allergy season, or a slow stretch you need to fill with scheduled work. Changing weights every week destroys the signal, because techs never get long enough on one set of priorities to move a number.

Do I need software to run this?

No. A spreadsheet with weights in a header row and a SUMPRODUCT formula does the math fine, and plenty of shops run for years that way. Software earns its cost when maintaining the sheet becomes the bottleneck, when you need the composite wired directly into payroll, or when you're past roughly fifteen techs and manual updating stops happening reliably.

Sources

• https://www.acca.org/ — Air Conditioning Contractors of America, service and maintenance standards
• https://www.ashrae.org/ — ASHRAE, HVAC system maintenance and performance guidance
• https://www.energy.gov/energysaver/maintaining-your-air-conditioner — U.S. Department of Energy on HVAC maintenance
• https://www.epa.gov/indoor-air-quality-iaq — U.S. EPA indoor air quality guidance
• https://www.bls.gov/ooh/installation-maintenance-and-repair/heating-air-conditioning-and-refrigeration-mechanics-and-installers.htm — Bureau of Labor Statistics, HVAC technician outlook
• https://hbr.org/2018/03/how-to-motivate-your-sales-team — Harvard Business Review on sales incentive design
• https://www.mckinsey.com/capabilities/growth-marketing-and-sales/our-insights — McKinsey growth, marketing and sales insights
• https://sloanreview.mit.edu/ — MIT Sloan Management Review on performance measurement
• https://www.sba.gov/business-guide/manage-your-business — U.S. Small Business Administration management guidance

Related on PULSE

• [How Do I Know Where, When, and How Many People to Schedule at Each of My Multi-Unit Retail Locations?](/knowledge/tl0001)
• [How Do I Figure Out How Many People to Schedule Each Day and at What Times for My Single Store?](/knowledge/tl0002)
• [How Do I Know How Many Cooks and Servers to Schedule Each Shift at My Pizza Restaurant?](/knowledge/tl0003)
• [How Many Salespeople Should I Schedule Each Day on My Furniture Store Floor?](/knowledge/tl0004)
• [How Do I Decide How Many Reps to Schedule at Each Store in My Mattress Retail Chain?](/knowledge/tl0005)