Through May 2026, U.S. shipments of central air conditioners and air-source heat pumps ran 3.69 million units, down 1.2% from the same stretch of 2025, according to AHRI. Gas warm-air furnaces were down 10.5% year to date. Residential storage water heaters were down 5.4% on gas and 6.6% on electric. Only heat pumps and oil furnaces grew. Manufacturers have been guiding residential volumes down 10% to 15% for the year, and January alone came in 29.2% below January 2025.
That is the install side of your business shrinking while your overhead stays put. The service side is the part that did not move. And most shops are quietly losing a third of it every twelve months without noticing, because nobody in the building tracks the one number that measures it.
The Install Market Stopped Carrying You
For four years, replacement volume covered a lot of operational sins. When the phone rang constantly, weak follow-up looked fine. When homeowners had cash, loose pricing went unnoticed. Cornerstone Advertising made this point bluntly in an August 21 piece: the home services gold rush is over, and 2026 is separating the shops that built businesses from the ones that rode a market.
The capital markets already priced this in. Private equity platforms pay 6x to 10x EBITDA for HVAC and plumbing companies with 50% or more of revenue coming from service contracts. They are not paying that premium for your truck fleet. They are paying it for revenue that shows up whether or not compressors fail this August. A shop at 30% agreement penetration can absorb a 20% drop in equipment revenue. A shop at 10% cannot.
Meanwhile the competitive picture got worse. Regional platforms with call centers and centralized marketing budgets are bidding on the same Local Services Ads and the same phone call you are. Customer acquisition cost went up while demand went sideways. Under those conditions, the cheapest revenue available to you is the customer who already signed an agreement and is about to let it lapse.
Renewal Rate Is the Number Nobody in Your Office Owns
Ask three people in your office what your maintenance agreement renewal rate was last quarter. You will get three answers, and none of them will be right, because most shops report the agreement count and never report the churn underneath it. A book that grows from 820 to 860 looks healthy. If you sold 340 new agreements to get there, you lost 300. That is a 63% renewal rate on a growing book.
Here is the benchmark ladder. Shops relying on manual office outreach land between 55% and 65%. Shops running automatic renewal with digital payment on file plus a scheduled reminder sequence land between 80% and 85%. Top-shop targets sit at 90% or better. Industry survey averages cluster in the 60% to 75% range depending on who is counting. The typical contractor loses 35% to 45% of agreement customers a year, and by most estimates 25% to 40% of those cancellations happen because the office never followed up. Not price. Not service quality. Silence.
Run the math on your own book before you read further. Take agreements active on January 1. Count how many of those same customers still hold an agreement today. Divide. Ignore new sales entirely, because new sales are what has been hiding the leak. If you have 900 agreements at an average of $220 a year, moving from 62% to 82% keeps 180 customers. That is $39,600 of recurring revenue, plus the repair and replacement work those 180 households will hand you over the next five years instead of handing it to whoever answers first.
Build the 60-Day Renewal Sequence This Month
The window that matters opens 45 to 60 days before the agreement expires. That is when the customer still associates your name with a completed visit and cancellation risk is lowest. Waiting until the expiration date turns a renewal into a sales call. Here is the build.
- Get the expiration dates into one field. ServiceTitan, FieldEdge, Simpro, Jobber and Housecall Pro all carry membership or agreement modules with an expiration date and a renewal status. If your dates live in a spreadsheet or in the office manager’s head, that is step one and it takes an afternoon.
- Put a card on file and turn on auto-renew. This single change moves renewal rate more than any message you will write. Pair it with an account updater through your payment processor so expired and reissued cards do not silently kill agreements. Disclose the auto-renew terms in plain language at signup, send a renewal notice before the charge, and make cancellation a one-step request. Shops that hide the terms trade a renewal today for a chargeback and a one-star review later.
- Send day 60, day 30 and day 7 touches, attributed to the technician. Not “Team ABC Heating.” Use the name and photo of the tech who ran the last visit. A message signed by Marcus who fixed the capacitor in June converts at a materially different rate than a template from the office. Most CRMs support merge fields for last technician.
- Route the reply to a human within two hours. Automation opens the conversation. It does not close a customer who writes back asking whether the plan covers their new water heater.
On tooling, ServiceTitan prices by quote and its Marketing Pro add-on now includes Atlas, the agentic layer the company rolled out for campaign recommendations, including the follow-ups teams skip when nobody owns marketing. Renewal outreach is exactly that category of work. Housecall Pro and Jobber both publish tiered plans in the low hundreds per month, with the automation and campaign features living on the upper tiers. Verify current pricing before you budget, because all three have moved list prices in the past year. If you are on a platform without a renewal workflow, a $50 a month scheduling tool plus a shared inbox beats another quarter of nobody calling.
For the message itself, give your assistant of choice the constraints instead of asking it to write an email. Something like: “Write a 70-word text from a technician named Marcus to a homeowner whose HVAC maintenance agreement expires in 30 days. Reference that he replaced a run capacitor on June 14. Ask one question: does he still have the right card on file. No exclamation points, no marketing language, eighth-grade reading level.” Then read it out loud before it ships. If it does not sound like Marcus, rewrite it.
What Actually Breaks a Renewal Program
Three failure modes account for most of it. First, discounting to save the lapse. The moment you offer 20% off to keep someone, you have taught your whole book to threaten cancellation every September. Hold price and improve the reason instead. Second, over-automating the exit. If your only cancellation path is a phone call during business hours, some share of those customers become complaints instead of renewals. Third, treating the agreement as a filter change. The agreements that renew at 90% come with something the customer notices, like priority scheduling with a real response window, a documented equipment history, or a flat diagnostic fee. Vague benefits renew like vague benefits.
There is also an honest tradeoff to name. Auto-renew raises renewal rate and it raises customer service load in the first cycle. Budget for a handful of “I did not know I was still enrolled” calls in month one. That is a cheaper problem than a book that shrinks 40% a year, but pretending it does not exist is how good programs get killed by the office in week three.
Pick your decision for September: pull the renewal rate on your active book before the end of next week, then choose one of the four steps above and ship it by October 1. Card on file is the highest-yield single move if you can only do one. The signal to watch over the next 30 days is your own agreement book, measured the same way twice, because a number you compute once is a fact and a number you compute monthly is a lever.
Atlas Unchained builds the automation layer that runs this sequence for trades shops that do not want another dashboard to babysit. If you want the renewal audit and the sequence spec, get in touch, and subscribe to get these breakdowns each weekday morning.
