Home Services Marketing: The Operator's Playbook

Most home services marketing advice is written as if your business were a shop. Get found, get clicked, get a lead, repeat. It treats demand as a smooth monthly number and capacity as unlimited.

Neither is true here. Demand arrives in spikes you cannot forecast and seasons you cannot move. Capacity is a fixed number of trucks with a fixed number of techs in them, and when marketing outruns that number the result is not growth, it is a three day backlog and a one star review. Your competitors are not two doors down either, they are anybody with a van and a Google Business Profile inside the same service area.

So this guide to home services marketing is written for the other reader: the owner or the multi-location operations and marketing lead at an HVAC, plumbing, electrical, roofing, pest, lawn, cleaning or garage door company. It covers the acquisition channels honestly, spends real time on answering the phone, and then goes where most guides never go, which is the list of past customers who had one job done and never called again. That list is the biggest single leak in this industry, and nobody else can market to it but you.

I run Winback Engine, a human phone reactivation service for multi-location service businesses, so you know where my bias sits.

What makes home services structurally different

Before channels, the mechanics. Four things shape every marketing decision here, and most generic advice ignores all four.

Demand splits in two, and the halves are won differently

Emergency work and planned work are different businesses that happen to share a truck.

Emergency work is a no heat call at 6am in February, a burst supply line, a dead panel. The customer has no loyalty to anyone, because the problem started twenty minutes ago. They search, call the first two or three results, and book with whoever answers and can be there soonest. You win it on being findable at the moment of panic and answering the phone. Brand, nurture, email and content decide nothing here.

Planned and replacement work is the opposite. A system replacement, a re-roof, a panel upgrade, a repipe. Tickets are larger, the customer takes weeks, collects two or three quotes, and is buying confidence as much as equipment. You win it on trust built before the quote: reviews, a recognizable local name, a past relationship, a tech who was in the house last year and did not make a mess.

Market only for emergency intent and you build a business of small tickets with no repeat path. Market only for planned work and you starve the techs between seasons. Most companies I talk to are tilted hard to the first and treat the second as luck.

Seasonality concentrates the year into weeks

The revenue year is not twelve even months. HVAC gets the first real heat wave and the first real freeze. Plumbing gets the holiday cluster. Roofing gets whatever the weather did last week. Lawn gets spring.

That makes marketing a question of timing. One expensive mistake is spending evenly across twelve months. The other is spending only inside the spike, when clicks cost the most and the schedule is already full, because during the spike you are capacity constrained rather than demand constrained and incremental paid spend often buys jobs you would have won anyway. The shoulder season is where the money is cheap and the schedule has room, which is also the right window for the past-customer work in the second half of this guide.

Your capacity is trucks and techs, so marketing can overshoot

A retail store with too much traffic has a good day. A service company with too much demand has a bad month, because the overflow becomes missed windows, rushed jobs, callbacks and reviews describing exactly that.

So the budget question is not “how do we get more leads”, it is “what does a fully booked week look like in revenue, and are we short of it or over it”. Over it, more spend buys worse delivery, and the right moves are capacity moves: dispatch efficiency, job sequencing, a hiring pipeline, higher prices, or pushing demand into the shoulder weeks. Near the ceiling, the money belongs in recruiting.

The service area replaces the storefront

A gym markets a building. You market a polygon.

Drive time is a cost that appears in no ad platform. A job forty minutes outside your core is less profitable than the identical job eight minutes away, and targeting that ignores it buys margin erosion. Multi-branch operators get the same problem one level up, with two branches quietly bidding against each other on the same overlapping suburbs.

Local visibility is also not one ranking but many. You can dominate the map results in your own town and be invisible six miles out, which is usually where the growth is, so local SEO here is a per-area exercise rather than one site-wide project.

If you run multiple branches or a franchise network, the structural layer on top of all this (brand control, local ad funds, channel ownership) is covered in our guide to franchise marketing. I will not rewrite it here.

The acquisition channels, honestly assessed

You need acquisition. Customers move, systems get replaced once a decade, and new service areas start from zero. Here is my honest read on the channels. I describe third party platforms only in general terms, because features and pricing change constantly and you should check them yourself.

Local SEO and Google Business Profile, per service area

The base layer. For emergency intent the map results take the click, and those are driven mainly by your Google Business Profile plus proximity to the searcher.

The unglamorous work that matters: one profile per real location, consistent name, address and phone, the correct primary category (the specific trade, not a generic “contractor”), services listed out, service area set honestly, real photos of real trucks and techs, growing review volume. Then a page on your own site per service area and per service, genuinely different town to town; templated pages with the city name swapped are not worth the hosting. Proximity is the hard constraint, so growth into adjacent areas leans on paid and reviews first.

Local Services Ads and paid search for emergency intent

For emergency keywords, paid placement sits above everything and the intent is about as pure as it gets. Google’s Local Services Ads product charges on leads rather than clicks and requires licensing and insurance verification, a real advantage in a category full of unlicensed competition. Standard paid search still has its place for service and brand terms and in areas where you are organically weak.

Three disciplines separate profitable paid search here from a bonfire: negative keywords maintained aggressively; dayparting matched to when your techs have capacity; and call tracking per campaign, because the conversion happens on the phone and a platform counting only clicks tells you almost nothing.

Reviews, because the homeowner is choosing between three vans

For planned work especially, reviews decide more often than price does. The customer has three quotes, cannot evaluate the equipment, and resolves the uncertainty by reading what the neighbors said.

This industry has an advantage most companies waste: your tech is in the house at the exact moment the problem is fixed and gratitude peaks. The companies that win reviews ask then, by text, from the tech, before the van pulls away. Track review velocity per branch and per tech, because a branch whose review count went flat did not get worse at service, somebody stopped asking.

Direct mail, neighborhood targeting and referral

Direct mail still earns its place because the targeting matches the business. You know which streets have housing stock of the age where systems fail, and a mailer to the forty houses around a completed job is cheap, relevant and credible. Measure it with distinct numbers or codes per drop, on cost per booked job.

Word of mouth is the best lead source here and it is not passive: ask at the end of good jobs and track who sends you work. Then there are the partners generic advice never mentions, realtors and property managers, home inspectors, general contractors, adjacent trades. A plumber who likes your electricians is a renewable lead source, and that is a relationship for someone senior to own, not a campaign.

Every channel above points at strangers, and every one gets more expensive each year because every competitor in your service area buys the same thing. There is one audience where you have no competition, and it is the one you already served.

Answering the phone is a marketing channel

This gets its own section because here a missed call is not a missed lead, it is a lost job that went to the next van. The customer with no heat is not leaving a voicemail.

Most operators under-measure it badly. They can tell me cost per lead to the cent and not what share of inbound calls went unanswered last Tuesday between 4pm and 6pm. The number is in your phone system and it is usually worse than anyone in the building believes, concentrated in the moments that matter most: the first warm day, the evening after a storm, lunch hour, Saturday.

What to do with it:

Pull the data monthly. Answer rate, abandoned calls and time to answer, broken out by hour, day and branch. Multiply missed calls by your booking rate and average ticket. That dollar figure is the honest cost of the gap, and it is usually the cheapest problem on the list to fix.
Separate booking from dispatch. When the person answering the phone is also coordinating techs, the phone loses. Every time.
Staff to the spikes, not the average. Overflow and after hours coverage are worth real money here specifically because of the emergency half of demand.
Measure the booking rate of the call, not just the pickup. A call answered and not booked is a different problem, usually script, pricing confidence or available windows. Track it per person. Then the uncomfortable arithmetic. If you are buying Local Services Ads and paid search to generate phone calls and some share of those calls goes unanswered, you are paying full price for leads and binning a portion of them. Fixing the phone beats raising the budget that feeds it.

The retention and reactivation layer, which is where the money actually is

Now the part almost no guide in this category treats seriously.

The one-job customer who never came back

Pull a list from your field service software of every customer with exactly one completed job more than eighteen months ago, no plan, no second visit. In every company I have looked at that list is large, often larger than the active customer base.

Think about what each record represents. Somebody in your service area had a problem, chose you out of three options, let your tech inside, paid the invoice, then had another problem two years later and called somebody else. Not because they were unhappy. Because they did not remember your name, and the first search result was right there.

That is the biggest leak in this industry and it is invisible on every dashboard. Nothing in your reporting shows the second job you did not get. Acquisition cost is a line item; a customer who quietly did not come back is not.

Why this list is the cheapest revenue available to you:

No media cost. You have the name, address, phone number and equipment history already. There is no auction.
Trust already happened once. The call starts from “we were out at your place in 2023”, not from a cold stranger.
You know the house, including how old that system was when you last touched it, which makes the reason for calling concrete.
No competitor can reach it. Anyone can outbid you on a keyword. Nobody else can call your customer list.

An illustrative model, assumptions stated, because I will not invent industry benchmarks. Say a branch has 2,000 single-job customers older than eighteen months, a worked campaign books 6 percent of them, which is 120 jobs, and the average ticket on that mix of diagnostic and repair work is $450. That is roughly $54,000 booked out of data you already own, before any of it turns into a plan or a replacement quote. Change the inputs to your own numbers in the ROI calculator.

Why doesn’t this happen already? Because reactivation is an operations problem wearing a marketing costume. Your CSRs are busy with today’s calls, and the automated “we miss you” email reaches somebody who has not thought about your company in three years and gets deleted. What works, consistently in the campaigns we run, is a human on the phone who can reference the last job, hear the real objection and put a time on the calendar in the same conversation. That is a capacity question, not a creative one. The framework is in our customer reactivation guide, and the reactivation call script breaks the conversation down line by line.

Maintenance plans and memberships, the structural fix

The one-job problem exists because nothing in a standard repair transaction creates a reason to come back. Maintenance plans fix that structurally: the customer pays monthly or annually, you perform scheduled visits, and now there is a recurring reason for your tech to be in that house. What a plan does for the P&L, in rough order of importance:

Shoulder-season work. Scheduled tune-ups are yours to place, so place them in the thin weeks. That alone changes your capacity economics.
The replacement sale becomes yours. The company inside the house twice a year quotes the replacement, and that is where the large tickets live.
Recurring revenue. A book of plan members is an asset that changes what the business is worth.
It stops the forgetting. Plan members cannot become one-job customers.

Honest notes, because plans get oversold as a cure-all.

Attach rate decides whether the program is worth running, and it is driven almost entirely by whether techs offer the plan on every job in a way that does not feel like a pitch. A plan nobody offers has an attach rate near zero and plenty of admin overhead. Track it by tech weekly and coach to it. The spread between your best and worst tech will be large, which tells you this is a training problem, not a product problem.

Plans are not free either. You owe scheduled visits, and those consume tech hours you could have sold, so priced below the cost of delivering them you have bought churn insurance at a loss. The payoff comes through replacement capture and repeat repair work, so if you do not track those downstream you do not know whether it works. And a plan sold to somebody who was never going to call anyone else is margin given away. The ones worth the discount are the drifters.

For the general mechanics of tiers, pricing and perks, see our guide on customer loyalty programs. What is specific to this industry is the three points above: shoulder-season placement, replacement capture and tech attach rate.

Expired plan members are the hottest list you own

Every plan program leaks. Cards expire, the auto-renew fails, somebody cancels during a tight month, somebody moves and never re-enrolls at the new address.

Those lapsed members are the highest-converting segment we see here, and that is our own campaign observation rather than a published figure. They understood the value and already paid for it, and a large share of cancellations were never decisions at all, they were failed payments nobody chased. Ask your software for everyone whose plan ended in the last twenty four months without renewing, sort by lapse date, call the top first. The ones who lapsed inside ninety days come back at a rate that makes the whole program look cheap. The payment-failure portion needs a different fix, so separate it out first; the mechanics are in reduce customer churn.

The post-job follow-up window

There is a short window after every completed job where the customer thinks well of you and is reachable. Most companies use it for the review request and then go quiet for years. What else belongs there:

The quality check. A call or text a few days out catches the small complaint before it becomes a one star review.
The deferred-work note. Your tech saw the water heater at the end of its life. That note belongs in the system with a date, and somebody should call on that date. It is almost always sitting unworked.
The plan offer, again. A customer who declined while stressed about the repair bill is a different conversation a week later.
The next-service reminder with a real reason. “Your system is due” works. “We miss you” does not.

If you want the general version of this across service verticals, see customer retention strategies for service businesses.

Measurement: the numbers that should run your marketing

Most reporting here stops at cost per lead, the least useful number in the stack. What belongs on the dashboard instead:

Cost per booked job, not cost per lead. Divide channel spend by jobs actually on the calendar from it. The gap between the two numbers is your intake performance, and it varies by channel: emergency paid search books immediately, planned-work leads take weeks and several follow-ups.

Average ticket, by channel and job type. Two channels at the same cost per booked job are not equal if one produces $180 service calls and the other $9,000 replacements. It also says whether discount-led advertising is buying customers or unprofitable work.

Revenue per truck per day. Your real capacity metric, and the one that says whether the next dollar belongs in marketing or in recruiting.

Membership attach rate. Plans sold divided by eligible jobs, by tech and branch, weekly. The leading indicator of next year’s repeat revenue.

Repeat rate. Of the customers you served twelve to twenty four months ago, what share have called again? This is the metric that exposes the one-job leak, and almost nobody computes it. Compute it per branch and watch it over quarters.

Reactivation cost per recovered job. Outreach cost divided by jobs booked from it, on the same slide as your paid cost per booked job. In our campaigns recovered jobs come in well below the cost of acquiring a stranger, and once both numbers sit in one table the budget argument gets short. For the framework behind valuing a customer rather than a job, customer lifetime value walks through it.

A 90-day plan

How I would sequence this with one quarter and a multi-branch company.

Days 1 to 30: measure what you are not measuring

Pull phone data by hour, day and branch. Compute answer rate and the dollar cost of missed calls from your own booking rate and average ticket.
Pull the single-job list per branch, older than eighteen months, and the expired plan member list separately, sorted by lapse date.
Compute repeat rate and attach rate per branch and per tech. The branch spread will be wider than you assumed.
Compute cost per booked job for every channel you can, and install call tracking where it does not exist.
Audit every Google Business Profile and service area page and fix the errors. Free, and it takes a week.

Days 31 to 60: fix the phone, then work the list

Fix intake first, because every other channel runs through it. Separate booking from dispatch, staff the spike hours, put a weekly answer rate in front of whoever owns it.
Make the review ask a required step at job close, and standardize the deferred-work note with a named owner for the callbacks.
Launch reactivation on your two thinnest-schedule branches, starting with expired plan members, because they convert fastest and will tell you whether the script works. Decide honestly whether your CSRs have the hours. In most companies they do not, which is where an outside calling partner earns its place.
Cut the paid spend that failed the cost-per-booked-job test, and add no new channels this month.

Days 61 to 90: compare, reallocate, make it a rhythm

Put reactivation cost per recovered job next to paid cost per booked job and rebuild next quarter’s budget from that comparison.
Roll reactivation to the next branches, now including the single-job list.
Coach attach rate off the weekly per-tech number and set a floor.
Fix the operating rhythm: answer rate and attach rate weekly, deferred-work list monthly, single-job list campaigned each quarter in the shoulder weeks, profiles and reviews audited quarterly.

By day 90 you have what few companies here have: every dollar carrying a cost per booked job, the phone treated as the channel it is, and the past-customer list managed rather than forgotten.

FAQ

What is home services marketing?

Home services marketing is everything an HVAC, plumbing, electrical, roofing, pest, lawn or cleaning company does to win and keep customers inside a service area: local SEO and Google Business Profile per area, Local Services Ads and paid search for emergency intent, reviews, direct mail and trade partnerships on the acquisition side; call intake performance, maintenance plans, post-job follow-up and reactivation of past customers and expired plan members on the retention side. Most published advice covers only the first half, which is where the money is spent, not where it is cheapest.

How much do home services companies spend on marketing?

There is no credible single percentage, and anyone handing you one is averaging across trades, markets and growth stages with nothing in common. Work backward from your own unit economics. Know your cost per booked job by channel and your average ticket on the work each channel produces, set a ceiling on what a booked job may cost, and fund only what stays under it. Then check capacity: if revenue per truck per day is near your ceiling, the spend belongs in hiring and dispatch, not more leads. Companies with large past-customer lists should fund reactivation before raising acquisition spend, because recovered jobs cost less than acquired strangers.

What marketing works best for emergency service calls?

Findability and pickup rate, in that order, and almost nothing else. The customer with no heat or an active leak decides in minutes, so the job goes to whoever is at the top of the search and answers the phone. That means Local Services Ads and paid search on emergency terms, a strong Google Business Profile in each area, and genuinely staffed intake including after hours. Content, email and brand campaigns decide planned and replacement work instead, a separate motion built on trust established earlier.

Are maintenance plans worth it for a home services business?

They are worth it when techs actually offer them on every eligible job and the plan is priced to cover the visits it commits you to. The payoff is not the plan fee, it is three other things: scheduled work you can place into thin weeks, being the company inside the house when the system needs replacing, and recurring revenue that raises what the business is worth. Track attach rate per tech weekly, because the gap between best and worst tech is where the program is won or lost.

How do you win back past customers in home services?

Size the two lists first: expired maintenance plan members, and customers with one completed job and no return visit in eighteen months or more. Work the expired members first, newest lapses first, because they convert fastest, and separate anyone who lapsed from a failed payment rather than a decision, since that is a billing fix and not a sales call. Then call. In our campaigns, a human who can name the last job, reference the equipment in that house, hear the objection and book a time in the same conversation outperforms email and text by a wide margin, because a call can handle an objection and an email cannot.

See what your past-customer list is worth

Most operators know the list exists. Few know what it is worth, and almost none have the calling hours to work it. Winback Engine runs human phone reactivation for multi-location service businesses: your automated flows run first, then our trained agents call the past customers and expired plan members those flows did not recover, and the booked jobs land on your schedule. We are priced on recovered revenue.

Start with a free audit of your past-customer and lapsed-member lists and we will size the recoverable revenue before you commit to anything. Run your own numbers first in the ROI calculator, or see how we work with multi-location and franchise networks on the franchise page. The same playbook in a different shape is in our guides to gym marketing, dental marketing and medical spa marketing.