The 1% Edge: How a Small Improvement in Targeting Adds Six Figures to Your Bottom Line

The Problem With "Spray and Pray"

Most home services contractors approach marketing the same way: cast a wide net of postcards, emails, paid ads, door knocks across their service territory and hope the phone rings. The channel doesn't matter. The logic is always the same: reach as many homes as possible and let the law of averages do its work.

It works, sort of. But nobody stops to ask the harder question: what if your success rate was just marginally better?

Not a different channel. Not more spend. Just a slightly smarter model for identifying which homes in your territory are most likely to need your services right now, and pointing every marketing dollar at those homes instead.

The math on that tiny improvement is startling.

The Back-of-the-Envelope Math

Let's keep this simple. Take a mid-size HVAC contractor and lay out the basics.

You have a service territory with roughly 500,000 homes. Your marketing budget, across all channels, can realistically reach about 100,000 of those homes in a given year. Of the homes you reach, let's assume that your current success rate is about 10%, which implies that 10% of the people you engage convert into a paying customer through some combination of direct mail, email, digital ads, and organic.

At an average job value of $2,500, here's what your current spray and pray approach looks like vs. just a marginally better targeting approach:

Metric	Current approach (10% success)	Improved approach with +1 pp improvement (~11% success)
Homes in service territory	500,000	500,000
Homes your budget can reach	100,000	100,000
Success rate	10%	11%
Customers acquired	10,000	11,000
Avg. job value	$2,500	$2,500
Revenue	$25,000,000	$27,500,000
Incremental revenue	:	—

Same territory. Same budget. Same channels. Same creative. An extra $2.5 million in revenue, from a single percentage point improvement in who you choose to reach.

And 1 pps isn't some moonshot improvement. It's the difference between using zip codes and home values as your only filters versus layering in permit history, equipment age, property transaction data, and customer look-alike models. It's going from "spray the whole territory" to "stack the deck slightly in your favor."

Why It Compounds

A single campaign improvement is nice. But targeting precision doesn't just help one mailer or one email blast. It creates a flywheel that compounds across your entire marketing operation, regardless of channel.

Compounding layer 1: Every campaign benefits.

Whether you run direct mail, email reactivation, paid social, or a combination, the same targeting model powers all of them. An improvement in predicting who your next customer is lifts performance across every channel simultaneously. Run four campaigns a year and that $150,000 single-campaign gain becomes $600,000 annually.

Compounding layer 2: Customer lifetime value.

A first-time customer who calls for one job doesn't stay a one-job customer. The average homeowner spends $7,500+ with their contractor over a 3 to 5 year relationship when retained through maintenance agreements and follow-up service. Those 60 extra customers from better targeting aren't worth $150,000. They're worth $450,000 in lifetime revenue. At the macro level, the 1,000 incremental customers from the back-of-the-envelope model above carry $7.5 million in downstream LTV.

Compounding layer 3: The model gets smarter.

AI-powered targeting improves with every campaign you run. Each batch of results teaches the model which home attributes, equipment ages, permit histories, and neighborhood signals actually predict conversion. A 1 pp improvement in year one often becomes 2 to 3 pps by year two. The flywheel accelerates.

Where the Improvement Comes From

Not every customer is equally easy to find, or equally likely to convert. The best targeting models understand the hierarchy of opportunity.

Tier 1: Your member customers. These are the easiest conversions. They already know your brand, your pricing, and your team. They're loyal. They come back. The opportunity here isn't acquiring them, rather it's identifying which members are sitting on aging equipment, approaching a decision point, or candidates for cross-sell into a different service line. AI can surface those signals from your own ServiceTitan data before the customer even picks up the phone.

Tier 2: Repeat and past customers. Homeowners who've used you before but haven't come back recently. There's already familiarity, which dramatically increases conversion likelihood. The question is timing: reaching them when they actually need service, not just when your campaign calendar says to.

Tier 3: Net-new prospects. The hardest group to capture. These homeowners have never done business with you. This is where targeting precision matters most, because you're competing for attention against every other contractor in the market, and most of these homes don't need service right now. The difference between blunt targeting (zip code + home value) and intelligent targeting (permit history + equipment age + property signals + look-alike scoring) is what drives value.

Traditional targeting uses blunt filters: zip code, home value, maybe homeowner vs. renter. That gets you to a baseline success rate.

None of these signals individually are magical. But stacked together, they move the needle from "reach everyone" to "reach the right ones." That's where 1 percentage point of improvement in prediction accuracy turns into millions in incremental revenue.

This Is What Arch Was Built For

Arch integrates directly with ServiceTitan and layers in property records, permit history, equipment age signals, and AI-scored similarity models to find the homes most likely to convert, across every channel you use. Every campaign includes holdout-based attribution, so you know exactly what worked. That holdout data feeds back into the model, making every future campaign smarter.

Request a demo to see how it works on your data