Internet sales up 3.8×, and a business that nearly doubled.
An inherited search account with a spend ceiling and no second channel. We rebuilt search into hubs, opened Meta with scripted problem-first video, and internet sales went from $1.65M to $6.2M in eight months.
What eight months produced.
Internet went from about a third of the company's sales to roughly 70% of them. Sales are net sales by lead source from EverDry's own reporting, not platform-modelled conversions.
A shrinking business, a capped account, and one channel.
EverDry Michiana sells a high-ticket, in-home service: every sale requires an appointment in the homeowner's basement. Lead volume on its own is a vanity metric. What matters is how many qualified homeowners agree to an inspection, at what cost, whether the account can absorb more budget without the cost per lead running away from it, and how much of that turns into closed sales.
When the company engaged Localizr, the business was shrinking and the account was inherited mid-flight: a fragmented set of legacy search campaigns competing against each other, and no paid social at all.
In the first eight months of 2025, total company sales were down 26% on the year before, and internet sales were down 32%. The channel with the most room to grow was the one shrinking.
Legacy campaigns were split by service in overlapping ways, so they were bidding against each other for the same searches. Every budget increase raised the auction price rather than the lead count. Past a certain monthly number, more money bought the same leads at a higher price.
Before the engagement, more than half of all search conversions came from people already searching for the company by name. Those are the cheapest leads in any account and they are not new demand. Blended into one number, they made a generic-search program that was barely working look acceptable.
Effectively all paid lead flow depended on Google. For a seasonal, weather-driven business, that is a single point of failure: a soft month in search is a soft month on the sales board.
Structure first, then a second channel.
Monthly media investment across all channels moved from roughly $19K to a $55–75K range, the ceiling the previous structure could not break. Four moves made that possible, in the order they were made.
Consolidate search into hubs, then scale into the ceiling
The legacy service-by-service campaigns were retired and rebuilt into a hub structure: one main-services hub for the region, plus a separate brand campaign so branded demand never shares a budget with generic demand. Brand is now reported on its own line instead of being used to flatter generic performance.
Open a second channel
A dedicated Meta Ads program launched inside the first quarter and scaled to a meaningful share of the media plan, adding hundreds of web leads plus confirmed inbound calls in its first six months. Cost per lead started above $1,000 in the launch month and settled into the low-to-mid hundreds once creative and audience testing found their footing — the normal shape of a new channel, and why launch-month numbers should never be used to judge one.
Make Facebook a scripting discipline, not a design exercise
Almost everything running on Meta is scripted short-form video, and the script is written before anything is filmed. The ads work because of what they say, not how they look.
Test upper funnel deliberately, and cap it
A YouTube / Demand Gen test ran alongside the core channels to see whether video demand could feed the search funnel. It generated tens of thousands of engaged clicks but no leads at an acceptable cost, so it was held to under 3% of account spend and reported the same week it became clear.
Same eight months, one year apart.
January through August 2025 against January through August 2026.
- Campaign structure
- Overlapping service-by-service campaigns bidding against each other
- Monthly media investment
- Roughly $19K, all of it on Google
- Adding budget
- Raised the price per click instead of the number of leads
- Branded share of conversions
- More than half, blended in with generic demand
- Demand channels
- One: Google search
- The business
- Total sales down 26%, internet sales down 32%
Rebuilt on hubs- Campaign structure
- One main-services hub for the region, brand in a campaign of its own
- Monthly media investment
- A $55–75K range across search and social
- Adding budget
- Search investment roughly 3×, clicks 3.0× and impressions 2.9×, at a flat cost per click
- Branded share of conversions
- About one in five, with non-branded conversions up roughly 3.7×
- Demand channels
- Three: Google, Meta and a capped YouTube test
- The business
- Total sales up 93%, internet sales up 3.8×
Branded search did not shrink. Genuinely new demand grew around it, which is the number that matters: the company now reaches homeowners who did not already know its name.
Creative is a script problem, not a design problem.
The hook does the work
Every ad opens on a specific, uncomfortable homeowner moment rather than on the company or the offer: "it's just a small crack," "the worst time to find out," "even if it looks dry," "one storm away." The first three seconds have to name it, or nothing after it gets watched.
One fear per script
Concepts are built one problem at a time — water during a storm, musty air and mold, hairline cracks widening, foundation settling, the repair bill that arrives later. Ads that cover the whole service menu test worse than ads that name one thing the homeowner has already noticed.
Local proof over production value
Named customer testimonials and creator-style UGC filmed in real Michiana basements consistently outperform polished studio cuts. The market recognizes its own weather, its own housing stock and its own accents.
Write the season into the script
Spring thaw and snowmelt, storm season, freeze-thaw cracking: the calendar decides which angle runs when, because a homeowner's willingness to book an inspection follows the weather.
Version it, retire it on evidence
Each concept runs in several scripted variants. The winning hook is kept and re-cut rather than rebuilt, creative is refreshed on measured fatigue instead of on a calendar, and concepts that stop producing are cut, not defended.
Measured on the sales board, not the ad platform.
Lead counts and cost per lead guided the work week to week. The scoreboard was the client's own sales report, January through August of each year.
Internet sales 3.8×
From $1.65M to $6.20M. The $6.2M through August is already 2.5× all of 2025, which closed at $2.51M, with four months of 2026 still to go.
Total company sales +93%
From $4.61M to $8.91M over the same eight months, already 34% above the whole of 2025.
From a third of the business to 70%
Internet is now the company's largest source of sales by a wide margin. Homegrown sales fell by about $640K; internet rose by about $4.56M. The business got bigger, not just reshuffled.
The highest-spending stretch was also the cheapest.
The account's highest-spending stretch of the engagement produced its lowest cost per lead and its highest lead volume on record. Efficiency normally degrades as budget grows, because the cheapest demand gets bought first. Cutting cost per lead while nearly tripling investment means the account was never efficient to begin with — it was small, and capped.
Cost per lead fell as the budget grew.
1,900+ tracked leads
Across search and social in the current year, against a prior-year baseline of a single channel producing a small fraction of that.
−49% cost per lead
From the low $300s early in the engagement to the mid $100s in the most recent full month, on platform-tracked search leads, while monthly investment nearly tripled.
Three channels instead of one
A soft month in search no longer means a soft month in the business. The highest-volume month on record was also the cheapest.
Don't take our word for it
Camden Kaminsky runs operations at EverDry Michiana. Here is what changed for his team, in his own words.
Camden KaminskyChief Operating Officer, EverDry Michiana
Check If You QualifySix things that made the difference.
Structure before budget
Consolidating overlapping campaigns is what made budget increases produce leads instead of higher click prices. No amount of bid management fixes an account that is bidding against itself.
Brand and generic, separated
Reporting them together hides whether an account is generating new demand. Splitting them made the real problem visible in week one.
The second channel, opened early
Meta went live inside the first quarter, not after search was declared finished. In a seasonal business, diversification is worth more than a marginal efficiency gain.
Creative as a script problem
The winning ads name the homeowner's specific fear in the first three seconds. Design does not decide that.
Sales, not just leads
Lead counts and cost per lead guided the work week to week, but the scoreboard was the client's own sales report.
Bad news, reported as fast as good
The upper-funnel test was measured, capped and reported honestly instead of quietly buried in the account.
Advertising figures come from the Google Ads and Meta Ads APIs for the periods stated, as of September 2026. Google "tracked leads" are platform-recorded calls and form submissions; tracking coverage was expanded during the engagement, so year-over-year comparisons are stated as ranges and directions rather than like-for-like counts. Meta lead counts use a single web-lead conversion event. Sales are net sales by lead source from the client's own reporting, January–August of each year, as of September 2026; "internet" is the client's own lead-source category, and 2026 includes January and February, before Meta launched. Nothing is annualized or projected.
Questions about this engagement.
How long did results take?
Structural work in the first two months, the first efficiency gains by month three, and the best cost-per-lead month later in the engagement. Home services accounts with in-home sales cycles need a full quarter before efficiency claims mean anything.
Does more spend always mean a worse cost per lead?
Not if the structure and measurement can carry it. Here investment roughly tripled while cost per lead fell — but only after the account was rebuilt. In the original structure, more budget genuinely did mean more expensive leads.
How do you know the growth is real and not just branded traffic?
Because brand is reported separately. Branded search went from more than half of all conversions to about one in five while total volume tripled, which means the growth came from homeowners who did not already know the company's name.
Is the sales growth just revenue moving from one channel to another?
No. Some of it is channel mix: homegrown sales fell by about $640K over the same period. But internet sales rose by about $4.56M, and total company sales grew by $4.31M. The business got bigger, not just reshuffled.
Want to know where your account's ceiling is?
Home services companies across Ohio, Michigan, Indiana, Wisconsin, Pennsylvania, and South Florida.
What does your company do?
Three quick questions, then pick a time for your call.
What's your company's annual revenue?
Our system is built for a certain size. This keeps the call useful for you.
How much do you spend on ads per month?
Across every platform: Meta, Google, TikTok, LSA, and other platforms.
You're a fit. Pick a time.
A 30-minute call, straight to your market.
Thanks. We're probably not the right fit yet.
Our system is built for home services companies at a certain size and ad budget. Below that, we'd be too much system for where you are right now, and we'd rather tell you now than take up your time on a call.
When the business grows into it, we'd like to hear from you again.