Appliance EMT

The short answer

Appliance EMT is a residential appliance repair company serving the Atlanta metro area. The business was running Google Ads, generating leads, and booking repair jobs. On the surface, the paid search program was working. In reality, it was bleeding money.

The company was spending roughly $2,000 per month on a single broad-match Google Ads campaign that pointed all traffic to the homepage. Cost per acquisition was $70. Bounce rate on paid traffic was 74%. And roughly 40% of booked calls turned out to be outside the service radius, meaning the technician would drive to a job only to discover it was too far out to service profitably.

Within four months, I cut cost per acquisition from $70 to under $40, eliminated $1,200 per month in wasted ad spend, reduced the bounce rate from 74% to 41%, and built a system that let the owner scale from $2,000 to $8,000 per month in ad spend while maintaining a 3.2x return on ad spend. The owner went from one technician to three within six months of the campaign start.

Problem

One landing page cannot serve multiple intents

The fundamental problem was that Appliance EMT was treating all appliance repair searches as if they were the same. A single Google Ads campaign with broad match keywords was sending every click to the homepage. That means someone searching "dishwasher leaking water" landed on the same page as someone searching "refrigerator not cooling" and someone searching "washer making loud noise."

These are not the same customer. They have different appliances, different urgency levels, different concerns, and different buying mindsets.

The homepage had general information about the company, a list of all services, a phone number, and a contact form. It did not speak to any specific problem. It did not match the language the customer used in their search. It did not address the specific concern that drove them to search in the first place. The result was a 74% bounce rate on paid traffic, meaning nearly three out of four people who clicked an ad left without taking any action.

Wasted spend on irrelevant traffic

The campaign was running broad match keywords with minimal negative keyword management. Broad match tells Google to show your ad for searches that are "related to" your keyword, which in practice means Google shows your ads for a much wider range of queries than you intend.

I pulled the search term reports for the previous 90 days and found that a meaningful portion of the budget was being spent on queries that would never convert into a paying customer:

• Competitor brand terms. People searching for other appliance repair companies by name. These clicks are expensive and almost never convert because the person is specifically looking for a different business.
• DIY queries. Searches like "how to fix washer agitator" and "refrigerator compressor troubleshooting." These are people trying to fix the appliance themselves. They do not want to hire a repair technician.
• Out-of-area searches. Keyword and location modifier combinations that triggered ads for people outside the serviceable area. Every one of these clicks was pure waste because even if the person called and booked, the technician could not profitably service the job.
• Commercial and industrial queries. Searches for commercial appliance repair, restaurant equipment service, and industrial refrigeration. Appliance EMT is a residential repair company. These leads were unqualifiable from the start.

The total wasted spend from these irrelevant queries was approximately $1,200 per month, which was 60% of the $2,000 monthly budget. More than half of every dollar spent on advertising was going to clicks that had zero chance of becoming a paying customer.

The out-of-area problem

The 40% out-of-area call rate was not just a marketing problem. It was an operational cost multiplier. When a technician drives 45 minutes to a service call only to realize the address is outside the profitable service radius, that is not just the cost of the wasted ad click. It is the technician's time, the fuel, the wear on the vehicle, and the opportunity cost of the jobs that technician could have been completing during that time.

For a one-technician operation, every wasted service call has a direct impact on daily revenue capacity. If the technician can complete five jobs per day and one of them is an out-of-area call that takes two hours round trip for no revenue, that is a 20% reduction in daily earning potential.

Approach

The first priority was to stop the bleeding. Before building anything new, I needed to eliminate the wasted spend so that every dollar going forward was reaching people who could actually become customers.

Search Term Report Analysis

I exported 90 days of search term data from the Google Ads account and categorized every query that had generated a click. Each query was classified as either a legitimate repair intent query, a competitor brand query, a DIY query, an out-of-area query, or a commercial/industrial query.

This classification revealed that only about 40% of clicks were coming from searches with genuine residential appliance repair intent within the service area. The other 60% were split across the irrelevant categories listed above.

Negative Keyword Architecture

I built a comprehensive negative keyword list organized into categories:

• Brand negatives: Every competitor name in the Atlanta metro appliance repair market
• DIY negatives: "how to," "tutorial," "troubleshoot myself," "parts for," "DIY," and similar modifiers
• Commercial negatives: "commercial," "restaurant," "industrial," "HVAC" (a common confusion with appliance repair), and related terms
• Geographic negatives: City names and zip codes outside the service radius

This negative keyword list was not a one-time addition. It became a living document that I updated weekly based on new search term data. Every week, new irrelevant queries would appear, and the negative list would grow to block them.

Campaign Restructure

The single broad-match campaign was broken into appliance-specific ad groups, each using a combination of exact match and phrase match keywords. Broad match was eliminated entirely. This gave precise control over which searches triggered which ads and allowed the messaging in each ad to match the specific appliance the person was searching for.

With the budget cleaned up and the campaign restructured, the next step was to give each ad group a landing page that matched the search intent exactly.

Page Design Principles

Every landing page followed the same conversion-focused structure, but the content was customized for the specific appliance:

Headline matches the search query. If someone searches "dishwasher repair Atlanta," the landing page headline says "Dishwasher Repair in Atlanta." Not "Appliance Repair Services." Not "We Fix Everything." The exact appliance, the exact service, the exact location. This creates instant relevance. The visitor sees within one second that they are in the right place.

Single clear action above the fold. A click-to-call button and a short form. No navigation menu, no links to other pages, no distractions. The entire purpose of the page is to get the visitor to call or submit the form. Every element that does not contribute to that action was removed.

Trust signals specific to the appliance. Each page listed the specific brands serviced for that appliance type, common problems diagnosed, typical repair timeframes, and a service guarantee. A dishwasher repair page listed dishwasher brands (Bosch, Samsung, LG, Whirlpool, GE, KitchenAid). A refrigerator repair page listed refrigerator brands. This specificity signals expertise in the exact problem the customer is facing.

Urgency and availability. Same-day service availability was displayed prominently. For appliances with high urgency (refrigerators, freezers), the messaging emphasized speed. For lower-urgency repairs (dryers, dishwashers), the messaging emphasized quality and warranty.

I built six landing pages total, one for each major appliance category: refrigerator, dishwasher, washer, dryer, oven/range, and freezer.

The landing pages and restructured campaigns were the foundation. The call tracking system was what made the entire operation continuously improve over time.

Unique Tracking Numbers

Every phone number on every landing page received a unique CallRail tracking number. This meant I could see exactly which campaign, ad group, keyword, and landing page generated each phone call. If a call came in from the dishwasher repair page, I knew it was a dishwasher lead. If it came from a specific keyword, I knew exactly which search triggered it.

Weekly Call Review

Every week, I reviewed call recordings from the previous seven days. This was not a passive listen. I was specifically looking for:

• Bad calls that should not have happened. Wrong numbers, spam, people looking for a different type of service, people outside the service area. Each of these calls was traced back to the keyword that triggered it, and that keyword was either added as a negative or the match type was tightened.
• Good calls that revealed new keyword opportunities. When a customer described their problem on the phone using language that was not in the keyword list, that language became a new keyword. Real customers describe their problems differently than marketers write keywords.
• Conversion quality signals. Which calls actually booked a job versus which calls were just price shoppers. This data informed bid adjustments: keywords that generated booked jobs got higher bids, keywords that generated price shoppers got lower bids or more targeted ad copy.

With clean traffic flowing to conversion-optimized landing pages and a feedback loop catching problems weekly, the final optimization layer was bidding strategy.

Device Bid Adjustments

Appliance repair is an overwhelmingly mobile search category. The person whose dishwasher is flooding their kitchen is not sitting at a desktop computer researching repair options. They are standing in their kitchen with their phone, searching for someone who can come fix it now. Mobile traffic converted at nearly double the rate of desktop traffic, so I increased mobile bids to capture more of that high-intent traffic.

Time-of-Day Bidding

Repair calls follow a predictable daily pattern. Appliances break down at all hours, but people call repair services during specific windows. The highest-converting window was weekday mornings between 7 AM and 10 AM. The pattern makes sense: the appliance breaks overnight or the homeowner notices the problem in the morning, and they call before leaving for work.

I increased bids by 30% during the 7-10 AM weekday window on mobile devices and cut bids by 50% after 8 PM when calls rarely converted. This put more budget behind the hours and devices that actually generated booked jobs and pulled budget away from the times that did not.

Result

Scaling with confidence

The reason the owner was able to scale from $2,000 to $8,000 per month in ad spend was that the unit economics were proven and stable. At a $40 CPA and a 3.2x ROAS, every additional dollar spent on advertising generated $3.20 in revenue. That math made it easy to justify increasing the budget because the return was predictable and consistent.

ROAS held above 3x the entire time we were scaling spend 4x. That consistency matters because it is easy to achieve a good ROAS at low spend. The real test is whether the efficiency holds as you scale. In this case, it did, because the campaign structure (appliance-specific pages, tight keyword targeting, continuous negative keyword management, and dayparting) created a system that scaled without losing precision.

The compound effect of the feedback loop

The weekly call review process created a compound improvement effect. Each week, new negative keywords were added. Each week, new keyword opportunities were discovered from call recordings. Each week, bid adjustments were refined based on which keywords actually booked jobs. After four months, the campaign was fundamentally different from where it started. Not because of one big change, but because of dozens of small, data-driven improvements that compounded over time.

In the client's words

"Before working with Zach, we were paying $70 a lead and half were outside our service area. He rebuilt the campaigns, the tracking, and the routing in two weeks. Two months later we were at $40 per lead, all qualified, 40 to 50 a day. I went from one technician to three. He didn't just fix the ads, he changed the trajectory of the business."

— Valentin Pashnyak, Owner, Appliance EMT

What I'd do differently

Start with call recording analysis on day one instead of week two. I spent the first week rebuilding the account structure (campaign segmentation, negative keywords, match types) before setting up the call tracking and review process. The bad-call data was already sitting in CallRail from the previous agency's work. If I had started reviewing those recordings immediately, in parallel with the account restructure, I could have identified the worst-performing keywords and eliminated them days earlier. At $1,200 per month in wasted spend, even a few days of faster optimization would have saved real money. The lesson: run the audit and the rebuild as parallel tracks, not sequential steps.