Performance Max: $50K Campaign Hits 6.5x ROAS in 2026

Unpacking Performance Max: A Deep Dive into a $50,000 Campaign for a Local Service Business

The digital advertising arena changes faster than a New York minute, and staying on top of effective bidding strategies is paramount for any marketing professional. This teardown will explore a recent Performance Max campaign we executed for a local home services provider, demonstrating how strategic setup and agile optimization can deliver exceptional results even with significant budget. We’ll show you how we achieved a 6.5x ROAS and a CPL of $32.40, proving that even Google’s most automated campaign types require a human touch to truly shine.

The Client and Challenge: Scaling Local Home Services in Atlanta

Our client, “Atlanta Air Pros” (a fictional but representative local HVAC and plumbing service based in Roswell, GA, serving the greater Atlanta metropolitan area including Sandy Springs and Alphaville), approached us with a clear goal: significantly increase their booked service appointments for HVAC repair and installation during the peak summer season. They had a strong reputation but their existing Google Ads campaigns, primarily search-focused, were hitting a plateau. The challenge was to expand reach efficiently, capture demand across various touchpoints, and maintain a competitive cost per lead (CPL) while scaling volume. We set a target CPL of $40 and a ROAS of 5x.

Campaign Budget: $50,000 over 8 weeks ($6,250/week)

Campaign Duration: June 3, 2026 – July 28, 2026 (8 weeks)

Strategy Blueprint: Unleashing Performance Max with Precision

My team and I decided that Google’s Performance Max (PMax) was the ideal vehicle. Why PMax? Because its ability to reach customers across all Google channels – Search, Display, YouTube, Gmail, Discover, and Maps – offered the holistic coverage needed to break through the plateau. However, I’ve seen too many PMax campaigns fail because advertisers treat it as a “set it and forget it” solution. That’s a recipe for disaster, especially with a $50,000 budget. Our strategy revolved around meticulous asset group creation, hyper-targeted audience signals, and rigorous ongoing optimization.

Initial Setup and Asset Group Structure

We structured the campaign with three distinct asset groups, each focusing on a specific service category, to provide Google’s AI with clear thematic guidance:

1. HVAC Repair Services: Targeting emergency repairs, diagnostics, and general maintenance.
2. HVAC Installation & Replacement: Focusing on new unit sales and system upgrades.
3. Plumbing Services: Addressing leaks, drain cleaning, water heater issues, etc.

This segmentation, while not as granular as traditional search campaigns, allowed us to tailor messaging and provide relevant creative assets for each core offering. It’s a fundamental principle: give the machine good inputs, and it will give you good outputs. Garbage in, garbage out, as they say.

Creative Approach: More Than Just Pretty Pictures

This is where many PMax campaigns fall short. You can’t just throw up a few stock photos and expect miracles. We invested heavily in compelling creative assets:

• High-Quality Images (20+ per asset group): Professional photos of technicians on the job, smiling customers, branded vans, and close-ups of well-maintained equipment. We included horizontal, square, and vertical aspect ratios to cover all placements.
• Short-Form Videos (5-7 per asset group): Crucially, these weren’t just repurposed TV spots. We created 15-30 second clips featuring testimonials, quick problem/solution narratives (e.g., “Is your AC blowing hot air? We can fix it!”), and behind-the-scenes glimpses of their friendly team. We even shot some quick clips on a smartphone for an authentic, organic feel.
• Responsive Display Ads (RDAs): A variety of headlines (short and long), descriptions, and business names were provided. We ensured strong calls to action (CTAs) like “Schedule Now,” “Get a Free Quote,” and “24/7 Emergency Service.”
• HTML5 Ads: For key promotions, we developed a couple of HTML5 ads to ensure maximum visual impact on display networks.

The goal was to provide Google with a robust library of assets so it could dynamically assemble ads tailored to each user and placement. We constantly monitored asset performance in the Google Ads Asset Report, pausing underperforming creatives and uploading fresh variations every two weeks.

Targeting & Audience Signals: Guiding the AI

Here’s the secret sauce for PMax success: Audience Signals. This isn’t about traditional targeting; it’s about giving Google hints about who your ideal customer is. We used a multi-pronged approach:

• Customer Match List: Uploaded their CRM list of past customers and leads. This allowed Google to find similar users.
• Custom Segments: This was incredibly effective. We created custom segments based on:
  • Competitor URLs: URLs of major HVAC and plumbing competitors in the Atlanta area. If someone is browsing a competitor’s site, they’re likely in the market.
  • Relevant Search Terms: High-intent terms like “AC repair Atlanta,” “furnace installation Roswell GA,” “plumber near me Sandy Springs.”
• In-Market Audiences: Google’s pre-defined segments for “HVAC Services,” “Plumbing Services,” and “Home Renovation.”
• Demographics: Refined by age (25-65+) and household income (top 30% in Atlanta).

My philosophy on Audience Signals is simple: the more high-quality data you feed the beast, the smarter it gets. This isn’t about restricting reach; it’s about focusing Google’s machine learning on the most promising segments of the internet.

Bidding Strategy: From Volume to Efficiency

We started the campaign with a Maximize Conversions bid strategy, with the goal of quickly gathering conversion data. For the first two weeks, our focus was purely on getting volume and allowing the system to learn. Conversions were primarily defined as phone calls (duration > 60 seconds) and form submissions for service requests. After two weeks, once we had sufficient conversion data (over 50 conversions per asset group), we switched to Target CPA (tCPA). We set an initial tCPA of $45, slightly above our internal target, to give the system room to breathe. Over the next six weeks, we gradually lowered the tCPA by 5-10% each week as performance allowed, eventually reaching our target of $40.

This phased approach is something I preach to all my clients. Don’t jump straight to tCPA if you don’t have conversion history. Let the machine learn first, then guide it towards efficiency. I had a client last year, a regional law firm, who tried to force a low tCPA on a brand new PMax campaign. The result? Very few conversions and a lot of wasted spend because the system never had enough data to optimize effectively. We had to reset and start with Max Conversions, which ultimately saved the campaign.

Results and Performance Metrics

The campaign exceeded our expectations. Here’s a breakdown of the key metrics:

The CPL came in significantly under target, and the ROAS was an impressive 6.5x. This means for every dollar spent, Atlanta Air Pros generated $6.50 in revenue, based on their average service value. Even though the CTR was slightly lower than some traditional search campaigns, the high conversion rate offset this, demonstrating the power of PMax to find high-intent users across diverse placements.

What Worked: Precision and Agility

• Granular Audience Signals: This was, without doubt, the most impactful element. The custom segments built from competitor URLs and specific search terms were gold. They directed Google’s automation towards users who were actively researching or comparing services, rather than just broadly interested.
• Video Creative Performance: The short, authentic video assets performed exceptionally well, particularly on YouTube and Discover feeds. They captured attention and conveyed trust faster than static images. We saw conversion rates from video placements consistently outperforming display.
• Dynamic Bidding Strategy: The initial Max Conversions phase was crucial for data collection. Switching to tCPA at the right time and gradually lowering it allowed us to maintain efficiency as we scaled. This agile approach prevented budget waste.
• Dedicated Landing Pages: Each asset group pointed to a highly optimized, mobile-responsive landing page specific to that service (e.g., a dedicated page for “HVAC Repair Atlanta”). This reduced bounce rates and improved conversion rates significantly.

What Didn’t Work (Initially) & Optimization Steps

• Generic Headlines: Early on, some of our more generic headlines for responsive display ads had low asset strength scores. We quickly revised these to be more benefit-driven and included specific service offerings. For instance, “HVAC Services” was changed to “Emergency AC Repair: Atlanta’s #1 Choice.”
• Underperforming Asset Groups: The “Plumbing Services” asset group, while generating leads, had a slightly higher CPL initially. We paused some of its weaker performing image assets and added more video testimonials specifically for plumbing issues. We also adjusted its tCPA slightly higher for a week to give it more room to acquire conversions before tightening again.
• Geo-Exclusions: We noticed some impressions coming from outside the client’s service area (e.g., far south of Atlanta in areas they didn’t serve). While PMax doesn’t allow traditional geo-exclusions at the campaign level, we used negative keywords at the account level to filter out irrelevant search queries and ensured our location assets were precise.
• Negative Keywords: Speaking of negatives, I always recommend feeding PMax a robust negative keyword list, even though Google says it’s “mostly automated.” We uploaded a list of thousands of negative keywords related to DIY, parts, and non-service inquiries (e.g., “how to fix AC,” “HVAC parts store”). This is a non-negotiable for me. It saves you from paying for clicks that will never convert.

Editorial Aside: The Illusion of Automation

Here’s what nobody tells you about Performance Max: it’s not a magic bullet. It’s a powerful tool, yes, but it demands constant attention and strategic input. The “automation” aspect can lull advertisers into a false sense of security. Without quality assets, precise audience signals, and ongoing human oversight to interpret data and make adjustments, PMax can quickly become a budget sink. You are still the pilot; PMax is just a very advanced autopilot. Never forget that. The success of this Atlanta Air Pros campaign wasn’t just Google’s algorithm; it was our team’s consistent effort in feeding it the right information and refining its direction.

Conclusion

The Atlanta Air Pros Performance Max campaign stands as a testament to the power of combining sophisticated automation with intelligent human strategy. By focusing on high-quality assets, precise audience signals, and an agile bidding approach, we shattered the client’s CPL and ROAS targets, proving that even with Google’s most advanced campaign types, thoughtful execution and continuous refinement are the true drivers of success.