Episode:
111
Claude + Google Ads
Listen on
Most businesses that try Google Ads fall into one of two camps: they give up after a few failed attempts, or they hand it off to a specialist and hope for the best. Alane Boyd and Micah Johnson chose a third option; they built the whole system themselves, using Claude as their AI-powered co-builder.
In Episode 111 of Automate Your Agency, the hosts share a behind-the-scenes look at every layer of the system: from the initial campaign setup to daily AI-driven analysis routines. Here's what they built, how it works, and what it took to get there.
The first decision Micah made wasn't about tools or platforms. It was about scope. Rather than trying to build the perfect campaign from day one, the team followed their "start small" methodology: get down to the basics, give Claude enough context, and iterate from there.
They fed Claude their offers, their messaging, examples of what had worked in the past, and some of their design style. From there, Claude researched keywords, shaped the campaign structure, and proposed ad copy directions; with the team steering at every step.
"It wasn't Claude built everything and we just pushed it live," Micah explains. "It took Claude's research and knowledge of paid ads, and we merged that with our expertise on what we're doing."
One of the most practical parts of this episode is the workflow the team developed for actually getting campaigns into Google Ads.
Claude was given read-only MCP (Model Context Protocol) access to Google Ads, deliberately not write access. This was a conscious choice. The team wanted Claude to function as an expert adviser, not an autonomous actor.
For the actual campaign creation, Claude Code handled the heavy lifting: writing JavaScript for tracking events, setting up Google Analytics key events, and critically, exporting perfectly formatted CSV files that the team could upload directly into the Google Ads desktop app and push live.
No copy-paste. No manual entry. No guessing whether the formatting was right.
This workflow also solved one of the most frustrating parts of Google Ads setup: the nuanced campaign settings. Which conversion actions should be primary vs. secondary? How do you assign values? What's the right bidding strategy for a brand-new campaign with no historical data? Claude walked the team through all of it: proposing recommendations, getting human approval, and refining until everything was right.
Once the campaigns were live, the team had another Claude-powered layer running in the background: a daily review routine that pulled data from Google Ads (via MCP), Google Analytics, Microsoft Clarity heat maps, and their lead database.
It was this routine that caught one of their most counterintuitive problems early.
They had built what Micah calls an "advertorial" landing page, a page that reads like a helpful blog post, with calls to action woven throughout. Traffic was coming in. Conversions weren't.
Claude flagged unusual "rage click" patterns in the heat map data within one to two days. When the team dug into Microsoft Clarity directly, they discovered the real issue: people weren't rage-clicking in frustration. They were clicking and dragging to copy text.
Visitors were finding the page through search, getting their question fully answered, copying the content, and moving on without ever needing to click a CTA.
"We were being too valuable right there at the very start," Alane said.
The fix was counterintuitive: give less. Tease the answer. Keep the hook but remove the resolution. The team identified the issue and made changes in under an hour; no agency, no week-long audit.
As leads started coming in, the team began analyzing their qualification data. They had built a four-tier system: fully qualified, partially qualified, needs review, and unqualified.
After 300+ leads, something wasn't adding up. What they had defined as "fully qualified" based on historical assumptions wasn't matching what they were seeing in the real data.
With Claude cross-referencing lead data, form responses, campaign performance, and qualification scores from their database, the team identified the issue in five to ten minutes: a significant portion of their "qualified" leads were individuals: students, job seekers, people who wanted to learn automation for their own careers, not businesses.
"We're in our silos too, people," Alane laughed.
The fix: simplify the qualification system to two tiers (qualified vs. unqualified), split the campaigns accordingly, and adjust keyword spend based on which terms were attracting which type of lead. Each change was logged as an experiment. Claude helped format the updated CSV files. The desktop app pushed it live.
One of the most underrated parts of this system is how it works for the people who didn't build it.
Alane wasn't involved in the technical build at all. But she's not in the dark. The team set up a private Slack channel where daily lead notifications and campaign analysis recommendations are posted automatically. She can follow along, react, and contribute ideas without needing to be in every technical decision.
For leadership reviews, a shared dashboard in Claude Cowork surfaces all the key metrics automatically. No one has to pull data before the weekly goals call. It's already there.
"We built it with Claude and Cowork, maybe took 30 minutes," Micah noted. "And nobody pays anyone to pull that data anymore."
Throughout the episode, one principle comes up again and again: Claude proposes. Humans approve.
It's easy to paint a picture of AI systems that self-optimize, run experiments autonomously, and just... work. That fantasy is tempting. It's also not what Alane and Micah built, and not what they'd recommend.
AI is only as good as the context it has. And the context that matters most: where the business is going, what the audience actually responds to, what experiments have already failed lives in the heads of the people running the company.
The goal isn't to remove humans from the loop. It's to remove the tedious, time-consuming work that keeps humans from doing what only they can do.
Alane and Micah are hosting two deep-dive workshops where they walk through the exact system covered in this episode:
🎧 Listen to Episode 111 of Automate Your Agency to get the full breakdown before the workshops, and come ready to build.
Alane Boyd and Micah Johnson just built a fully functional Google Ads lead system using Claude, and it's generating real leads without a specialist, an agency, or weeks of setup. In this episode, they pull back the curtain on every layer of the build: from campaign research and ad copy to conversion tracking, CSV exports, and daily AI-powered analysis routines.
If you've ever tried to run paid ads and gotten lost in the weeds of conversion tracking, keyword strategy, or Google Analytics event setup, you know how painful it is. Most businesses either give up or hand it off to an expensive agency. Alane and Micah chose a third path, and the results are already coming in.
In this episode, you'll learn:
If you're ready to stop guessing on paid ads and start building systems that actually work, press play now.
for Founders, Owners, and Leaders
AI evolves faster than most strategies can keep up. The Mastermind is designed for business owners to navigate it together.
Bi-weekly sessions covering the latest AI topics, monthly office hours, and direct access to facilitators.
Disclosure: Some of the links above are affiliate links. This means that at no additional cost to you, we may earn a commission if you click through and make a purchase. Thank you for supporting the podcast!
For more information, visit our website at biggestgoal.ai.
Explore our Courses/Programs:
is a visionary leader and serial entrepreneur with two successful SaaS exits under her belt. Recognized as a Top Leader under 40 and a finalist for Top Companies to Watch in 2021, Alane's expertise spans operations, sales, marketing, and technical skills. A published author and a mentor to many, she is passionate about impact-driven, result-oriented leadership.
is an accomplished entrepreneur and advisor, known for his ability to bridge the gap between business requirements and technical execution. With a knack for identifying system gaps and implementing solutions, Micah has been recognized as a Top Leader under 30 and has significantly contributed to scaling businesses for large brands and manufacturers across the US.
.svg.webp)
