Webinar Recap: Claude Code in Action with Noah Learner, SEO Community

In this AirOps webinar, Noah Learner joined Eoin Clancy, VP of Growth at AirOps, to build a Google Search Console Model Context Protocol (MCP) from scratch using Claude Code. They demonstrated how SEO teams can move from idea to working tool faster when they define the problem clearly and keep the initial build simple.

The biggest theme throughout the session: product thinking beats prompt tinkering when you want useful outputs.

Top 5 takeaways

1. Specs matter more than perfect prompts
   Clear success criteria helped Noah move from blank folder to working MCP faster.
2. Claude Code can teach while it builds
   Noah used Claude as both coding assistant and explainer for unfamiliar technical decisions.
3. Local-first MCPs are practical for real workflows
   A local Node.js and SQLite setup kept the build easier to secure and test.
4. A Search Console MCP can surface real SEO opportunities
   The demo covered low-hanging fruit, cannibalization, trends, and traffic-drop analysis.
5. MCPs are often the starting point, not the final product
   Text-based outputs are useful, but visual apps make insights easier to explore and act on.

Best practices and key learnings

This webinar gave marketers and SEO teams a clear path from curiosity to execution. The key lessons focused on scoping, learning fast, building safely, and turning raw data into decisions.

Start with a spec, not a giant prompt

Noah kept the build grounded by defining the endpoints, storage, language, and expected behavior before Claude wrote code. That approach gave the model a better target and gave Noah a faster way to judge progress.

"Talking with Claude Code is like… In general, it's all about the spec, and it's all about acting like a product owner or product manager." — Noah Learner

Think of yourself as a product manager, not just a prompt writer. Write down the tools, APIs, storage layer, and success criteria before you start. Ask Claude to fill in implementation details after you define what the finished tool should do. Interrupt and redirect the model when it starts going down the wrong path.

[Insert image asset: workflow graphic showing scope, planning mode, package approval, OAuth, and testing outputs.]

Use Claude Code as a builder and a teacher

Noah didn't treat Claude like a black box. He asked it to explain tradeoffs, hosting choices, and technical concepts so he could make better decisions while the build moved forward.

When you hit a concept you don't understand, ask "make me smarter about this." Use Claude to compare options across cost, speed, efficiency, and security. Change direction early when the build stops matching your goal.

Keep early MCPs local and narrow

Noah chose a local Node.js and SQLite setup because it lowered friction and reduced risk for an early build. That choice also made the tool easier to test, easier to install, and easier to keep inside strict data environments.

Key implementation decisions:

• Start with one or two endpoints instead of trying to support everything at once.
• Keep sensitive data local when your team needs tighter control over security.
• Build in a familiar environment like VS Code so you can see the repo, terminal, and files at once.

Turn Search Console data into SEO actions

The MCP did more than pull metrics from Google Search Console. It grouped, filtered, and analyzed data so Noah could find quick wins, diagnose drops, and spot patterns in user intent.

"You can find low-hanging fruit. You can detect cannibalization. You can find trending queries, find trending pages. You can even analyze a traffic drop." — Noah Learner

Look for low-hanging fruit SEO opportunities by combining impressions, rankings, and click data. Use grouped views to separate brand from non-brand, query groups, and URL segments. Treat Search Console as a source of audience intent, not just a ranking dashboard.

[Insert image asset: infographic showing search analytics, URL inspection, low-hanging fruit, cannibalization, trends, and traffic-drop analysis.]

Treat the MCP as the first product step

Noah made a strong point here: a working MCP proves the logic, but it doesn't always give teams the best interface for ongoing use. Once the analysis worked, he immediately wanted a more visual app that let him explore data faster.

Use the MCP to validate the workflow before you invest in a richer interface. Move to an app when your team needs visual exploration, live querying, or drill-down analysis. Build in small steps so each experiment teaches you what the next product should include.

Connect external data with internal context

Eoin's AirOps walkthrough showed what happens after you connect the data layer. He combined Search Console, AI search visibility, Slack conversations, and recorded customer context to turn insights into ranked content opportunities.

This approach works because it:

• Pairs quantitative search data with qualitative inputs from sales, support, and internal teams.
• Scores ideas by traffic, citation and mention rate scoring, and business relevance before you brief content.
• Pushes the winning ideas straight into content workflows so teams can move from insight to output fast.

Putting these MCP insights into practice

AI tools now let SEO and content teams build lightweight products without waiting on a full engineering cycle. That shift matters because teams can test ideas faster, learn from real workflows, and shape better tools around actual user needs.

Start with one narrow use case, one clear success definition, and one trusted data source. From there, you can expand into richer workflows that connect search data, brand context, and content production in the same system.

Why this matters for modern SEO teams

Teams that think like product owners will move faster with AI than teams that chase perfect prompts. Start small, define success, and let each build teach you what the next layer should do.

Want to turn ideas like this into repeatable SEO and content workflows? Book a call.

How AirOps helps SEO teams build on these insights

AirOps helps teams connect the same kinds of inputs covered in the webinar (like search data, AI visibility, and internal knowledge) then turn them into repeatable workflows. Your team can move from opportunity discovery to briefs, content creation, and optimization without stitching together a pile of one-off tools.

• Connect SEO, AI search, and internal context in one workflow.
• Turn insights into briefs, refreshes, and new content faster.
• Give your team a practical path from MCP-style experimentation to production-ready systems.

Ready to see how that works in practice? Book a call.