"Earning the Right to Automate" | Eoin Clancy
Listen to episode
About this episode
About our guest — Eoin Clancy
Eoin Clancy is the VP of Growth at AirOps, the AI search and content engineering platform powering content workflows at companies like Webflow, Ramp, and Notion. He came up through the same GTM engineering path our audience is on — six years at Telnyx going from growth engineer to head of growth, then founded Build AI First in late 2023. At AirOps, Eoin scaled the company from $2M to $15M ARR in 11 months and runs the AirOps content engineering cohort that's become one of the most concrete training programs in AI search.
Core takeaways
Polarization Is Accelerating, Not Closing: The spectrum between top operators and the median has never been wider. Laggards are still figuring out Zapier; the top 10% are spending days in Claude Code and shipping at a different rate. AI doesn't replace people — it replaces the ones who don't use it well. Eoin's call to action: if you're actively learning today, you're already in the top 10%, and the only job is to stay there.
Information Gain Is The Whole Game In AI Search: Search engines and the LLMs powering them reward unique context that isn't in the training data. Where that context lives — in your customer calls, support tickets, internal Slack threads, expert heads — is now reachable through MCPs and agents. The competitive moat is the proprietary substrate, not the publishing pipeline. Cosine similarity is your enemy: undifferentiated content reads as gray, replicated, and the algorithm penalizes it.
AI/Human Collaboration Is A Modulator, Not A Switch: The "either let the AI do everything or do it all yourself" frame is broken. The right question per workflow: which steps reduce human involvement (legal review codified into an agent over time), which steps increase it (subject-matter-expert capture from your engineers and salespeople), and what's the hybrid that gets to 10 out of 10 quality. The pure ends get to 6 or 7.
Earn The Right To Automate: High-velocity experimentation comes before infrastructure. AirOps ran 5–10 manual webinars before codifying anything; Ramp ran years of trigger experiments before building internal sales tooling. Premature automation locks in process before learnings exist. Take 10 shots, learn from the misses, codify only what's repeatedly worked.
Top quotes
> "I'd much rather take 10 shots on goal, five of them hit. Two of them are failings that you learn more about your audience or the market on, and then you develop from there."
> "If you actually don't know what works and what doesn't, you're, you haven't earned the right to go automate it yet."
> "Internally, I'm known as the guy who drinks our own champagne. So drinking your own champagne is my preferred method."
> "If you're actively learning today, you're probably in the top 10%. And that's just where you want to make sure that you stay."
Referenced tools and resources
AI assistants & dev: Claude Code, Claude, ChatGPT, Cursor
AI search & content engineering: AirOps, Perplexity, Google AI Overviews, MCP
GTM & ops infrastructure: Salesforce, Clay, Slack, Gong, Intercom
Image generation: Nano Banana
Analytics: GSC, GA4
Workflow automation: Zapier
Timestamps
(04:23) Welcome to S2E4 — the SEO to AEO shift, Eoin's path from growth engineer to VP Growth at AirOps
(07:06) What's actually changed in GTM, content, and context engineering in the last year
(09:53) The polarization of skills — top operators racing ahead, the median falling behind
(12:00) AI won't replace people, it should empower them — Ross Simmons quote, anti-doom framing
(13:49) Hype vs reality in AI-powered content — where information gain actually lives
(17:25) Don't let your agent fan out a thousand pages — it does more harm than good
(17:55) Cosine similarity and the gray middle — why undifferentiated content loses in AI search
(20:18) Build vs buy plus bundling and unbundling cycles — the TV subscription analogy
(22:30) Bundling is the next move — tool consolidation with MCP and Claude Code orchestration on top
(24:33) The maintenance burden of vibe-coded apps — late-night pings and product-owner drift
(25:30) Pick your battles — the nano banana lesson on waiting out the better tool
(27:14) What Eoin's team optimizes for — high-velocity experimentation over single-bet projects
(30:31) Switching to AI search — why no acronym (AEO, GEO, LLMO) has won yet
(32:56) Top operator misconceptions — "AI content is bad" and "we have nothing unique to say"
(35:00) The hybrid is the answer — modulating AI and human involvement per workflow step
(41:07) What you can do this week — sales calls plus GSC plus Slack into Claude or ChatGPT
(46:49) Where content engineering sits at AirOps vs Webflow vs Ramp — and why it sits under growth
(48:42) True north metric for AI search — mention and citation rate, branded search as correlation
(50:34) Drinking your own champagne — AirOps runs its full GTM motion through AirOps
(52:13) Hiring the next GTM engineer — sit between sales and marketing, build infra plus boost team efficiency
(56:55) Closing — the gap is widening, keep the human in the loop, where to find Eoin
Where to Find Eoin
AirOps
Where to Connect with Jared & Matteo
Jared Waxman, GTM Engineer School Co-founder: LinkedIn
Matteo Tittarelli, GTM Engineer School Co-founder: LinkedIn, X, Website, Newsletter
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit gtmengineerschool.substack.com
More AI podcast episodes
Browse all →Want to find AI jobs?
Join thousands of AI professionals finding their next opportunity