GTM Engineer School Podcast

Listen now | The founder of HyperGrowth Partners unpacks why hybrid GTM teams beat full automation, why true AI adoption stays under 1%, and why the next decade belongs to the Olympist of every field.About our guest — Guillaume CabaneGuillaume “G” Cabane is the founder and general partner of HyperGrowth Partners, where he and his team work hands-on with companies like Ramp, Neon, Zapier, AirOps, n8n, and Attention. He’s the recognized godfather of GTM engineering — pioneering the Clearbit reveal loop, personalized outbound, and the experimentation-as-system playbook before any of it was standard practice. Before HyperGrowth, G was VP Growth at Drift, Segment, and Gorgias, helping each one scale through multiple step-changes in revenue, and most recently served as interim CMO at Ramp through their run from $10M to over $1B in ARR.Core Takeaways* Hybrid GTM Beats Full Automation: G has yet to see a single scale-up successfully operate without a GTM team. He tested full automation against a hybrid setup (humans + AI) at Ramp himself, and failed to beat the hybrid baseline. Other people brought him their prompts and also failed. The AI-replaces-your-SDR-team posts are hype; SDR and BDR teams are still in active hiring across his entire portfolio.* True AI Adoption Stays Under 1%: Most marketing teams have not done any meaningful AI work beyond a GPT prompt or two. When G builds something for them and shows the output, they’re happy with the output but don’t want to learn how to build it themselves. The technologists who actually compound value with AI are a tiny fraction — G estimates less than 1% of the population — and the gap between them and everyone else is going to widen every quarter.* SaaS-for-SaaS Is the Risky Bet — Tech-for-Non-Tech Stays Safe: Build versus buy is shifting back to build. Niche subscriptions are getting harder to justify when you can vibe-code the same thing in an hour. SaaS sold to other software companies is the riskiest position — either LLMs start buying for them or enterprises start building in-house. Selling technology to non-tech companies (Ramp is the canonical example) stays safe; those buyers won’t suddenly start coding.* Be the Olympist of Your Field: There won’t be hundreds of millions of Python developers in the future. There will be a few thousand master-level specialists who can beat the best LLM in their narrow domain. The path for younger talent is to pick a niche, dedicate ten years from age 15 to 25, and become the world-class expert. Most Olympians are out of college by 25. Why would tech be different.Top Quotes“I tried beating that hybrid setup with a full auto and I failed. And other people have come to me and gee, there must be a way to, we can improve the prompts. And they have failed.”“True AI adoption, in the sense that you guys in this audience thinks about it, is going to be limited to a very small percentage of the population. I wouldn’t be surprised if it’s less than 1%.”“I think SaaS for SaaS is extremely risky because either like LLMs are going to buy from that land or because large enterprise and mid markets are going to stop buying and like just start building.”“You’ve got to be the Olympist of your field. That is how you survive. That is how you win.”Referenced Tools and Resources* AI & Building: Claude Code (referenced throughout as “Open Claude” and “Cloud Code”), GPT* Workflow Automation: Zapier, n8n* Content Production: AirOps* Data Infrastructure: Salesforce (via MCP), vector databases* Case Studies: Ramp (tech-for-non-tech archetype), Manifest (legal AI), Netic (applied AI in real-world spaces)Timestamps* (00:00) Cold open and intros* (04:05) What G’s three-and-a-half years at Ramp changed about GTM engineering* (06:45) What has and hasn’t changed about the growth team in 2026* (08:30) The AI-replaced-our-GTM-team posts are hype* (10:30) Hybrid wins, full automation loses — and G tested it himself* (11:35) Is the GTM Engineer role being diluted, or is it the same person with a new title* (14:30) Why true AI adoption stays under 1% of the population* (17:45) The two societies splitting around AI and how fast the gap is widening* (20:40) When everyone’s bot acts in their name, response rates collapse across every channel* (24:30) The effective craziness framework — tech love plus rigor plus creativity plus psychology* (28:10) What happens when LLMs start buying from LLMs* (31:50) Have you given Claude Code your credit card yet* (35:30) Are skills the new IP — and where defensibility goes* (40:00) Build versus buy is shifting back to build* (43:35) Why specialist agents still need specialists* (46:30) 2026 is the year you either build or step off the progress train* (50:30) SaaS for SaaS is the risky bet; tech for non-tech stays safe* (52:55) Trust the model output instead of needing to know everything* (54:30) Be the Olympist of your field* (56:50) What G would tell college graduates todayWhere to Find Guillaume* LinkedIn* HyperGrowth PartnersWhere to Connect with Jared & Matteo* Jared Waxman, GTM Engineer School Co-founder: LinkedIn* Matteo Tittarelli, GTM Engineer School Co-founder: LinkedIn, X, Website, Newsletter Get full access to GTM Engineer School at gtmengineerschool.substack.com/subscribe

Listen now | The co-founder and co-CEO of Sendoso unpacks the full ZIRP-to-agents arc, why physical mail still cuts through in an AI-saturated world, and how to plan teams when capability replaces headcount.About our guest — Kris RudeegraapKris Rudeegraap is the co-founder and co-CEO of Sendoso, the sending platform that lets revenue teams trigger physical gifts, direct mail, and branded swag right from their CRM, enrichment tools, and automation workflows. Sendoso has raised over $150 million (including a $100 million Series C from SoftBank Vision Fund), operates in 14 countries, and is trusted by revenue teams at Gong, Verkada, and Meta. Before founding Sendoso in 2016, Kris spent about a decade in software sales as an account executive at Talkdesk and other startups — where he was already hand-packing swag boxes and going to the post office on lunch breaks. The aha moment came when he saw the gap between CRM data and physical sending, and built the platform to close it. In February 2026, Sendoso’s Clay integration went live, making personalized physical sends as composable as any other Clay action.Core TakeawaysThe Full ZIRP Arc, Lived From Both Sides: Sendoso scaled the headcount-heavy model during ZIRP — over $150M raised, around 400 employees, a 50-person SDR floor. Then the market corrected and Kris had to rebuild with a fraction of the team, an agency layer for outbound, and AI agents handling what dozens of people used to do. Kris is one of the few founders who scaled the old way first, then publicly rebuilt — so the playbook he shares is earned, not theoretical.Ask “Who Would You Hire?” Instead Of “What Should We Automate?”: When Sendoso started building internal agents, the framing of “what should we automate?” produced a workflow list, not an investment-grade priority list. Reframing it as “if you could hire somebody today, what would they do?” turned the same conversation into employee-led role design. Field marketing wanted an associate to run 50 events — Sendoso shipped a field marketing agent. Account management wanted contract renewal busy-work eliminated — Sendoso shipped a contract analyst that pre-populates the next renewal from the last. Leadership wanted real-time win-loss instead of quarterly spreadsheets — Sendoso shipped a clickable analyst built on Gong, email, and Salesforce data.Capability Count Replaces Headcount As The Planning Unit: The old math was head-count × ramp-time × quota = pipeline. The new math is tasks × capability per agent + the humans who run them. Kris’s framing: capability count, not headcount. Forecasting changed too — instead of multiplying SDR seats by meeting targets, the new model accounts for how many Clay tables and plays are generating pipeline, which is harder to model but more honest to how work actually happens now.Inbox Full, Mailbox Empty — Why Direct Mail Becomes The Anti-Saturation Channel: Email saturation is structural and accelerating. Kris’s framing: he’s not anti-email, he’s anti-email-only. The omni-channel buyers’ journey includes email, social, ads, direct mail, phone, and field — and direct mail is the physical channel that AI can’t cheaply replicate. Sendoso’s smart-delivery agent already hits 95%+ confidence on whether to ship to home or office, and the Clay integration that just launched makes personalized sends a button-press inside any GTM engineer’s table.Top Quotes“I am not anti-email, but I’m anti-email only.”“if you could hire somebody today, what would they do? And that’s unlocked a lot of ahas for us because they’re like, ooh, I’d hire another field marketing associate for the 50 field events.”“you really need to think about like, what is the capabilities and the capacity of, of these agents, but also the employees that are kind of running them or reporting into them.”“the contract effort was like a hour long busy work and some of these come managers were doing, know, dozens and dozens of these renewals a week.”Referenced Tools and ResourcesOutbound + orchestration: Clay (with the new Sendoso integration), UserGemsWorkflow automation: N8N, Cloud CodeAgent infrastructure: Databricks, LangChain, AWS BedrockCRM and data: Salesforce, Gong (call transcripts ingested by the win-loss analyst)Internal interaction: Slack (primary surface for non-builders to talk to agents)Sending platform: Sendoso — Kris’s company, the gifting and direct mail platform behind the entire conversationTimestamps(03:07) Welcome to S2E8 — Matteo’s intro to Kris and the full transformation arc at Sendoso(05:36) Jared opens — what did the first version of Sendoso look like, and who was the first customer(05:59) Coffee Sender — the $5K Upwork project that became Sendoso(07:47) What’s stayed the same in GTM over the last decade(09:56) Hype vs reality — where GTM engineering is genuinely a step change(10:29) The Sendoso SDR redesign with Clay and UserGems — taking one step back to take ten forward(12:17) What changed when you went from 50 SDRs to AI-augmented hybrid roles(14:06) How forecasting changed when capability replaced headcount in the model(16:16) Where Sendoso’s narrative still holds — direct mail as the anti-saturation channel(16:46) “I am not anti-email, but I’m anti-email only” — the omni-channel thesis(18:05) Peak Sendoso — what the GTM motion looked like at 400 employees(20:55) The hardest decisions during the downsizing — and advice for founders sitting where Kris was in 2022(23:00) How the company structure transformed — GTM engineering as the new core(24:25) Early wins — the contract analyst that saves hours per renewal, the real-time win-loss agent(28:19) The backlog of agents — and the shift from “automation” to “digital colleague”(30:27) The current stack — Cloud Code, Databricks, LangChain, AWS Bedrock(32:26) Slack as the primary interface for digital colleagues(33:00) Are we cutting edge — and does it even matter(34:15) Will employees fear being AI’d away? What Kris is hearing from the team(36:10) The case for direct mail in every GTM stack — no more excuses(38:36) Where to send the gift — Sendoso’s smart-delivery agent at 95%+ confidence(42:43) The Clay integration that just went live(44:40) How the GTM org reports up — and why GTM engineering sits close to the CEO(48:08) Capability count, not headcount — the planning shift(49:51) Agency model and fractional experts — when to lean external(52:23) Closing — advice for VPs of marketing who haven’t started with gifting yetWhere to Find KrisLinkedInSendosoEmail: [email protected] Get full access to GTM Engineer School at gtmengineerschool.substack.com/subscribe

Listen now | The CEO of FullEnrich unpacks why one data vendor is never enough, how cold calling still wins when the list is right, and why audience-first beats GTM-engineering-first.Brought to you byThis could be your logo!Reach out to Jared to sponsor the next episode of this podcast and reach 5000+ modern GTM operator and founders.About our guest — Benjamin DouablinBenjamin Douablin is the CEO and co-founder of FullEnrich, the waterfall enrichment platform serving 3,000+ customers across the US, Europe, APAC, MENA and Latam. FullEnrich aggregates 20+ data vendors and queries them sequentially until you get the email or mobile number you actually need — then plugs into the CRM, data warehouse, automation tool, or MCP server you already run. Ben came up through pure sales tech with stints at Sales Ramp and Jellyfish before building an internal data-aggregation tool that became FullEnrich two and a half years ago.Core Takeaways* One Data Provider Is Never Enough: A generalist returns roughly a 30% enrichment rate against most addressable markets. Coverage breaks down by region, by vertical, and by segment. The winning shape is a managed orchestration layer that picks the right provider per query in real time, not a stack of disconnected vendors that operators have to manually benchmark.* Your List Is The Strategy: FullEnrich’s first 40 customers came through cold calls Ben placed himself. The unlock wasn’t the script. It was the prep — convincing yourself the call is a blessing for the prospect before you dial. If there’s no real issue to solve, don’t book the meeting. Open with a “because” that names a problem peers in their industry face, never their own performance, and let them choose the one that resonates.* Audience-First Beats GTM-Engineering-First: The discipline is a means; the audience is the end. FullEnrich serves three distinct audiences (GTM teams, lead-gen and talent agencies, product builders embedding the API). Each one has different acquisition motion, product needs, contract structure. Build squads around the audience, not the tooling — and let GTM-engineering capacity serve the squads where it actually moves the metric.* Stop Building 25-Slide Decks: The bar to execute has dropped; the bar to convince has stayed high. That asymmetry is the lever. If you’re a VP with an idea, hire someone to ship the zero-to-one in days, demo the result, then politick. Pitching internally before you’ve proven anything is the slow path that closes your learning window.Enjoying this episode so far? Subscribe for free to new posts.Top Quotes“I don’t know anyone that is really happy with their data providers, even if they multiply them.”“If you don’t know who you want to talk to, just don’t even think about it, don’t start, because you will get rejected, it will be hard.”“I never start with the tools or with the skills. I always start from my audience or end users.”“Having people that know how great look like is very important because the danger of AI is just very well articulated MBA type of intern that know how to write down very articulated way. But when you go into details, you need to be picky.”Referenced Tools and Resources* Data and enrichment: FullEnrich (aggregating 20+ vendors)* CRM and data infrastructure: Salesforce, Snowflake* Workflow automation: Clay, n8n, Zapier* AI and agents: Claude Code, MCP server (FullEnrich), HubSpot MCP* Movement origins: “Forward-deployed engineer” (Palantir), GTM engineer (Clay)Timestamps* (02:37) Welcome to S2E6 — the data infrastructure layer of GTM engineering, Ben’s bio (Sales Ramp, Jellyfish, FullEnrich)* (04:13) Ben’s intro — FullEnrich origin, 20+ data vendors, usage-based model, integrations everywhere* (07:13) What’s changed in GTM engineering — the Palantir comparison, why Clay launched the movement* (09:12) Beyond GTM — recruiter and talent acquisition use cases, Snowflake as system of record* (12:07) Where the movement is being oversold — ex-spray-and-prey agencies rebranding as engineers* (14:50) Does GTM engineering 10x the rest of the sales org? The hire profile question* (16:22) Claude Code hype — siloed adoption, FullEnrich’s MCP server, the chat-as-blank-page problem* (20:36) Why one data provider isn’t enough — coverage gaps, the benchmarking pain* (23:08) The orchestration layer — managed waterfall by industry, region, segment* (26:02) Cold calling as a growth lever — your list is the strategy* (28:36) Tone of voice on the call — don’t push, don’t sound salesy, ask permission* (29:51) Have a reason to call — the “because” frame that earns attention* (30:37) Don’t touch the prospect’s ego — name the problems peers face, let them pick* (31:50) What’s next for the enrichment industry — global coverage, verification, decision layer* (34:28) Hiring for GTM in the AI era — what separates good from great* (42:47) AI as the well-articulated MBA intern — polish hides shallow work* (44:20) How FullEnrich structures its GTM org — three audiences, three squads* (50:00) Agencies vs in-house — main motion in-house, $20–50K budget for new motions* (53:41) Stop-doing for VPs — demo the zero-to-one before building the deck* (55:56) Closing — audience first, list is strategy, AI with growth mindsetWhere to Find Benjamin* LinkedIn* FullEnrichWhere to Connect with Jared & Matteo* Matteo Tittarelli, GTM Engineer School Co-founder: LinkedIn, X, Website, Newsletter* Jared Waxman, GTM Engineer School Co-founder: LinkedInThanks for reading GTM Engineer School! Subscribe for free to receive new posts and support my work. Get full access to GTM Engineer School at gtmengineerschool.substack.com/subscribe

Listen now | The VP of Marketing at Airbyte unpacks why every company is now a go-to-market company, the warm outbound playbook that doubled pipeline, and what separates the operators worth hiring in 2026.Brought to you byThis could be your logo!Reach out to Jared to sponsor the next episode of this podcast and reach 5000+ modern GTM operator and founders.About our guest — Mario MoscatielloMario Moscatiello is the VP of Marketing at Airbyte, the open source data integration platform with 500+ connectors that moves data between sources and warehouses, CRMs, or activation tools. If you are syncing product signals, stitching together enrichment sources, or moving data between systems at scale, Airbyte is probably already somewhere in your stack. Mario came up the developer-tools path: growth at Pusher in London, then GitBook, then a stint as Principal at Flex Capital and board observer at Strapi, then Head of Growth at FluteStack. He had been advising Airbyte since 2020, four years before joining full-time, so when he stepped into the VP role he already knew the product, the community, and the motion cold. Since joining, Mario has built what he calls a warm-outbound motion that triangulates GitHub signals, product usage, and pricing-page intent through Common Room, Octave, and Clay. That approach doubled Airbyte’s pipeline growth rate.Core Takeaways* Foundations Before Automation Is The Real Step Change: GTM engineering is a step change only when product market fit is in place and revops data is clean. AI is a multiplier of whatever foundation exists. Bad data equals bad signal equals bad results, and AI in the mix is a multiplier effect. Pre-PMF teams hacking GTM with AI just create more damage faster. The two foundations are the PMM hat (personas, competitor, market, what works) and the revops hat (clean fields, clean enrichments, clean attribution). Without those two, more activity equals more noise.* Owning Workflows End-To-End Is The Leverage Unlock: Every team member can now own a workflow from idea to live. Paid SEO goes from idea to keyword research to blog post with assets and published in an hour. The SDR manager goes from prospect list to email copy to launched campaign in the same hour. AI compresses the wait between specialists. The right framing is force multiplier per role, not headcount replacement. Anthropic and OpenAI are hiring engineers AND account executives at the same time they ship tools that supposedly replace both. Five strong people doing the work of fifteen, not one star doing the work of ten — because that one star carries unrepeatable key-man risk.* Warm Outbound Is Signal Triangulation, Not Message Volume: The doubling of Airbyte’s pipeline growth rate did not come from blasting cold sequences. Two specific plays. First, events: dump the post-event lead list into Common Room, rank by who has signed up for the product or used the open source repo, and let SDRs only call the warm subset. Second, PLG signups: when an engineer signs up, outreach to them, AND prospect for decision-makers in the same org, AND warm those decision-makers with ads BEFORE the SDR call. The Twilio “Ask Your Developer” campaign at scale. Even when a motion has to be cold, the question is how to warm it up. ABM at Series B is now defensible if you know your audience.* Hire Barrels, Not Ammunition. Then Outsource The Deep Expertise: Mario hires generalists who can take a project end-to-end without a playbook over narrow specialists. The best SDR he ever hired was selling pest control door to door. Four traits to look for. Agency: just do the thing, do not tell me you will plan to do the thing. Curiosity: the playbooks that worked five years ago do not work now. Taste: AI brings the cost of writing copy and code to zero, and taste is what differentiates. Chip on the shoulder: something to prove. Then complement the in-house team with agencies for deliverability, paid media, and scaled outbound, so the team focuses on managing agents and workflows rather than becoming a CPM expert.Enjoying this episode so far? Subscribe for free to new posts.Top Quotes“Bad data equals bad signal equals bad results... When AI is in the mix, that’s a multiplier effect.”“For me, it’s really about drive over experience, agency over experience. I don’t care if you’ve never done this job. The best SDRs I hired was a person that was selling pest control products like door to door.”“AI brings the cost of writing copy, writing code, and everything to zero. What is going to differentiate is your taste and deep understanding of who you’re selling to.”“The biggest thing you should stop doing is just doom scrolling social media and sending your team everything that other people are doing... Just start listening to what your customers are saying.”Referenced Tools and Resources* Data infrastructure: Airbyte, Salesforce, Supabase* Signal hub: Common Room* Outbound orchestration and messaging: Octave, Clay, Instantly, Outreach* Audience sync and ads: Vector* Sales intelligence: Gong* AI assistants and dev: Claude Code, Claude, Cursor, ChatGPT, Lovable* Documentation and written culture: GitBook, GitHub* Frameworks and references: Barrels vs Ammunition (Keith Rabois, Founders Fund and Vinod Khosla), Twilio “Ask Your Developer” campaignTimestamps* (05:11) Welcome to S2E5 — Matteo’s intro on data infrastructure, Mario’s bio, the warm-outbound motion that doubled pipeline at Airbyte* (08:18) Investor to operator return — every company is a go-to-market company when software costs go to zero* (09:55) Workflow ownership end-to-end — paid SEO, SDR manager going idea to live in an hour, technology no longer the blocker* (11:28) Dev tools historical context — Auth0, Pusher, segment plus APIs, GTM engineering not new for dev tools* (13:33) Step change vs oversold — PMM and revops as the two foundations, AI as multiplier of garbage too* (16:50) The brand and product marketing comeback in the AI slop era* (17:48) Force multiplier vs headcount replacement — Anthropic and OpenAI hiring engineers AND AEs, key-man risk* (19:30) Markdown files in a GitHub repo vs notebooks on laptops — knowledge that survives the people* (23:14) Build vs buy and stitching — Vercel and Ramp can build, most teams should stitch via MCP* (25:00) MCP server unlock — making any tool talk to any tool, free from vendor integration roadmaps* (28:02) Mario’s overnight cron job that cleans his daily notes plus a CLAUDE.md file knowing his priorities* (28:43) Stack walkthrough — Common Room, Octave, Clay, Instantly, Vector, Gong, Outreach* (32:32) Warm outbound play one — events leads ranked through Common Room before any SDR call* (33:30) Warm outbound play two — PLG signups plus decision-maker prospecting plus warming via ads (Twilio at scale)* (35:33) ABM at Series B is now defensible if you know your audience* (36:41) Barrels vs ammunition (Keith Rabois) — drive over experience, the door-to-door pest-control SDR* (41:57) Outsourcing the deep expertise — deliverability agency, paid media agency, the team focuses on agents and workflows* (47:18) Four traits for great GTM engineers — agency, curiosity, taste, chip on the shoulder* (49:33) Stop doom scrolling, start listening — the Gong digest workflow as the one habit to startWhere to Find Mario* LinkedIn* AirbyteWhere to Connect with Jared & Matteo* Matteo Tittarelli, GTM Engineer School Co-founder: LinkedIn, X, Website, Newsletter* Jared Waxman, GTM Engineer School Co-founder: LinkedInThanks for reading GTM Engineer School! Subscribe for free to receive new posts and support my work. Get full access to GTM Engineer School at gtmengineerschool.substack.com/subscribe

About our guest — Eoin ClancyEoin 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 takeawaysPolarization 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 resourcesAI assistants & dev: Claude Code, Claude, ChatGPT, CursorAI search & content engineering: AirOps, Perplexity, Google AI Overviews, MCPGTM & ops infrastructure: Salesforce, Clay, Slack, Gong, IntercomImage generation: Nano BananaAnalytics: GSC, GA4Workflow automation: ZapierTimestamps(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 EoinWhere to Find EoinLinkedInAirOpsWhere to Connect with Jared & MatteoJared Waxman, GTM Engineer School Co-founder: LinkedInMatteo Tittarelli, GTM Engineer School Co-founder: LinkedIn, X, Website, Newsletter Get full access to GTM Engineer School at gtmengineerschool.substack.com/subscribe

About our guest — Kevin WhiteKevin White is the head of marketing at Scrunch AI, where he's building visibility infrastructure for the post-LLM web — analytics and optimisation that show brands how they're being represented inside ChatGPT, Perplexity, Claude, and Google AI Overviews. Before Scrunch, Kevin spent a decade marketing for the companies that defined the modern operator stack — Common Room, Retool, and Segment — and has advised teams at Ashby and Deepnote. He's one of the most pragmatic operators in the AEO space, with a working hypothesis that "there are no experts in this market yet" and a habit of running controlled experiments instead of taking anyone's word for it.Core takeawaysBots Are Your New VIP Visitor: Retrieval bots from ChatGPT, Perplexity, Claude, and Gemini are now hitting sites at meaningful volume — on Scrunch's own site, bot traffic exceeds human visits. Each retrieval call has a real human with commercial intent behind it. The work for marketers and engineers is shifting toward making sites cheap to crawl in tokens (Markdown and JSON over heavy JavaScript), so retrieval bots get more useful information and your brand surfaces in more answers.Wave 1 GTM-E Was Enrichment. Wave 2 Is Bespoke Trigger Hunting: The first wave of go-to-market engineering consolidated firmographic enrichment for SDRs. That layer is mature. The alpha now lives one step upstream — in identifying the specific sequence of triggers that puts a buyer in commercial-intent mode (a product action, a competitive event, a regulatory shift) and building the workflow to catch it at scale. Tools like Cloud Code have collapsed the build cost from a 6-figure software contract to a $20/month subscription.Hire The Vibe Coder Over The Content Writer: Vibe-coded interactive tools — site graders, prompt generators, real-time domain audits — are now more compelling outbound offers than gated PDFs. The same build gets reused as outbound asset, programmatic SEO play, sales enablement surface, and self-serve qualification interface. Kevin would hire a vibe coder over a content writer 100 times out of 100. Elena Verna at Lovable just hired one full-time. The move is no longer fringe.Reddit Citations Are TOFU. Lower-Funnel Intent Lives In The Long Tail: Most "win at AEO" advice fixates on Reddit and Wikipedia. Kevin's data says those surfaces cover top-of-funnel referential prompts — "what is X" — not the comparison and evaluation prompts that drive commercial intent. The prompts you actually want to win for are answered by long-tail niche publications. Optimise where your buyer's lower-funnel question lives, not where the volume looks biggest on a leaderboard.Top quotes> "We have more bot traffic now than we have human visits to our site."> "If I had the choice between hiring a vibe coder to build really cool tools and someone on the content side of things who's going to write a bunch of like white papers, I'm definitely going to hire the vibe coding person 100 times out of 100."> "There's not really an expert in the space. I would say instead of listening to me or others, go out there, create controlled experiments, see if those experiments yield the right kind of results that you're looking for."> "As marketers, you would typically dismiss in the past — bot traffic, this is not going to give me any information on the user experience. But now, it's like, I want that bot traffic, because there's a person with intent behind that bot."Referenced tools and resourcesAI assistants & dev: Cloud Code, Cursor, Claude, ChatGPTGTM & enrichment: Common Room, Clay, Clearbit, Apollo, Outreach, Artisan, 11xAI search visibility: Scrunch AI, Perplexity, Google AI OverviewsCitation surfaces: Reddit, G2, Trustpilot, WikipediaWeb infra & CDN: Cloudflare, Akamai, VercelAnalytics: GA4, PostHog, Similar WebTimestamps(00:04) Welcome to S2E3 — Kevin's path from Segment to Retool to Common Room to Scrunch AI(03:00) What's actually changed in GTM engineering — Wave 1 enrichment vs. Wave 2 trigger reverse-engineering(04:55) Reverse-engineering signals back from buyer commercial intent(09:25) The biggest failure mode — list-building beats message-coaching, every time(12:00) SDRs as multipliers, not dispensable headcount(14:00) Stack the plays as patterns emerge — and don't stop hiring once it works(16:25) Where GTM engineering is winning — and the untapped industries with greenfield opportunity(18:15) Cloud Code as today's tool of choice — and why Kevin won't be loyal next quarter(20:40) Worked example: paid spend × declining organic = Scrunch ICP(23:05) Who maintains the sprawl of vibe-coded tools? Enter the AI architect(24:43) The vibe coder vs. content writer hiring decision — 100 out of 100 times(27:30) One vibe-coded tool, four surface areas — outbound, SEO, enablement, self-serve(30:30) AEO, GEO, AI search — and why Kevin stays acronym-agnostic(33:14) There are no AEO experts yet — run controlled experiments instead(36:00) Reddit and Wikipedia cover top-of-funnel. Long-tail niche pubs cover lower funnel.(39:00) Two Scrunch personas — marketing (CMO/SEO) and engineering (CTO/CIO)(42:30) Bots are now your VIP visitor — Scrunch's own traffic data(44:00) Three categories of bots: traditional search, training, and retrieval(44:48) The token economy of crawlability — Markdown and JSON over JavaScript(48:46) The mental shift from "filter bot traffic" to "want bot traffic"(53:12) Hire deep IC experts and let them stack AI on top — the new shape of marketing teamsWhere to Find KevinLinkedInScrunch AIWhere to Connect with Jared & MatteoJared Waxman, GTM Engineer School Co-founder: LinkedInMatteo Tittarelli, GTM Engineer School Co-founder: LinkedIn, X, Website, Newsletter Get full access to GTM Engineer School at gtmengineerschool.substack.com/subscribe

About our guest — Zach VidiborZach Vidibor is the co-founder and CEO of Octave, the first context engine built specifically for GTM teams. Think of Octave as a GitHub for your go-to-market context — a structured home for your ICP, your value props, your competitive opinions, and every piece of institutional knowledge that should shape how AI and humans represent your company in market. Before founding Octave, Zach spent a decade as an operator inside companies that defined how modern GTM is done — LinkedIn, DocuSign, and Dropbox — seeing from the inside how strategy leaks and compresses by the time it reaches the frontline. Octave is his answer to that problem: a shared repo so every agent, workflow, and person on the team represents the company consistently.Core takeawaysContext is the moat, AI is the commodity: Every team now has Claude, GPT, and Gemini. Using AI is table stakes. The alpha has to sit one layer up — in your first-party context: your codified ICP, your competitive opinions, the customers you know you don't serve and why. The lens you put between your company and the market is the defensible thing; the model behind it is not.Wave 1 was filling fields. Wave 2 is codifying opinions: The first year of GTM engineering was dominated by enrichment — "we can now look up 30 DevOps engineers and drop them into an email." Problem solved. The next wave is interpreting what 30 DevOps engineers mean for how you act on that account — and the best teams aren't optimizing yesterday's assembly line. They're 3D-printing new motions from scratch with context as the starting input.If you skip ICP, you will regret it: LLMs are sycophantic optimists. Without your opinions baked in as constraints, they iterate plausibly against any problem indefinitely — "GPT splits atoms until infinity." ICP isn't a static deliverable; it's a testable hypothesis the system grades itself against. Teams that skip this foundation end up producing AI-slop at scale.Retire the set-and-forget mindset: The market's resting heart rate has gone from 60 to 120 beats a minute. VPs who still expect high conviction before every decision will lose to the ones who shift on a dime. Ten new competitors arrive per quarter. Category-defining tools appear on one-month cycles. The mindset for 2026 isn't "decide once"; it's "decide fast, shift faster, keep your context repo clean enough that changing direction is cheap."Top quotes> "Using AI, that's not a differentiator, right? You have to put alpha on top."> "What about, instead of an assembly line, we 3D print this thing?"> "The resting heart rate of the market has gone from 60 to 120 beats a minute."> "Don't just expect the models to know what you know."Referenced tools and resourcesContext & Messaging: Octave (GitHub for your GTM context)Data & Prospecting: Clay, Salesforce, HubSpotAI & Agents: Claude, Claude Code, Claude Co-work, OpenAI (GPT), Google GeminiPast operator context: LinkedIn, DocuSign, Dropbox (Zach's prior career)Timestamps(03:10) Welcome to season 2, and why the season shifts from hype to operating system(04:00) Meet Zach Vidibor — co-founder and CEO of Octave, the context engine for GTM teams(06:20) Why context is the lens between your company and your markets(08:00) How GTM engineering has evolved in the last year — Wave 1 to Wave 2(09:50) Using AI isn't a differentiator — you have to put alpha on top(10:20) Why LLMs are sycophantic optimists and what that means for GTM work(13:10) Where GTM engineering is real vs. where it's oversold(18:30) The "3D print this thing" reframe — beyond assembly-line optimization(19:30) Claude Code hype and what it means for solo operators vs. the enterprise(22:10) The coordination problem Claude Code creates in scaled sales orgs(25:00) Cut vs. grow — why you need to pick the goal before deploying AI(26:00) "Slice the world into 50 subverticals" — what different goals actually look like(28:00) Defining context engineering, and why ownership is still TBD(30:00) Institutional knowledge: the "we don't sell to higher ed" problem(34:30) The infrastructure monitoring case study — developer outbound at scale(36:00) Brownfield vs. greenfield routing and persona-specific context(39:40) Why the goal is positive interactions, not just demos(41:00) Matteo on the PMM perspective — teams that skip ICP(42:30) "If you skip, you will regret"(43:30) How Octave structures its GTM team — forward-deployed engineers reporting to sales(47:00) Hiring bar for great GTM engineers — taste plus systems thinking(50:40) "Go-to-market is super quantum"(52:30) What VPs should stop doing immediately(53:40) The resting heart rate of the market — 60 to 120 beats a minuteWhere to Find ZachLinkedInClayWhere to Connect with Jared & MatteoJared Waxman, GTM Engineer School Co-founder: LinkedInMatteo Tittarelli, GTM Engineer School Co-founder: LinkedIn, X, Website, Newsletter Get full access to GTM Engineer School at gtmengineerschool.substack.com/subscribe

About our guest — Yash TekriwalYash Tekriwal is Head of Education at Clay, where he builds the programs, content, and partnerships that help GTM operators learn one of the most powerful — but complex — tools in the modern stack. Before running education, he was Clay's first founding GTM engineer, building out the core sales process, demo tables, and custom POC model that supported the company's back-to-back years of 6–10x revenue growth. A former high school computer science teacher, 2x founder (Lectureless, Radify Labs), and self-described expert generalist, Yash brings an unusual mix of classroom chops and operator instincts toCore takeawaysThe Layered Grunt Work Problem: AI doesn't delete grunt work — it moves it up a layer of abstraction. What used to be "write your weekly update" becomes "read everyone else's updates," and that shift unlocks more interconnected team conversations downstream. The leverage is in what becomes newly possible, not just in hours saved.Computational Thinking Beats a CS Degree: Having written Python doesn't prove you can think computationally. The real test is transferability — can you port a skill from Claude to ChatGPT, or from Python to JavaScript, with light pointers? If not, you've been given fish, not taught to fish. In a toolscape that reshapes every 90 days, transferable skills are the highest-signal indicator of top talent.Three Personas Hit Three Different Walls: GTM-background operators get lost in computational thinking and burn credits before they learn. Ops-background people with no-code experience need to learn workflow thinking — where Clay fits vs. where it doesn't. Engineers are often the hardest persona because syntax knowledge doesn't guarantee systems thinking; the abstraction layer is the new challenge.Automate by Category + Time to Value: Every automation either replaces manual work with identical output OR unlocks capability that wasn't possible before. Prioritise by the 5–10 minute test: can you get 80% of the way there? If yes, ship it. Optimise the last mile of any automation only after you've covered the broad gains. Pareto Principle is always in effect.Top quotes> "What I think AI is doing, which is still a step change forward, is it is moving the grunt work up one layer of abstraction."> "Just because you have a computer science degree does not mean that you know how to think computationally."> "We will probably enter the most entrepreneurial generation of the past couple of decades because you don't need all the things or the obstacles that were in your way of starting a business before."> "Take the 20% that you can automate to give you 80% of your returns on the things that you need. Do that for as many things as you can before you start to try and optimize the last mile of your automations."Referenced tools and resourcesSales & Revenue: Clay (Ads, Audiences in beta), Salesforce, AttentionAI & Agents: Claude, Claude Code, OpenClaw (formerly Cloudbot), OpenAIProductivity & Ops: Notion (custom agents, meeting recorder), Dust, SlackWorkflow Automation: N8N, Zapier, AirtableLearning: Clay University, Code Academy, Algorithms to Live By (book)Timestamps(02:34) Welcome back to the GTM Engineer Podcast season two, and why Yash is the right person to open it(05:20) How GTM engineering has evolved in the last year, and the Andreessen parallel to software engineering titles(07:19) The evolution from software engineer to frontend to forward-deployed — and what that means for GTM engineering next(07:38) What Yash would miss most if GTM engineering disappeared tomorrow(08:48) AI doesn't delete grunt work — it moves it up a layer of abstraction(09:16) Where the hype is overselling and what still has to happen manually(11:29) The two-category framework for deciding what to automate first(14:37) The 5–10 minute time-to-value test for any new automation(16:00) Why we're entering the most entrepreneurial generation in decades(16:20) Why larger companies struggle — and the fear of job loss that blocks adoption(18:36) The transferability test: Python → JavaScript, Claude → ChatGPT(20:53) Playing with OpenClaw for EA workflows — meeting briefs from calendar(21:19) Claude Code vs Claude Co-work vs OpenClaw — what's actually different(24:59) Security and permissions when giving agents tool access(31:13) What's new at Clay: Ads GA and Audiences beta(32:34) Three learner personas and why each hits a different wall(37:17) Tool picks beyond Clay: Notion custom agents and Attention(42:48) The highest-leverage skills for GTM engineers this yearWhere to Find YashLinkedInClayWhere to Connect with Jared & MatteoJared Waxman, GTM Engineer School Co-founder: LinkedInMatteo Tittarelli, GTM Engineer School Co-founder: LinkedIn, X, Website, Newsletter Get full access to GTM Engineer School at gtmengineerschool.substack.com/subscribe

About our guest — Laurens NysLaurens Nys is the founder of GTM Sigma, a studio that builds AI-led GTM systems. He is a prominent voice in the GTM engineering community, known for his practical job descriptions for the role and his advocacy for dynamic, signal-based TAM lists. Laurens is an expert in N8N a workflow automation tool, and uses it to create powerful and efficient GTM motions.He is passionate about building systems that help companies acquire, retain, and grow their customer base. Laurens —  also a lead instructor at GTM Engineer School cohort 2 — is a a strong proponent of using automation to drive efficiency and growth.Core takeawaysGTM Engineering from First Principles: A GTM engineer is someone who builds systems to get, keep, or make customers worth more. It’s about applying an engineering mindset to the entire go-to-market process.The Convergence of Factors: The rise of GTM engineering is the result of several factors, including the end of the “growth at all costs” era, the increasing importance of efficiency, and the advent of AI.The System is Key: The specific CRM or tools used are less important than the underlying system. As long as the tools have a decent API, a GTM engineer can build an effective system around them.The Intelligence Layer: The core of a modern GTM stack is the intelligence layer, which for Laurens is N8N coupled with AI. This is where the data is processed, and the decisions are made.Finding the Constraint: To effectively design a GTM system, it’s crucial to identify the bottleneck in the customer journey. This allows the GTM engineer to focus their efforts on the area that will have the greatest impact.Top quotesOn GTM engineering: “A GTM engineer is just someone that builds systems to get, keep, or make customers worth more.”On the importance of systems: “I don’t really care about CRM as long as I can interact with it, meaning it has an API that’s not complete s**t. I’m fine.”On the modern GTM stack: “You have a database where obviously all the data lives. And then two, you have kind of the intelligence layer or kind of the brain of the operation. And for me, that’s N8N coupled to either workflows or some sort of an AI system.”On identifying the bottleneck: “The top of the bottle usually is the bottleneck. So outbound is a very common one.”Referenced tools and resourcesCRM: Atio, HubSpotLLM: OpenAI (ChatGPT), ClaudeEnrichment/Scraping: Bright Data, Rapid APIWorkflow Automation: N8NTimestamps(02:08) Laurens’ definition of GTM engineering(03:07) The factors that led to the rise of GTM engineering: efficiency and AI(04:33) Lightning Round: Favorite CRM (Atio, HubSpot)(05:13) Lightning Round: Top LLM (OpenAI/ChatGPT, Claude)(05:46) Lightning Round: Top enrichment tool (Bright Data, Rapid API)(06:54) Laurens’ top overall GTM engineering tool (N8N)(07:09) The most underrated GTM engineering tool (Bright Data)(08:22) The building blocks of Laurens’ GTM stack: database, intelligence layer (N8N), and interaction layer(10:14) Identifying the bottleneck in the customer journey(12:53) A deep dive into a GTM play for a company with a planning API(16:19) A walkthrough of the N8N workflow for the planning API use case(20:27) How to get good at N8N: project-based learning(21:53) Emerging skills for GTM engineers: GTM knowledge and technical fundamentals(23:16) The importance of mental models and learning how to think(26:18) Advice for aspiring GTM engineers: figure out your skill gaps and fill them(28:42) Why Laurens switched from Clay to N8N(29:48) How to build maintainable GTM systems in a rapidly changing tool landscape(31:33) The future of GTM engineering: will we be rebuilding systems every year?How to connect with LaurensLinkedInGTM Sigma Get full access to GTM Engineer School at gtmengineerschool.substack.com/subscribe

About our guest — Nico DruelleNico Druelle is the founder of The Revenue Architects, a consultancy that helps B2B SaaS companies build and scale their revenue engines. He is a leading voice in the GTM engineering community, recognized for his early advocacy of the role and his expertise in building signal-driven GTM motions for companies like Attention, Descript, and Preply. Before launching The Revenue Architects, he led GTM ops at Melio, scaling pipeline with advanced workflow tools.Core takeawaysThe Evolution of Rev Ops: GTM engineering is a consolidation of skills from traditional Rev Ops, Marketing Ops, and data engineering. The modern GTM engineer is an architect, a data expert, and an executor, all in one.The Power of Consolidation: A single GTM engineer can replace a team of specialists, leading to increased speed, efficiency, and ROI. This consolidation reduces friction and allows for faster iteration and validation of growth experiments.The Modern GTM Stack: The core components of a modern GTM stack include a data layer (data warehouse), an orchestration layer (Clay, Cargo), an engagement layer (Unify), and a CRM (Salesforce).The Scarcest Resource: The ability to set up and iterate on a holistic GTM system is the most valuable and scarce resource, not the data or the tools themselves.Top quotesOn GTM engineering: “Go-to-market engineering is a discipline of orchestrating first party data and third party data into a system of action, system of engagement to execute a given vision, a given go-to-market strategy.”On the GTM engineer’s role: “He’s that glue that comes in, just runs experiments, know, test things out, get some results filled back from the market and keep on iterating. And ultimately the uniqueness of that position is that he generates pipeline.”On the evolution from Rev Ops: “If rev ops was the before and go to market is the now or after, I think there is a bit of a consolidation of function of skills.”On the value of a GTM engineer: “The scarce resource is basically the ability to set up that system altogether as a holistic solution and iterate on it to build a defensible system to design growth. That is the real value in this.”Referenced tools and resourcesCRM: SalesforceLLM: OpenAI (ChatGPT), ClaudeEnrichment & Orchestration: Clay, CargoEngagement: UnifyWorkflow Automation: N8NTimestamps(02:13) Nico’s definition of GTM engineering(03:58) The before and after of GTM engineering(06:24) The GTM engineer as an architect, plumber, and electrician(08:19) Why the GTM engineer role is a consolidation of multiple roles(09:45) The benefits of consolidation: speed and less friction(12:10) Lightning Round: Favorite CRM (Salesforce)(14:18) Lightning Round: Top LLM (OpenAI/ChatGPT)(16:19) Lightning Round: Top enrichment tools (Cargo and Clay)(17:23) Nico’s top GTM engineering tools (Unify)(19:16) The core building blocks of Nico’s GTM stack(23:12) The role of a tool like Default for PLG companies(24:25) Tradeoffs in designing GTM stacks: modularity vs. speed(27:13) A deep dive into a PQL nurturing flow built for Descript(31:26) The importance of evaluations (evals) in AI model performance(37:53) Essential skills for aspiring GTM engineers: data literacy, tool fluency, and business acumen(40:10) How to acquire GTM engineering skills(42:14) The importance of feature engineeringHow to connect with NicoWhere to find NicoLinkedInThe Revenue Architects Get full access to GTM Engineer School at gtmengineerschool.substack.com/subscribe