Lenny's PodcastJeanne DeWitt Grosser: How one GTM engineer replaced 10 SDRs
GTM engineers shadow top reps, then encode workflows as inbound lead agents; six weeks later Vercel held conversion rates with 1 SDR doing the job of 10.
CHAPTERS
- 0:00 – 5:50
Why GTM matters more in the AI era (and what listeners will learn)
Lenny sets up the episode around a surge in demand for go-to-market help. Jeanne frames how AI has intensified competition, making differentiation and execution in GTM strategically critical.
- •AI creates crowded markets where GTM execution becomes a key differentiator
- •Shift from “problem/pain” messaging to “why we’re better than alternatives”
- •Enterprises often buy to reduce risk more than to pursue upside
- •Preview of topics: GTM roles, GTM engineering, AI agents, tooling, segmentation, pricing, comp
- 5:50 – 8:44
Defining go-to-market as the full customer lifecycle (not just sales & marketing)
Jeanne broadens GTM beyond the classic sales-and-marketing view to include every function that touches customers or revenue. She argues GTM must become an integrated lifecycle rather than siloed Venn diagrams across teams.
- •GTM includes marketing, sales, technical sales roles, CS/support, and partnerships
- •Siloed strategies (e.g., different segmentation frameworks) create friction and inefficiency
- •Hyper-specialized GTM roles are likely to collapse into fewer, broader roles
- •Best practice: map the end-to-end journey like you would a product experience
- 8:44 – 11:23
What’s changed in modern GTM: consultative selling, forward-deployed engineering, and AI
Jeanne explains how consumption models and AI have pushed GTM from transactional selling to deeper, consultative engagement. She highlights forward-deployed engineering as a way to embed with customers and feed learnings back into product.
- •Consumption pricing shifts sales toward relationships and long-term expansion
- •AI buyers know they must change but often don’t know what “to”
- •Forward-deployed engineering helps customers implement and helps vendors learn
- •GTM increasingly involves “consulting-like” best practices and enablement
- 11:23 – 15:54
The rise of the GTM engineer: rebuilding outbound personalization with agents
Jeanne introduces the go-to-market engineer role through Stripe’s earlier attempt to build a ‘company universe’ for targeted outbound. What was too hard in 2017 becomes feasible now with AI-driven workflows and agents.
- •Stripe’s “Project Rosalind”: a database of companies + attributes for targeted outreach
- •Goal: Mad-Libs-style personalization driven by structured company signals
- •Earlier attempts struggled with false positives even with data science support
- •At Vercel, similar concepts work now by encoding workflows into AI agents
- 15:54 – 20:31
How AI agents actually get deployed in sales (human-in-the-loop, where it works first)
Jeanne explains how Vercel operationalizes agents: start with legible workflows, shadow top performers, and keep a human reviewer until quality is consistently high. She shares where agents help now (inbound/outbound research, install-base) and where they’ll take longer (deep enterprise).
- •Start with workflows you can write down (more deterministic = higher success)
- •Train by shadowing top reps to capture the real multi-tab workflow
- •Deploy with humans reviewing/editing before sending to protect brand quality
- •Use agents to raise seller “customer time” from ~30–40% toward ~70%
- 20:31 – 23:48
Real results: one agent replaces 10 SDRs (and improves speed-to-lead)
Jeanne shares concrete metrics from Vercel’s inbound agent rollout: a single GTM engineer built it quickly, the team kept conversion rates flat, and response speed improved. The SDR team shifts up the value chain instead of being eliminated outright.
- •Built by one GTM engineer in ~6 weeks at ~25–30% time
- •Reduced inbound SDR coverage from 10 people to 1 QA reviewer
- •Held lead-to-opportunity conversion flat while reducing touches/time to convert
- •Reallocated headcount to outbound rather than pure cost-cutting
- 23:48 – 26:25
Sales roles 101: SDRs vs AEs (and how AI changes the ladder)
To ground non-sales listeners, Jeanne defines SDR and AE responsibilities and how the buying motion changes by segment. She also argues AI will compress the traditional career path by automating repetitive SDR work and moving people sooner into higher-skill selling.
- •SDRs generate pipeline (inbound qualification + outbound prospecting)
- •AEs run the sales process and close deals
- •Selling complexity increases from SMB → mid-market → enterprise committees
- •AI can replace much repetitive prospecting, especially outside deep enterprise
- 26:25 – 33:59
When to hire GTM engineering—and what the ideal profile looks like
Jeanne argues GTM engineering forces earlier rigor: documenting sales best practices, creating playbooks, and instrumenting processes. She describes the ideal GTM engineer as someone technical with real GTM context (often former sales engineers) who can encode best practices into systems and agents.
- •Agents require a replicable playbook; startups often lack this despite traction
- •GTM engineering pushes earlier investment in process + data (like RevOps)
- •Founders typically hire first salesperson around ~ $1M ARR when repeatable
- •Best GTM engineers often come from sales engineering: technical + GTM fluency
- 33:59 – 40:39
The modern GTM tool stack: Gong + ‘Deal Bot’ for win/loss, coaching, and GTM “bug fixing”
Jeanne highlights Gong’s new power when paired with agents, describing how Vercel built internal bots that analyze calls, Slack, and emails to diagnose why deals are really lost. They evolve this into real-time coaching and weekly GTM ‘sprints’ to fix objection-handling and enablement gaps.
- •Run agents over Gong transcripts + internal artifacts to find true loss reasons
- •Example: “lost on price” was actually weak economic-buyer engagement and ROI proof
- •Real-time Deal Bot nudges reps in Slack (missing stakeholders, weak follow-ups)
- •Use agent insights to run GTM improvement sprints like engineering bug-fixing
- 40:39 – 46:37
Build vs buy for sales bots: why internal agents can be cheap, fast, and context-rich
Jeanne explains why Vercel often builds agents in-house: it’s faster than people assume, surprisingly inexpensive to operate, and better leverages proprietary workflows and context. She also notes the near-term fragmentation of AI tooling and procurement friction as companies experiment.
- •Some internal agents can be prototyped in days; more complex ones in weeks
- •Operating costs can be tiny compared to headcount (e.g., ~$1k/year for an agent)
- •Internal context (content, workflows, systems) is often the “alpha” for agent quality
- •Tool sprawl + procurement slows adoption; build can bypass some constraints
- 46:37 – 1:00:37
GTM as a product: designing a differentiated buying journey + practical tactics
Jeanne lays out her thesis: as software commoditizes, the experience of being sold to becomes a key differentiator. She shares concrete moves—like Stripe’s architecture whiteboarding discovery and always adding value—to turn sales from a transaction into an experience.
- •Differentiation increasingly comes from the buying experience, not just features
- •Replace boring discovery with collaborative sessions (e.g., whiteboard architecture)
- •Principle: add value at every touchpoint—even if the customer doesn’t buy
- •Tactics: unique insights, benchmarking, blueprints/well-architected guides, better discovery (listen more)
- 1:00:37 – 1:09:31
Segmentation primer: beyond SMB/mid-market/enterprise to growth, model, and workload signals
Jeanne explains segmentation as carving the market by how customers buy differently, then shows how Stripe and Vercel layered additional dimensions to improve targeting and messaging. She emphasizes segmentation as a company-wide language, not a sales-only exercise.
- •Base segmentation by size, then add dimensions that change value and motion
- •Stripe: add growth potential (consumption model) and business model (B2B/B2C/marketplace)
- •Vercel: add traffic/CrUX rank for “promote” and workload type (e-comm vs crypto vs SaaS)
- •Approach: analyze what drives revenue and where you repeatedly win; keep it to ~3 key attributes
- •Segmentation should inform product decisions too; align company-wide on the framework
- 1:09:31 – 1:14:00
Sales orgs engineers respect: product depth, GM mindset, and turning sales into R&D signal
Jeanne shares her litmus test for AEs: engineers should mistake them for PMs because of their product depth. She argues top GTM orgs are equal parts revenue generation and R&D, translating high-volume customer conversations into roadmap signal.
- •Litmus test: an AE should pass as a PM in front of engineers
- •Product depth builds credibility and improves selling quality
- •Sales as R&D: high customer touch volume can become roadmap signal
- •Discern signal vs noise; know when to objection-handle vs request true product work
- 1:14:00 – 1:22:37
PLG, pricing, and sales leadership realities: ceilings, pricing as product, and comp tradeoffs
Jeanne reframes PLG as powerful but capped, requiring sales motions to sustain growth as deal sizes increase. She then shares pricing lessons from Stripe and Vercel, and closes with her hot take on sales comp rigidity and why she likes mixing traditional sellers with consulting/banking profiles.
- •PLG is still relevant, but it has a ceiling; big deals rarely close self-serve
- •Don’t wait too long to build sales + outbound; it takes time to become predictable
- •Pricing should align value and cost; avoid freemium-by-default
- •Vercel example: unbundling enterprise features to self-serve increased efficient growth
- •Sales comp creates rigidity in fast-changing markets; diversify hiring backgrounds for stronger consultative selling
- 1:22:37 – 1:26:01
Lightning round: mottos, resilience, and the ‘yes/no/maybe’ lesson from diving
Jeanne shares the mottos that shaped her leadership style and how competitive diving taught resilience and repetition under pressure. She ends with a sales mindset shift: treat ‘no’ as useful data and avoid getting stuck in ‘maybe.’
- •Mottos: “When the going gets tough, the tough get going” and “Where there’s a will, there’s a way”
- •Diving as a precision, repetitive sport that forces immediate retry after failure
- •Obsession with excellence and replicable outcomes maps well to sales forecasting
- •Sales framing: “Yeses are great. Nos are great. Maybes will kill you.”
