From Finding $50M in Customer Support Calls to Building a GPS for Revenue Teams: Claudia Natasia's journey to founding Riley

From Finding $50M in Customer Support Calls to Building a GPS for Revenue Teams: Claudia Natasia's journey to founding Riley

Episode Description Summary Most companies are sitting on customer data that could reshape their product strategy—they just can't read between the lines to find it. In this episode of the Growth Wizards Podcast, host Geoffrey Lugli sits down with Claudia Natasia, co-founder and CEO of Riley, for a sharp conversation about turning messy customer data into revenue intelligence. Claudia started her career as a data scientist, where she built an embeddings model that analyzed thousands of customer support calls and uncovered a hidden insight: customers thought a loyalty program was actually a payments product. That single finding drove nearly $50 million in revenue growth in under a year. It also planted the seed for Riley, which she describes as a GPS for revenue—an always-on intelligence engine that tells revenue teams which accounts to prioritize, where churn risk lives, and what moves will actually move the needle.  Claudia and Jeff get into how Riley differentiates in a crowded AI market by building proprietary models that understand each customer's internal nomenclature from day one, why she refuses to do seat-based pricing for a data product, how 80% of their inbound unexpectedly came from rev ops leaders instead of product teams, and what she's learned about finding companies that are ready to buy now versus those still experimenting. If you lead revenue, run a GTM team, or are building an AI-powered product, Claudia's perspective on solving revenue problems instead of AI problems is worth your time. Timestamps * [00:27] – Claudia's journey from data scientist to CEO * [02:17] – The $50 million insight hiding in customer support calls * [04:35] – Riley's biggest GTM challenge: finding companies ready for intelligence, not just Q&A * [06:20] – How proprietary models and custom ontology cut through AI noise * [08:40] – The GPS metaphor: why we have it for driving but not for business * [12:52] – Why seat-based pricing contradicts data democratization * [15:06] – The ICP pivot: how rev ops became Riley's core buyer * [17:23] – Finding your voice to stand out when every channel keeps changing Takeaways * Listen beyond what customers say explicitly—build models that read between the lines of support calls and product usage * Differentiate your AI product by solving a business outcome (reduce churn, grow ACV) rather than selling the technology itself * If you want to democratize data access, don't gate it behind seat-based pricing * Let your ICP find you through inbound signals, then follow that data even when it contradicts your original thesis * Use a strong, opinionated voice backed by data to cut through crowded channels—every time you come back to it, it wins * Anchor technical products in metaphors people already understand (GPS for revenue) to make them instantly accessible Guest LinkedIn: https://www.linkedin.com/in/claudianatasia/ [https://www.linkedin.com/in/claudianatasia/] Company website: https://www.askriley.io

From Pre-Diabetic to Product Coach: Building AI That Actually Changes Behavior with Randhir Vieira, Managing Partner at Petra Coaching LLC

Summary Handing every employee an LLM license isn't an AI strategy — it's barely scratching the surface. In this episode of the Growth Wizards podcast, host Jeff Lugli sits down with Randhir Vieira, Managing Director at Petra Coaching and a 20+ year product leader with stops at Yahoo, Headspace, Omada Health, and HealthPy. Randhir shares what he's learned coaching digital health leaders through one of the fastest-moving technology shifts in a generation, and why the companies winning with AI are the ones pairing tools with governance, purpose, and a deep understanding of the user. He opens up about his own journey from pre-diabetic to losing 15% of his body weight through lifestyle change, and how that personal experience shapes his belief that meaningful transformation — whether in health or in business — happens in small, deliberate steps. The conversation also dives into what makes high-performing teams tick (from Saturday Night Live to emergency rooms), why psychological safety is a business-critical issue, and the single most impactful exercise a new manager can run with their team. Whether you're leading a health tech startup, navigating an AI rollout, or stepping into your first management role, Randhir's practical wisdom will leave you with something to try tomorrow. Timestamps [00:30] – Randhir's scenic path through five startups and three health tech companies [02:45] – Assessing where your company really stands on AI maturity [04:46] – Small AI habits that compound: building your own agent [08:16] – Why health tech can no longer stay in its silo [10:20] – Designing AI experiences that stand out from the noise [12:47] – Building feedback-friendly teams and the cost of silence [15:37] – The team norms exercise every new manager should run [18:49] – Parting advice: just get your hands dirty Takeaways * Tie every AI initiative back to a clear business metric, not just adoption numbers * Identify your most annoying recurring task and build an AI agent to handle it * Pair AI with human care rather than treating it as an all-or-nothing choice * Run a 60–90 minute team norms session to make implicit expectations explicit * Speak up earlier in your career — your perspective is more valuable than you think * Pick one small, meaningful problem and start experimenting with AI today You think you're behind on AI adoption, here's a visual that proves otherwise: https://x.com/NoahEpstein_/status/2025605338779496797/photo/1 [https://x.com/NoahEpstein_/status/2025605338779496797/photo/1] AI companies are pulling ahead, and here is why: https://tomtunguz.com/communication-tax-small-orgs/ [https://tomtunguz.com/communication-tax-small-orgs/] To learn more about Randhir and his work, visit www.petracoaching.com [http://www.petracoaching.com] Connect Guest LinkedIn: https://www.linkedin.com/in/randhirv [https://www.linkedin.com/in/randhirv]  Company website: https://petracoach.com [https://petracoach.com]

Greg Davis, Chief Revenue Officer at Markmonitor

Leading the Merge: CRO Greg Davis on Customer-First GTM, AI Signals, and the .brand Opportunity Summary Selling a complex offering while integrating two companies and riding the AI wave?  Greg Davis, Chief Revenue Officer at Mark Monitor Group, breaks down how to simplify complexity, scale trust, and drive growth. Greg shares how the Mark Monitor–Com Laude merger came together around a “customer-first, white-glove” ethos—structuring leadership early, building SME overlays, and aligning sales with operations for seamless handoffs.  He details how AI now powers their go-to-market: a 28-language website chatbot, regionalized content, intent signals for targeted outreach, and strict data guardrails.  Greg also unpacks two repeatable plays—leaning into GlobalBlock in 2024 to protect brands (and double expected growth) and mobilizing for ICANN’s new .brand application window with specialists, a focused landing page, and fast expert routing. He closes with timeless lessons: lead with problems (not products), co-author ROI with customers, and keep sales, marketing, and RevOps speaking the same language. Timestamps [00:45] – From finance and trading to CRO: coaching, simplifying complexity, and commercial growth [02:56] – Translating technical value to outcomes: multi-stakeholder selling and co-authoring ROI [04:47] – Inside the Mark Monitor–Com Laude merger: culture fit, GTM integration, and customer-first guardrails [08:38] – Building SME overlays and “white-glove” ops alignment to serve global accounts at scale [10:09] – AI in GTM: 28-language chatbot, regionalized site, intent signals, and strict data protection [14:58] – Customer-first campaigns that win: GlobalBlock as a protective play (and growth driver) [17:49] – The .brand window: why ICANN’s new round matters for trust, security, and control [20:57] – What he’d learned sooner + parting advice: problem-first selling and GTM alignment Takeaways - Lead with customer problems—then co-author success metrics and ROI with buyers. - Operationalize “customer-first” in mergers: set leadership early, align systems, and design white-glove handoffs. - Deploy multilingual chat + regional content to cut speed-to-lead and meet customers where they are. - Use intent signals to personalize outreach; retire generic sequences. - Add SME overlays across sales for deep, complex conversations without slowing deals. - Lean into market shifts that protect customers (e.g., GlobalBlock) even if they seem to cannibalize short-term revenue—and prep now for the ICANN .brand application window.

Stefan Kontschinsky, Product Management Director at Ping Identity

Pricing AI Without the Guesswork: Outcome-Based Models, Credits, and CFO Predictability Summary If pricing AI is breaking your forecast, this episode shows you how to regain control.  Ping Identity’s Product Management Director, Stefan Kontschinsky—formerly of PwC and Okta—breaks down how to price AI in a way CFOs can trust and customers value.  He explains the critical difference between what you sell (per user, per usage, per outcome) and how customers buy (contracting and consumption), then shares concrete levers to restore predictability: real-time consumption dashboards, clear limits, and smart overage policies.  Stefan makes the case for outcome-based pricing as a differentiator, with examples from Intercom, Fin.ai, and even SpaceX’s “cargo delivered” model. He walks through a practical 2x2 of pricing tactics (auto-upgrades, burst protection, committed capacity, pooled credits) and demystifies credits with a resource-unit approach he deployed at Okta.  You’ll also learn his monetization flywheel—quantify, create, track, and communicate value—and how AI can unify siloed sales/CS data to inform packaging and prove ROI. We wrap with what’s next: more optionality (bundles, seats, credits), how to hedge against rising model costs, and why reducing vendor lock-in is now a pricing strategy. Timestamps [00:21] – Guest intro and monetization journey: PwC to Okta to Ping Identity [02:34] – Pricing AI 101: what you sell (user/usage/outcome) vs. how you buy (contract/consumption) [05:07] – CFO predictability: consumption dashboards, limits, and proactive overage policies [06:34] – Stand out with outcomes: examples from Intercom/Fin.ai and SpaceX’s “cargo delivered” [08:14] – Tactics that work: auto-upgrades, burst protection, committed capacity, pooled credits, resource units [11:06] – The monetization flywheel: quantify, create, track, and communicate value across Sales and CS [14:01] – AI for pricing/packaging: un-siloing data to shape bundles and prove value [15:54] – What’s next: optionality, rising token costs, and reducing lock-in risk Takeaways - Separate what you sell from how customers buy to avoid defaulting to pure usage billing. - Build CFO-friendly predictability: show consumption in real time, set limits, and define overage policies. - Pilot outcome-based pricing where you can; it differentiates and aligns price with delivered value. - Offer committed capacity and pooled credits; use auto-upgrades and burst protection to balance risk. - Standardize value quantification and close the loop with CS to track baselines, outcomes, and ROI. - Use AI to unify pricing, sales, and CS data so packaging decisions reflect real customer behavior and value.

Robert MacDonald, Chief Product Officer at Strategic Education

From Partnerships to Agentic AI: Lumavac’s Robert Macdonald on Scaling Without Sprawl Summary If AI pilots are popping up across your org but nothing sticks, you may be building on cracked foundations, fueling “agent sprawl,” and trying to automate the past. Robert Macdonald, co-founder of Lumavac and former leader at Google, Pinterest, and Fox (plus a stint as Chief Product Officer in EdTech), shares how to make agentic AI real—and scalable. Drawing on 25 years in partnerships and product, Robert breaks down a practical path: start with one use case, fix technical debt, immerse cross-functional teams, and align on a shared playbook. He unpacks Lumavac’s hybrid approach—Velocity—a combination of tool and hands-on guidance that centers on spec-driven development, context engineering, and secure system agents. Robert also reveals his partnership blueprint, inspired by Stuart Diamond’s “Getting More”: build on authentic value exchanges, spot early red flags, and co-develop with product and engineering to move faster together. Expect concrete advice for CISOs, execs, and GTM leaders on establishing ownership, reducing chaos, and turning curiosity into consistent outcomes. Timestamps [00:45] – Robert’s path: Google, Pinterest, Fox, EdTech CPO, and why he founded Lumavac [03:22] – The partnership playbook: value exchange, transparency, and “Getting More” [07:41] – Early red flags: guarded partners, low investment, and resetting with better questions [13:24] – Top AI mistakes: cracked foundations, agent sprawl, and automating the past [16:12] – Immersion over memos: getting execs and non-technical teams hands-on [18:55] – Where to start: single use cases, ownership, and calming CISO concerns [20:54] – Inside Velocity: spec-driven development, context engineering, and system agents [27:39] – Differentiation and GTM: hybrid tool + services, co-building, and “start with one” Takeaways - Build partnerships on explicit value exchange and transparent goals; use “Getting More” to co-create outcomes you couldn’t reach alone. - Heed early red flags—rigidity, reluctance to share, and low time/resource commitment—and pause instead of pushing. - Fix technical debt before scaling AI; don’t build on cracked foundations. - Curb agent sprawl with standards, clear ownership, and a shared playbook. - Reimagine workflows, don’t just automate the past—pursue productivity and efficiency together. - Accelerate adoption through immersion: get cross-functional teams hands-on via spec-driven development, context engineering, and secure system agents; start with one use case, then scale.