Blog Posts:
Agentic RAG: retrieval that reasons for GTM
Most AI agents retrieve once and respond. Agentic RAG retrieves, evaluates whether the answer actually holds up, and retrieves again before acting. For GTM teams trying to deploy AI that reasons correctly about accounts and deals, that distinction is everything.
Classic RAG runs one pass: embed the query, fetch matching chunks, generate an answer. When the first retrieval misses, the answer is wrong and confident. Agentic RAG adds a decision loop: the agent plans what to look up, evaluates what came back, notices gaps, reformulates, and tries again. The SoK research paper formalizes this as a sequential decision process under partial observability, and it documents the failure modes that make this hard in production: compounding hallucination propagation, where one bad retrieval contaminates the entire chain; memory poisoning, where false information persists across runs; and retrieval misalignment, where topically related but decision-irrelevant evidence drives a confident wrong answer. For GTM, the stakes are higher than a wrong answer on a document search. When retrieval is chained to tools that write, a reasoning error becomes a wrong action against a real prospect. RevSure's position is that grounding the agentic loop in a resolved context graph, not a vector index, is what prevents those failures. A context graph keeps entities resolved, tracks state over time, and maintains evidence provenance, giving the agent something trustworthy to reason over rather than disconnected passages.
To understand why retrieval architecture is the make-or-break layer in GTM AI, read the full article here.
How to Build a Sales Hiring Profile Using Coaching Data
Most sales hiring profiles are written from intuition and borrowed job descriptions. Six months later, half the hires are struggling, and nobody can explain why because the criteria were too vague to predict anything. Your coaching data already contains the answer.
Every rep scored on every call for the past 6 to 12 months has a behavioral profile built from real selling data: which methodology criteria they mastered fastest, which behaviors correlate with quota attainment, which coaching patterns predicted long-term success. The article walks through a four-step process for extracting that profile: identify your top 20% of performers by win rate, pull their coaching score patterns across the first 90 days, isolate the 3 to 4 behaviors that appear early and resist coaching, and build a roleplay that tests those exact behaviors. The signal that emerges consistently: industry experience does not predict methodology adoption, coachability predicts long-term success better than initial skill, and talk ratio in the interview almost always predicts talk ratio on real calls. When a candidate with 2 years of experience scores 70% on the behavioral evaluation and a 7-year candidate scores 40%, the data says hire the 2-year candidate. The profile built from real rep performance is more predictive than any resume credential or interview impression.
For the full framework, including roleplay design, scoring logic, and how to update the profile quarterly as your market shifts, explore the complete guide here.
Why Most AI Pilots Fail to Scale (It's Not the Model)
A recent MIT study found that 95% of generative AI pilots fail to reach production. The barriers are almost never the model. They are data quality, disconnected systems, unclear governance, and lack of operational ownership.
- Poor data quality: Most organizations discover their CRM problems only after deploying AI on top of them, and by then, the pilot is already compromised.
- Disconnected systems: A pilot that runs beside the real workflow never changes an outcome. The model has to sit inside the process, not next to it.
- Unclear governance: Without defined rules on data access and risk, legal blocks the rollout or shadow AI spreads unchecked. The article points to a practical framework for getting this right before deployment.
- No operational ownership: Someone has to own the workflow, monitor performance, and measure business outcomes. The research shows exactly who that should be, and it is not the central AI team.
The article draws a clean line between "pilot theatre" (a demo that succeeds in controlled conditions) and operational AI (a system that changes how a revenue workflow actually runs). The fix is designing for production from day one: starting with a business problem, not a technology; building on clean data; integrating into existing systems; defining governance before deployment; and measuring cycle time, conversion rates, and revenue impact, not model accuracy.
For the full framework, including the pilot-to-production checklist and the operating model fix, take a look at the complete article here.
Podcast Episodes:
How to Hold the CRO Seat for 6 Years with Jeff Perry
The average CRO tenure at a public company is under 18 months. Jeff Perry, CRO at Carta, has held the seat for six years. This 48-minute episode of The CRO Spotlight Podcast, hosted by Warren Zenna, gets into what that actually requires and why most revenue leaders fail before getting there.
Jeff's central argument challenges the reflex to solve growth problems by adding headcount. At Carta, the move was the opposite: prioritize extreme efficiency, shift away from standard growth metrics, and invest in cross-functional alignment that outpaces quota-carrying rep additions. He covers how Carta evolved from a single equity management tool into a multi-dimensional platform serving different stakeholder groups, how he scaled entirely new business units simultaneously without losing operational focus, and why resetting leadership repeatedly (rather than fixing the underlying system) is one of the most common and most destructive patterns on revenue teams. On AI, his position is practical: trust-based selling is becoming more valuable, not less, as automation takes over the routine execution work. He also argues for aggressively auditing the technology stack and favoring in-house innovation over unnecessary third-party sprawl.
For a direct, experience-backed account of what it takes to hold the CRO seat long enough to actually change a business, listen to the full episode here.
CPQ in the AI era | with Colin Gerber, VP RevOps & Strategy at Socure
CPQ has gone through three distinct eras, and most organizations are still operating with tooling and mental models from the previous one. This 41-minute episode of RevOps Lab features Colin Gerber, who brings 17 years of CPQ experience and a clear-eyed view of where the function is actually heading.
Colin traces the evolution from Zuora's billing-wrapper days through Salesforce CPQ's deep CRM integration to today's platform-agnostic tools like DealHub that are built for consumption-ready, SKU-heavy, globally distributed pricing. He shares how Socure pressure-tested its pricing model in Excel for a full year before building anything in a CPQ system, and how they eventually designed a bi-directional entitlements object that ties CRM, CPQ, and billing into one source of truth. The episode also covers the most common CPQ implementation mistakes (over-complication and workflow bloat), how AI is being used to auto-generate Solution Readiness Documents, how RevOps's role in CPQ has shifted toward guardrails and pre-deal margin modeling rather than configuration ownership, and what consumption-based forecasting will require from CPQ infrastructure.
For any RevOps leader navigating CPQ selection, implementation, or the shift to usage-based pricing, access the full episode here.
Scaling Revenue Without Headcount with Natalie Furness
By: Attributed - A podcast by Dreamdata
Everyone wants growth. With budgets frozen and headcount off the table, the question is how to get it. This 56-minute episode of Attributed by Dreamdata features Natalie Furness, Founder and CEO of RevOps Automated, in conversation with host Steffen Hedebrandt on what revenue organizations actually need to scale without scaling headcount.
Natalie's argument is operational and specific: the lever is not more people, it is better systems, smarter data use, and automation applied to the right workflows. The conversation covers where AI is genuinely replacing manual work and where it is not, how RevOps leaders can unlock efficiency across marketing, sales, and customer success simultaneously, the common mistakes companies make when trying to scale too quickly, and what high-performing revenue teams are doing differently in 2026. She also addresses the governance and structural changes required for AI to compound in value over time rather than create a more expensive version of the same broken process.
For a practical, experience-grounded conversation on building a revenue engine that grows without compounding headcount, tune in to the full episode here.
Webinars:
Signal Rich, Action Poor: What Al-Native Revenue Teams Do After the Insight
By: RevGenius ft. Airspeed
Most GTM teams have more signals than they can act on. Intent data, engagement scores, product usage flags, CRM activity, call intelligence outputs. The problem is not a lack of signal. It is that most organizations have no reliable system for turning signals into coordinated action before the window closes. This session from RevGenius addresses exactly that gap.
The session covers how to identify which signals in your current stack are actually predictive versus which are generating noise, how to build workflows that convert signal into timely, coordinated outreach across sales, marketing, and CS, and what the operational infrastructure looks like for teams that have solved the execution problem rather than just the data problem. For RevOps leaders tired of hearing about intelligence tools that never change what reps actually do, this is the conversation that reframes where the real work lives.
To close the gap between what your data is telling you and what your team is actually doing, register for the full session here.
Growth Ain't a Vibe: The Math Behind Revenue Growth
By: RevOps Co-op ft. Revenue Operations Associates
Growth targets start in the boardroom. Making them real is a math problem, and RevOps is uniquely positioned to do the math. This session from RevOps Co-op, led by Matthew Volm (CEO of RevOps Co-op) and Steve Busby (Founder of RevOps Associates), breaks down the numbers behind sustainable revenue growth in a way that translates directly into operational decisions.
The session unpacks how pipeline generation, conversion rates, sales capacity, productivity, retention, and expansion interact as a system, and shows how changes to any one lever ripple through the others. Attendees will learn how to translate an ambitious growth target into its underlying operational requirements, identify the constraints most likely to prevent the business from reaching its goals, evaluate which levers are actually worth pulling versus which look good on a slide, spot faulty assumptions before they derail the plan, and communicate the math behind growth more effectively to GTM and executive leaders.
For any RevOps leader heading into annual planning season, register for the full session here and come prepared to pressure-test your assumptions before leadership does it for you.
