Sam Blond and Jason Lemkin joined together for a live SaaStr Workshop Wednesday looking at what has and what hasn’t changed in 2025 in GTM in general and in the age of AI.
They did a deep dive looking at top learnings not just from their own ecosystems, but from the top CROs Sam has interviewed recently on the CRO Confidential series: the CRO of Toast, the CRO of Databricks, the VPS of Windsutf/Codeium, the CRO of Wiz and more!
4 Unexpected GTM Learnings from CROs of Wiz, Toast, Codeium/Windsurf and Databricks
Through Sam’s conversations with leading CROs in his “CRO Confidential” series, several counterintuitive insights have emerged:
1. Wiz: Territory Matters More Than Tech Stack
Colin Yasukochi (Wiz CRO) demonstrated that even on the path to a $32B acquisition, territory design was more impactful than AI tools. Wiz discovered a direct correlation between territory size and productivity: the tighter the territories, the more effective their sales team became.
Key insights from Wiz’s approach:
- They maintained a strict 5:1 ratio of revenue growth to headcount growth
- When they compressed territories by 15%, they saw productivity increases of nearly 20%
- They found that reps with focused territories spent significantly more time with high-value prospects
- Their territory planning involved a sophisticated blend of:
- Company size parameters
- Industry-specific targeting
- Security maturity indicators
- Geographic considerations beyond simple regional boundaries
While competitors were chasing AI-powered sales tools, Wiz’s biggest efficiency gains came from this territory optimization. The unexpected learning: AI-powered tools took a backseat to fundamentals of territory planning, and this discipline helped them maintain predictable growth despite the pandemic forcing them to go hybrid.
2. Toast: Domain Expertise Isn’t Always Transferable. Even in the Age of AI.
Despite Toast’s need for restaurant industry understanding, Jonathan Vassil (Toast CRO) discovered something counterintuitive about hiring for their challenging low-ACV ($10K), low-margin (20-30%) sales environment with intense competition. Contrary to conventional wisdom, hiring based solely on restaurant experience wasn’t effective.
Toast’s surprising sales talent discoveries:
- Reps with 5+ years of restaurant experience often underperformed compared to those with none
- Deep industry knowledge sometimes created blind spots about how technology could transform operations
- Former restaurant managers tended to empathize too much with objections rather than overcoming them
- Their highest performers shared these traits instead:
- High emotional intelligence and ability to build rapport quickly
- Comfort with a consultative, educational approach
- Tenacity and resilience through lengthy sales cycles (up to a year)
- Ability to translate technical benefits into practical business outcomes
The most successful Toast reps weren’t restaurant veterans but those who could quickly absorb enough industry knowledge while bringing fresh perspectives on transformation. Toast built a specialized training program that could turn almost any disciplined, people-focused rep into a restaurant tech expert in under 60 days.
The unexpected revelation: people skills and adaptability often outperformed domain expertise. Their best performers came from diverse backgrounds like insurance, payroll services, and even education, but shared an ability to connect with restaurant owners’ challenges quickly.
3. Codeium / Windsurf: Enterprise Deals Need Different AI Support
Graham Murphy (Codeium CRO) discovered a fascinating dichotomy in how AI impacted their sales motion at different deal sizes. While scaling from 3 to 75 go-to-market team members, Codeium / Windsurf found that AI transformed their sales process in completely different ways depending on the size of the customer.
Kodium’s two-track AI sales approach:
- For SMB and mid-market deals:
- AI-powered demos drove incredible efficiency
- Automated personalization increased conversion by 40%
- Product-led trials with AI assistance closed deals with minimal human touch
- Some deals closed with just 15-20 minutes of human selling time
- For enterprise deals ($500K+):
- AI couldn’t replace what Murphy called “confidence engineering” – the ability to instill confidence in large organizations making significant commitments
- AI shifted from closing tool to trust-building assistant
- Top performers used AI to create detailed ROI models based on customer-specific data
- The most successful approach used AI for pre-call research and objection prediction, not actual customer interactions
- Security and procurement teams required extensive human reassurance about AI implementation
The unexpected finding: Codeium’s top enterprise reps spent more time on fewer deals, using AI behind the scenes rather than as a customer-facing solution. Their enterprise sales required a completely different approach to AI integration, focusing on risk mitigation and trust building rather than feature explanation or efficiency. This created almost opposite playbooks for their SMB vs. enterprise motions, complicating their enablement strategy.
4. DataBricks: Pricing Power Trumps Efficiency
CRO Ron Gabrisko’s surprising revelation during his time taking DataBricks from $1B to $3B was that focusing on pricing power created more leverage than operational efficiency. This went against the conventional wisdom in enterprise AI companies that were racing to implement AI-powered sales tools.
DataBricks’ pricing power strategy:
- They doubled down on technical sellers with deep expertise rather than trying to make generalists more efficient
- Their hiring profile shifted to favor engineers who had transitioned to sales over traditional sellers
- They invested heavily in a specialized pricing team with industry-specific expertise
- Key metrics from their approach:
- 22% higher average selling price when technical sellers led deals
- 40% reduction in discounting on enterprise contracts
- 35% increase in multi-product deals
- 63% improvement in expansion revenue within the first year
Their most important discovery: when selling to technical buyers (like data scientists and ML engineers), the perceived value gap between best-in-class and second-best solutions was enormous. By focusing their teams on establishing technical differentiation rather than sales efficiency, they could command premium pricing that dwarfed any operational improvements.
The counterintuitive lesson: While everyone else focused on using AI to improve sales operations and productivity, DataBricks found that positioning their solution strategically to command premium pricing created vastly more business impact (by a factor of at least 5x) than incremental operational improvements through AI. This shifted their investment from sales productivity tools to deep technical enablement for their sellers.
And a few other learnings:
1. The Mech Advantage: The Future is Human + AI Integration
Jason proposes that the killer app isn’t point solutions but “the mech” – a comprehensive system where:
- Digital AI assists the AE in every meeting
- AI remembers every conversation (unlike humans)
- AI serves as a true co-pilot, supplying product details, integration information, etc.
- AI handles prep, summarization, follow-up all in one system
Rather than 28 narrow point solutions funded by VCs, the future is an integrated system that does everything.
2. The In-Person Edge: 3X Conversion Rates That AI Can’t Beat
Both Brex and Toast found the same remarkable statistic: in-person sales generate 3X higher conversion rates than remote sales. At SaaStr, meeting sponsors in person didn’t dramatically change close rates but increased deal size by 30%.
Sam shared an anecdote about an SDR from TalkDesk who showed up at Brex’s office – the only person ever to do this, and he still remembers it years later. This raises the question: can AI help enhance these in-person experiences or will it push sales further into digital interactions?
3. Sales Efficiency Remains Constant Despite Tools. It’s Never Really Gotten Much Better, In Fact. Will AI Be Any Different?
Despite all the sales tech we’ve deployed in the past decade (Gong, Outreach, etc.), the fundamental efficiency metrics haven’t improved significantly:
- Wiz maintained a 1:1 ratio of revenue growth to headcount growth (5x revenue, 5x headcount)
- AE productivity and quotas haven’t dramatically increased with tool adoption
- Scratchpad saves time but doesn’t necessarily translate to more deals closed
This pattern suggests AI sales tools might follow the same trajectory: helping with tasks but not fundamentally changing productivity.
4. Technical Products Require Technical Sellers (Even With AI)
Ron (DataBricks CRO) maintains that technical founders still need technical sellers. This is particularly true when:
- Selling to developers
- Selling in regulated industries
- Vertical SaaS environments
- Dealing with complex solutions like data lakes
The critical question: can AI transform generalist sellers into subject matter experts? One startup found their digital SE tool exposed that their sales team was asking embarrassingly basic technical questions, leading them to fire half the team. But could better AI training transform this dynamic?
5. Cold Calling Still Works (And AI Might Enhance It)
Against conventional wisdom, Rippling found cold calling surprisingly effective. In a world of email overload, phone calls stand out. The key insights:
- Different touchpoints matter, and consistency counts
- People receive exponentially more cold emails than cold calls
- Being different creates cut-through in a crowded market
The AI angle here includes “voicemail before ring” technology, where AI leaves voicemails without calls, and real-time follow-up for inbound leads.
6. Sales Culture Matters More Than Tools Per Toast’s CRO
Why doesn’t everyone adopt high-conversion in-person sales tactics if they work so well? Sam and Jason suggest the answer is organizational culture and expectations:
Toast demonstrates a “ruthlessly efficient” sales culture where:
- Sales efficiency metrics are shared publicly during earnings calls
- Territory compression is used strategically to boost productivity
- Daily metrics are used to drive performance
This stands in contrast to companies where sales execs want “six-figure jobs that are work from home and mainly inbound driven.” AI tools won’t change this fundamental cultural dynamic.
The Bottom Line: AI isn’t yet revolutionizing sales the way it is transforming software development.
But the seeds are being planted for significant change, particularly around:
- Comprehensive co-pilot experiences
- Convergence of traditional go-to-market functions
- Enhanced training and enablement
- Smarter pipeline management
The winning approach isn’t just about AI tools—it’s about integrating AI capabilities into a ruthlessly efficient sales culture that maintains the human connection that drives conversions.
Keep watching for that “Cursor/Windsurf moment” in sales—it’s coming, but it hasn’t arrived yet.
And catch-up on the full CRO Confidential series here, deep dives with the CROs of Wiz, Rippling, Windsurf, Toast, Splunk, Zapier, Apollo, Gong and so much more:
