AI productivity in sales is not the same as AI ROI. Saving time on research and prep is real, but if that time savings doesn’t translate to more pipeline or lower cost-to-acquire, it isn’t moving your business. In this episode, Gabe Larsen, CRO of Atonom, breaks down where AI is actually creating measurable results in sales and GTM teams, and where teams are confusing efficiency with impact.
About Gabe Larsen
Gabe Larsen is the CRO of Atonom, an AI company building always-on AI agents that run end-to-end work autonomously. His career spans leadership roles at Kustomer, Meta, and InsideSales.com, where he served as VP of Growth and hosted the Sales Secrets podcast. He’s known for a science-driven approach to sales that has helped him close major deals and train thousands of reps.
Check out Atonom at atonom.ai | LinkedIn | X | YouTube
Connect with Gabe: LinkedIn | Instagram
What you’ll learn
- Why saving time with AI is not the same as creating measurable business impact
- Where Gabe is seeing AI produce real ROI across coding, customer service, SDR work, and recruiting
- Why perfect prep docs and research bots still fail when reps lack the fundamentals
- How digital employees could change the way sales and customer-facing teams operate
- Why GTM leaders need to map the process before buying another AI tool
The SaaS golden age and what it got wrong
The SaaS growth era created a generation of sales leaders who optimized for activity and pipeline coverage without being forced to connect those metrics to actual business outcomes. When revenue was easy, efficiency was enough. When a few reps on a rising tide could close anything in the funnel, whether your tools were actually creating revenue didn’t matter.
That era is over. CAC has gone up, markets have compressed, and leadership teams are now being asked to justify every dollar in their stack. That pressure has landed directly on AI tools, and it’s exposing a lot of teams that bought efficiency without buying results.
Efficiency is not the same as results
Here is the distinction Gabe draws. A rep who used to spend 40 minutes on pre-call research and now spends 10 has genuinely gained something. But if that rep still can’t run a discovery call that moves the deal forward, the time savings doesn’t show up anywhere in the P&L.
AI productivity in sales becomes measurable ROI only when the time or cost savings connects to a revenue or cost outcome. A perfect prep document that a rep can’t use because they lack the fundamentals isn’t ROI. It’s a more expensive version of the same problem.
The question to ask about every AI tool in your stack: where exactly does this create revenue or reduce cost? If you can’t answer that specifically, you’re paying for efficiency theater.
Where AI is actually producing results
Gabe names specific areas where he’s seeing real ROI across the teams he works with.
Coding is the clearest example. The productivity gains in engineering are measurable, well-documented, and directly tied to output.
Customer service is producing results because the inputs are structured, the responses are largely rule-based, and the volume justifies the automation.
SDR work and recruiting are showing early positive signals, but the gains are more variable and depend heavily on the quality of the underlying process and the fundamentals of the people using the tools.
The common thread: AI performs best where the process is clear, the inputs are consistent, and success is measurable. Ambiguous processes produce ambiguous results.
Why good tools still fail when fundamentals are missing
The rep who can’t run a strong discovery call doesn’t improve because they walk in with a better prep document. The prep document surfaces information they don’t know how to use.
This is the most frustrating version of AI adoption on sales teams. Leaders invest in tools, reps use the tools, and the number doesn’t move. The tools aren’t the problem. The tools are accelerating a process that was already broken.
Getting the fundamentals right first isn’t the slow path. It’s the only path that makes the AI investment pay off.
Digital employees and where the market is heading
Gabe’s work at Atonom is focused on what he calls digital employees: AI agents that run complete, end-to-end processes autonomously rather than assisting humans with discrete tasks. Most current AI tools augment human work by handling individual steps. Digital employees handle the full workflow.
The implications for sales and customer-facing teams are real, and the timeline is shorter than most leaders are planning for. The teams building the operational infrastructure now will have a significant advantage when that shift accelerates.
Map the process before you buy the tool
The practical takeaway from this conversation: stop evaluating AI tools based on features. Evaluate them based on the process you’re trying to improve, and only automate the parts where automation actually creates revenue or reduces cost.
If you can’t draw the process on a whiteboard and point to specifically where a tool creates measurable business value, you’re buying a research bot. Research bots save time. They don’t move the number.
Map the process. Find the leverage point. Then buy the tool that targets that specific point. Everything else is noise.
What is the difference between AI productivity in sales and AI ROI?
AI productivity refers to time or effort savings from using AI tools. AI ROI refers to measurable business impact: more revenue, lower cost-to-acquire, or reduced cost-to-serve. The two are not the same. A rep who spends less time on pre-call research but still loses deals at the same rate has gained efficiency without gaining ROI. The ROI test is whether the time or cost savings connects directly to a revenue or cost outcome that shows up in the business.
Where is AI producing real ROI in sales and GTM teams right now?
Gabe Larsen identifies the clearest current ROI in coding, customer service, SDR automation, and recruiting, in roughly that order of confidence. The areas producing consistent results share a common structure: defined inputs, measurable outputs, and clear success criteria. Areas with ambiguous processes, variable inputs, or judgment-dependent outcomes tend to produce inconsistent results regardless of the quality of the tool.
Why do AI sales tools fail when rep fundamentals are weak?
AI tools accelerate existing processes. They don’t fix broken ones. A rep who lacks the ability to run a strong discovery call doesn’t improve by walking into that call with a better AI-generated prep document. The document surfaces information the rep doesn’t know how to use. If the underlying skill is missing, the tool produces a faster version of the same failure. Investment in AI tools without investment in fundamentals is a way to fail more efficiently.
What is a digital employee in the context of AI and sales?
A digital employee is an AI agent designed to run a complete, end-to-end workflow autonomously rather than assisting a human at individual steps. Where most current AI tools help reps move faster on specific tasks, digital employees are built to handle the full process without a human in the loop. Companies like Atonom are building infrastructure for this model, which represents a meaningfully different architecture from the productivity tools most sales teams are using today.
How should GTM leaders evaluate AI tools for their sales teams?
Start with the process, not the tool. Map the workflow you’re trying to improve and identify specifically where automation creates measurable revenue or cost impact. If you can’t point to a specific leverage point where the tool creates business value, the tool is unlikely to move your metrics. Evaluate every AI investment against one question: does this create revenue or reduce cost in a way I can measure? If the answer isn’t clear before you buy, it won’t be clear after.
John Barrows helps sales leaders decide whether to replace or rebuild their teams for the AI era. For 25+ years he has worked with the world’s most demanding sales organizations, including Salesforce, LinkedIn, Google, Amazon, and Okta, building the frameworks that became Filling the Funnel and Driving to Close. Today he advises CROs and VPs of Sales on AI readiness, team restructuring, and go-to-market strategy, drawing on exposure to every type of B2B sales organization over that span. John believes sales is a science, not a personality contest. His training focuses on the fundamentals that hold up regardless of what the market or the technology does next. He is the host of Make It Happen Mondays, author of I Want to Be in Sales When I Grow Up, and an LP at GTMfund.
Get John’s weekly sales insights: subscribe to the JB Sales Learning Lab Newsletter.
