Make It Happen Mondays

The episode contains solid substantive claims about AI ROI across specific domains (coding, customer service, SDR, recruiting) with concrete examples, but relies heavily on the same few frameworks and repeats them throughout. The 'pretend productivity' concept is valuable, but much of the middle section devolves into general agreement rather than novel analysis. Larsen provides useful specificity on the four areas where AI shows real ROI, but the conversation recycles talking points rather than building new insight.

“Those four areas are the following. One, coding. So cursor. You know, those one, you know, programmer can do the work of two or three people. They're seeing real ROI there. There is, I think, great case studies there. That's number one. Number two is customer service.”

“real roi, not productivity, which I'm not saying is bad, it's fun. But real roi, Those four areas are coding, customer service, SDR, recruiting.”

The critique of 'pretend productivity' is useful but not novel - the efficiency vs. results gap has been widely discussed in AI circles. The idea of mapping processes before buying tools is textbook Accenture consulting playbook recycled here. Larsen does offer a genuine contrarian take on not replacing entire sales teams wholesale (centering rather than extremes), but the core arguments lean heavily on conventional wisdom about fundamentals and process-first thinking.

“Watch where you are on the spectrum. Don't be on the two extremes because it's crazy.”

“People process technology, you know, it's kind of like the old school thoughts.”

Larsen's background is genuinely strong: he scaled InsideSales from $5M to $150M ARR, then built Customer into a $1.2B Meta acquisition, and is now actively operating Autonom with founder Dave Elkington. He has real practitioner scars ('I got scars all over the place of where AI failed') and is implementing this at scale across hundreds of clients. This is not a thought leader or career podcaster - he's in the trenches executing on AI agent deployment. His credibility is high and relevant.

“we went from about 5 million to 150 in ARR during the heyday of the sales acceleration era. From there, he went on to customer and helped build that into a $1.2 billion acquisition by Meta.”

“I've now implemented a hundred or hundreds of clients, it's work, man. It really is work to get it across a whole team to have that roi.”

Larsen provides concrete examples: the 17-minute customer service call, a company reducing recruiters from 30 to 7, $14M in cost savings from eliminating 400 BPO reps, and specific revenue figures for different AI categories (Cursor at $3B, Sierra/Intercom at $30-500M, AI SDR companies at $2-3M). However, many claims lack specificity - he references 'companies' and 'a client' without naming them, and some assertions about where AI works are stated without supporting data. The specificity is moderate, not exceptional.

“30 recruiters. They're down to seven, um, because AI is now doing 200, 300 hours of recruiting for them each month.”

“I've seen 14 million in cost savings in customer service, John. 14 million. They were they eliminated almost 400 offshore BPO reps that with using AI”

Barrows asks solid opening questions and follows up on concepts like 'pretend productivity,' but often lets Larsen deliver long monologues rather than challenging him. There are moments of genuine push (asking about the spectrum of automation, the transition strategy), but the host frequently agrees and amplifies rather than probes. The conversation reads more as two practitioners commiserating than one testing the other's claims. Barrows doesn't push back on the strength of evidence or ask harder questions about edge cases where the four-area thesis breaks down.

“But you know, I don't know. And I think now is, is a lot of people are having a crisis. You know, uh, they're just really trying to figure out like what do I do in this market?”

“So, what is pretend? What are people like kidding them? Like based on what you're seeing out there, what this pretend productivity thing”