Is AI Making Sales Worse, or Making It Easier to Stand Out From the Crowd? A Conversation with John Barrows

Most of the AI conversation in sales is about speed and volume. When I sat down with John Barrows, founder and CEO of JB Sales, for the Sales Operator Series, he proposed that AI is making it harder to get attention, but easier to stand out.

‍

John has trained sales teams at some of the best-known companies in tech for more than 20 years, and I've known him for over a decade. His view is that we spent years teaching a generation of sellers to behave like robots, and now AI is able to automate that work even further. The opportunity, according to John, belongs to the reps and leaders who go the other way.

Here's a recap from our conversation on why predictable revenue turned sales into an assembly line, why curiosity beats personalization at scale, and the three most insulting questions you can ask a prospect today.

You can also watch the full episode here on Youtube.

Why is it easier to stand out in sales when everyone uses AI? 

That was my opening question for John. He says AI has created personalization at scale in a better form than cadence tools ever did, but it's still mostly AI slop that all sounds the same. What cuts through is what he calls the "give-a-sh*t" factor. He told a story about Morgan Ingram, one of his early hires (and a recent guest on the Sales Operator Series) who was doing everything right with cadences and messaging but had plateaued. John's advice was that the results won't change until you start treating the prospect as a person who gets their kids to school, sits in traffic, and has a boss—not as a cadence or box to check.

AI can write a convincing email on a topic you've never experienced, he pointed out, but reps still tell him every email they receive has "something off" about it. That missing human factor doesn't scale, which is why John dislikes the phrase "personalization at scale." The point of personalization is that it isn't supposed to scale.

He also made a compelling point about activity metrics. The "16 to 17 touches" statistic cadence vendors love to cite never mentioned that the response to the 17th touch is usually "take me off your list." His point is that buyers aren't just ignoring bad outreach anymore. Rather, they're actively avoiding it and remembering the company that sent it.

Is curiosity the sales skill of the future? 

John thinks so, and he's careful to say you don't need to be the industry expert or the thought leader. You just need to know enough to be curious. When you ask, "Tell me more about that," "give me an example," "why does that matter," and genuinely want to know the answers, the person on the other end can feel it.

Can curiosity be taught? John thinks genuine curiosity is partly something you're born with, but you can also develop it if you get interested in your customers, your product, and your industry. For managers, that makes the question less about teaching techniques and more about getting reps truly interested in the problems they solve.

He also argued that the most valuable thing a rep can improve is business acumen, and that it's never been easier to build. In the past, you had to live through a merger to understand what a VP of sales at a $10 million company goes through when it's acquired by a $20 billion one. Now, a curious rep can ask ChatGPT or Claude about the phases of acquisition and what that buyer is probably dealing with, and come away with a decent understanding of the process.

The catch is that we've gone from a search engine to an answer engine. For curious people, that can be great. For people who just copy, paste, and send, John says, he has no idea what the value of the human is. "Fast food knowledge," if you ask me.

How did Predictable Revenue break sales? 

John was particularly blunt on this point. As much as he's benefited from the model, he believes Predictable Revenue "ruined sales" by turning it into an assembly line of SDR, BDR, AE, and CS roes. He gave a talk at Pavilion about a year ago called "Anything But Predictable." 

His argument it that we over-indexed on science in sales (process, tech, metrics) and lost the art (empathy, curiosity, and fundamental problem solving). We taught a generation of sellers to act like productivity-obsessed robots, and that's not their fault. As he put it, leaders are the ones who handed out the trophies.

He applies that thinking to the SDR role in particular: 

• SDR is the most expensive marketing you can buy. John shared stats from about a year ago that said that fewer than half of SDRs ever become AEs, and the average SDR stays at a company for 16 months. The idea of a cheap talent pipeline into AE roles isn't holding up.
• SDRs belong under marketing. John sees most of his clients doing this today, since the role is really about impressions.
• The next step should be customer success or account management, not AE. There, reps develop empathy, see what happens when a bad deal gets signed, learn closing in a relatively safe setting (renewals, cross-sells, upsells) and arrive as AEs with real references and use cases.

He also thinks buyers' tolerance for the ADR-AE-SE handoff chain is now "zero." If a rep can't add value in the first conversation, the buyer can go to ChatGPT or Claude, or talk to customer success or engineering instead.

John's prediction is that we're heading back to full-cycle sales, plus account management. That means sellers who own a patch, have AI agents and other resources behind them, and understand the P&L of their territory. He also thinks the sales process matters less than the buying process, and that stages are mainly useful for organizing the CRM and forecast.

I'd come to a similar view from a different angle. I've been hearing from folks that sales may swing back toward how it worked in the 1990s, when there was no SDR role and sellers owned relationships and became the face of the brand to an account.

What can sales leaders do right now? 

John offered two practical ideas on this point.

Run a monthly persona workshop. Here's the format he described.

1. At the start of each month, get your GTM org together for about two hours and pick a personal, such as CIOs in healthcare.
2. Spend about 30 minutes researching what that persona cares about, using AI or whatever tools you like.
3. Bring in an existing customer who fits that exact role for a short Zoom. Reps can ask contextual questions, not "what keeps you up at night," but "we read this about your role. Is that accurate?"
4. Break into groups to build messaging around the priorities. Marketing and ops assemble it into a sequence.
5. Split the accounts into test groups. For example, five reps with 20 accounts each gets you 100 accounts per sequence. One group runs all email, another adds phone calls, and another adds video.
6. At the end of the month, a team lead reviews the data and the team repeats with a new persona.

John's point here is that the exercise builds business acumen, aligns sales and marketing, puts reps in front of real customers, and gives you data on what works for each persona. A tactic that tanks for one persona doesn't mean the channel doesn't work.

Give reps a structure, not a script

John's observation is that younger reps have been structured their whole lives and excel inside a framework. Hand them a script and they'll read it word for word. Give them a structure to test within, such as "this month, we'll all ask this one question," and then collect data on whether it made a difference. That's a stark contrast to the nightmare of every rep running their own cadences and AI tools.

This suggestion resonated with me from the manager side as well. I regularly talk to leaders who dozens of managers who can't tell what anyone is doing or whether anyone is being held accountable. It's like the gaps you see as a sports coach. You can be great at drills in practice, but being skilled in isolation doesn't always translate to the game (or the live sales call).

What are the most insulting questions you can ask a prospect today? 

John's list of the three most insulting things you can say to a buyer right now are: 

1. Tell me about your business.
2. What are your priorities?
3. What keeps you up at night?

His reasoning is that a seller can get those answers from ChatGPT or Claude in mere minutes, so asking them puts all the effort on the buyer to explain rather than on the seller to do their homework. The alternative is to prep with AI and lead with a hypothesis. Like, "Before we get started, I noticed you're doing X. My hunch is you're dealing with a couple of scenarios like Y. Am I on the right path?" That gives the buyer an anchor to react to, and John says it earns far more credibility and detail than a generic open-ended question.

He backed it up with an experiment. About a year ago, he took 20 meetings with SaaS vendors on products he was genuinely interested in. At the start of each, he asked what the rep already knew about him and his company. Fourteen of the 20 couldn't say he does sales training. I've had a similar experience when I occasionally take calls from competitors: reps often have no idea we compete for the same business.

What I'm taking away

I think this conversation is a useful counterweight to the idea that the answer to AI is more AI. And to be clear, John is in no way anti-AI. But his point is that tools can't supply the care, curiosity, and business acumen that make a conversation valuable, and the old assembly-line model deliberately trained those characteristics out of a lot of sellers.

For leaders, the practical question is what rhythm you have in place to build those skills on purpose. A monthly persona session is a small, repeatable system a team can start next month, and that's the kind of consistent, structured practice we keep coming back to at Ambition.

To hear the whole conversation, including John's thoughts on the "core values exercise," where the fundamentals will come from when AI absorbs entry-level work, and his take on the infinite game, watch the full episode on the Sales Operator Series. You can find John at jbarrows.com, where you can get to his newsletter and YouTube channel.

You can find me here on LinkedIn.