AI in Sales: Stop Looking for Problems, Start Solving Friction

The enterprise AI adoption playbook is broken. Companies are hiring consultants, launching transformation initiatives, and building AI-first strategies, only to see 95% of large AI projects fail.

The issue isn't the technology. It's the approach.

In a recent conversation with go-to-market leaders Allison Macalik, Jeff Margolese, Hannah Bloking, and John Patton, a pattern emerged that challenges conventional AI adoption wisdom: The most successful teams aren't asking "how do we use AI?" They're asking "where does friction exist?"

The Real Barrier Isn't Technology—It's Fear

Hannah put it bluntly: "The biggest barrier to adoption that I continue to see is just fear." Not fear of the unknown, but fear of the concrete: will I get in trouble for using this? Will it replace my job? Should I even be using it for this use case?

This fear is compounded by a fundamental misunderstanding of how AI should be deployed. As John observed, "A lot of what we see with AI right now is a solution looking for a problem." Teams announce they have an agent, and when asked what it does, the answer is "anything you want." That's not helpful, it's paralyzing.

Embed, Don't Replace

The breakthrough insight from this conversation: AI adoption succeeds when it's embedded into existing workflows, not introduced as a separate initiative.

Think about how SaaS adoption works. A new tool that's siloed from enterprise systems dies on the vine. But when functionality is embedded where people already work? Adoption soars.

The same principle applies to AI. Hannah's team doesn't create new mechanisms to introduce AI, they put it into places that already exist. No separate training sessions on "prompt engineering." No new workflows to learn. Just friction, removed.

Jeff echoed this with a practical example: his digital sales team created AI twins for both the rep and the SE, trained them to monitor customer signals, recommend solutions, and even role-play conversations, all without disrupting existing workflows or requiring proprietary data access.

The Data Reality Check

Here's an uncomfortable truth that John laid bare: "The data we actually had inside of our system from the sales team was about logging hours and logging activities. Not helping you sell."

Most CRM data exists to keep managers off reps' backs and forecasts ready for Monday meetings. It's not designed to make anyone better at selling. Yet teams are building AI models on top of this data and wondering why they're not seeing transformative results.

The quality of AI output depends entirely on the quality of data input. If your data set reflects compliance behavior rather than sales excellence, your AI will optimize for compliance, not revenue.

As John put it: "If I had 6 hours to fell a tree, I'd spend 4 hours sharpening the axe." In AI terms, sharpening the axe means getting your data right first.

The Iron Man Suit, Not The Replacement

Perhaps the most powerful reframing came from John's analogy: "I'm not trying to take you out of the Avengers, I'm trying to give you an Iron Man suit."

This perspective shift is critical. Autonomous agents that completely replace humans are scary—and rightfully so. But AI that augments human capability? That makes someone wildly more capable than they were before? That's empowering.

Hannah described this as "leveling the playing field"—enabling people to do things they couldn't do before, filling in gaps where their partner's complementary skills would traditionally be needed. The technical seller can now handle more business conversation. The relationship-focused rep can dive deeper into technical details.

Three Principles for Practical AI Adoption

Based on this conversation, here's what's working:

1. Start with friction, not features. Don't ask "how can we use AI?" Ask "what's creating friction in our workflows?" Then see if AI can reduce it.

2. Embed into existing workflows. Every new tool, system, or process you introduce is a barrier to adoption. Put AI where people already work.

3. Fix your data before you scale. If your data set reflects the wrong behaviors, your AI will optimize for the wrong outcomes. Be brutally honest about what behaviors your systems actually capture.

The Vindication of Technical Selling

Jeff made a provocative observation: for 30 years, the traditional sales rep had cursory product knowledge and focused on relationships while SEs stayed in the background demonstrating functionality. That model is breaking.

In an AI-driven world, everyone needs to be fairly technical. CIOs are no longer non-technical business leaders—software has eaten the world, and AI has eaten software. The SE's ability to translate complex solutions into understandable, usable formats is becoming the core sales skill.

Some companies are already turning SEs into quasi-sales reps, adding sales managers only for negotiation and closing. The role that could explain, demonstrate, and contextualize technical capability is moving from supporting act to center stage.

The Empathy Edge

The most unexpected insight came from John's vision of "unlimited empathy." AI doesn't get tired, frustrated, or impatient. When an insurance customer chats with an AI agent, satisfaction scores are higher because the agent asks "Are you okay? Were you injured?" instead of rushing to case close.

For populations requiring repeated explanations, Alzheimer's patients, complex support cases, anxious customers, AI can provide patience that even the most empathetic humans eventually exhaust.

This isn't about replacing human connection. It's about using AI's unique capabilities, unlimited patience, consistent empathy, instant availability, to make human interactions better when they do happen.

What's Next

The leaders in this conversation aren't waiting for perfect AI strategies or complete organizational buy-in. They're experimenting. They have weekly AI practice sessions. They build POCs with customer data. They try things, fail fast, and iterate.

As Jeff noted: "I dropped out of a meeting on native AI go-to-market strategy to do this conversation." The irony? While companies debate top-down AI strategies, the bottoms-up experimentation is already showing results.

The future of AI in go-to-market isn't about transformation initiatives. It's about finding friction, embedding solutions, and making people more capable—one workflow at a time.

The takeaway: Stop treating AI as a strategy. Start treating it as a tool that solves specific, high-friction problems in places where people already work. The teams winning with AI aren't the ones with the biggest budgets or most sophisticated LLMs. They're the ones asking better questions about where work is hardest, and making it easier.

Watch the full conversation here.