The Broken Algorithm: Why Sales Enablement Hasn't Moved the Needle | AmpUp

Why enablement hasn’t moved the needle, and why that’s finally about to change.

I was terrible at sales. Genuinely terrible.

Before AmpUp, I was CTO and founder at ThoughtSpot. I didn’t get near sales until we were past $50M in revenue. My involvement was always like a cameo. I’d walk in, dazzle the technical buyer with the vision, and slip out before anybody started talking about pricing.

When I started AmpUp as CEO, that stopped working. Now I was the rep. I had to earn the second meeting. I had to do discovery. And I was bad at it. So I hired a sales coach.

The moment everything clicked

He was reviewing one of my calls. I was trying to convince the customer their homegrown call recording system was inferior and ours was better. He stopped me: “Amit, the algorithm you’re using to decide what to say is fundamentally broken. It’s never going to work.”

I tried again. This time I told them their solution was made of glue and tape and wouldn’t scale. He said: “You did the exact same thing. Different words.”

The problem wasn’t that I lacked knowledge. Dazzling the customer was my habit. Under pressure, that’s what I did. And habits don’t get fixed by knowing better.

What he was drilling into me: customers don’t buy because you convince them. They buy because you diagnose their problems so well that they convince themselves. They see what they have to act on, and they see they have to act now.

Deals don’t die because of a knowledge gap. They die because of an execution gap. People don’t do the right thing under pressure.

The supply cabinet problem

How many of you have run an enablement session, great feedback, high energy, everyone loved it, and three weeks later everyone’s doing exactly what they were doing before? Most of you.

There’s the Ebbinghaus curve to explain it: 90% of what you teach evaporates within days. But I’d argue the deeper problem is that knowledge was never the gap in the first place. Execution is.

For the last 20 to 30 years, enablement has been about filling the supply cabinet: battle cards, pitch decks, use case libraries. You hope the soldier grabs the right weapon at the right time. It doesn’t work. Nobody is thinking about changing the architecture of how an organization learns. That’s where the money is.

There have been endless attempts to fix it. Give everyone AI tools. Hire GTM engineers. Bolt AI onto existing platforms. All of these automate the armory. They make it faster to fill the cabinet. They don’t change what happens in the room.

Mine is Dazzle. The moment a technical buyer says “can I see the demo?” every fiber in my being wants to start sharing my screen, and discovery dies right there.

I have a colleague who’s the opposite. Great relationship guy. Knows everyone’s kids’ names, their vacations. But when it comes to asking hard commitment questions, he’d rather keep smiling. His deals roll from quarter to quarter and never close. I call him Ted Lasso with a CRM.

Traditional enablement gives both of us the exact same session. Same slides, same framework, same role play. It doesn’t work for either of us, because our weaknesses are completely different.

One friend in enablement at Databricks put it this way: we run a hospital with 30 medicines and 100 patients. We don’t know which patient needs which medicine, so we give all 30 to all 100. And everyone dies.

Even if you know the rep, every sales room is different. If you’re talking to a healthcare company, you need to lead with a HIPAA-compliant reference customer. If you’re talking to a bank CDO, you need data exfiltration controls before they’ll have a real conversation. No rep can memorize all of this and surface the right thing in the moment. This is exactly the gap AmpUp’s meeting prep is built to close.

Why now? Memory.

For the first time, we have AI that’s smart enough to read every call, every CRM record, every email, and actually understand what’s going on. That part has been commoditized in the last few months.

The harder part, the part very few teams know how to do, is memory. Right now, your AI agents only get smarter when Anthropic ships a new model. They don’t get smarter by watching your deals.

Memory is about looking at every deal with one goal: understand what makes a deal progress, what makes conversion rate go higher, what makes it tank. Then build all of that like a chess engine. Every micro-strategy that worked gets assembled into a piece of memory that thinks with you in the moment.

What this looks like inside AmpUp: memory for every rep (their strengths, their weaknesses, where they choke under pressure), memory for every deal (who’s in the room, the minefields, the value drivers), and memory for the org (how deals get stuck, how they move, which tactics work in which verticals). All accessible instantly, without burning 10 million tokens to find one piece of information. It’s the same idea behind the AmpUp Sales Brain: pattern recognition that compounds with every conversation.

Build the Iron Man suit. Stop filling the armory.

Your job as an enablement professional is no longer to fill the armory with the right weapons and give everybody the same weapon. Your job is to build the exoskeleton that lives with the soldiers in the field and thinks with them in the moment, anticipates their problems before they happen.

What that looks like: the system knows my weakness is Dazzle. So when it gives me prep material the morning of a call, it brainstorms three discovery questions specifically for that deal and makes me practice them. So when the pressure comes, I’m prepared to hold the line and say, I’m not doing a demo until I understand the problem.

For the other rep, the relationship guy who avoids hard asks, it flips. His prep is about knowing exactly what commitment he needs to walk away with, with practiced lines ready to go. That’s what AmpUp’s sales roleplays are for: rehearsing the exact moment you’re most likely to choke.

We looked at roughly a million calls across different organizations and reduced them to about five million moments that actually mattered, moments that changed the trajectory of the deal.

Then we looked at the moments where the rep didn’t do the optimal thing. Here’s what we found:

42% of the time, the rep already knew how to behave correctly. They’d done it before in another deal. They just didn’t do it under pressure.

30% uplift. The experiment group outsold the control group within 5 weeks of a clinical-trial-style pilot with a $5B revenue company.

2-3x. Bottom quartile reps, when most engaged with the system, sold at the same rate as top quartile reps.

A real deal, a real miss

A rep is in a late-stage deal: $3.7M, economic buyer and champion both on the call. The economic buyer quietly drops this line: “We now have a new AI review committee that approves all AI purchases.”

The rep nods. Two minutes later, asks a couple of questions about paper processing.

A previous-generation AI tool scores this: Discovery 3.2/5. “Try asking more open-ended questions.” Directionally useful. Completely misses the most important moment that could kill the deal.

What the system should have flagged: AI governance committee, red flag. Who’s on it? What have they blocked before? What’s the risk to this deal? And it should have drafted the follow-up email before the rep even closed their laptop. This is the difference AmpUp’s sales coaching is designed to make: catching the moment that matters, not just the average.

Your job is changing. It’s getting better.

This isn’t all AI. AI can write an email, but it takes a lot of human judgment to write a good one. The enablement team seeds the hypotheses, shapes what matters, catches the mistakes, and the AI does make mistakes. Sometimes as silly as misgendering a rep. Sometimes raising objections that have never once appeared in the field.

The collaboration is the whole product. Without the judgment layer, the system doesn’t work.

Your job is moving from offline enablement to online enablement. From filling the armory to building the Iron Man suit. You’re going to be the designers.

The goal is to be the team that walks into a leadership meeting and says: we are going to move revenue by 30, 40, 50 percent over the next year, and here’s exactly where we’re attacking the gaps. That’s a very different conversation than QBR adoption metrics and session completion rates.

---

My system now knows what my weaknesses are. I’m genuinely better at discovery. But recently, I was in a meeting and I thought, I’m going to go for a fifth discovery question. “What’s the cost of not deploying this for another quarter?” It came out totally inauthentic. I’m still working on it. The system knows where I am. I’m getting better.

That’s what the system is supposed to do. Not make you someone you’re not. Help you become the best version of what you already are, when the pressure’s on.

---

Book a Demo with AmpUp

Ready to see how AmpUp can transform your sales team? Schedule a demo with AmpUp  and discover how AI-powered sales coaching delivers measurable results.

---

Frequently Asked Questions

Q: What is the difference between a knowledge gap and an execution gap in sales?

A knowledge gap is when a rep doesn’t know what to do; an execution gap is when they know but fail to do it under pressure. AmpUp’s research found that 42% of the time reps already knew the right move and just didn’t execute it in the moment. Traditional enablement fixes knowledge gaps, but most deals die from execution gaps.

Q: Why does traditional sales enablement fail to change rep behavior?

Most enablement focuses on “filling the supply cabinet” with battle cards, decks, and one-size-fits-all training, which evaporates within days per the Ebbinghaus forgetting curve. It doesn’t account for each rep’s individual weaknesses or the specifics of each deal. AmpUp instead delivers personalized, in-the-moment prep that targets where each rep actually struggles.

Q: What does “memory” mean in AI sales coaching?

Memory is the ability for an AI system to get smarter by watching your deals, not just when a new model ships. AmpUp builds memory for every rep, every deal, and the whole org, assembling the micro-strategies that move deals like a chess engine so the right insight surfaces in the moment.

Q: Does AmpUp replace the enablement team?

No. AmpUp is built on human-AI collaboration: enablement seeds the hypotheses, shapes what matters, and catches the AI’s mistakes. The judgment layer is the whole product, shifting enablement from offline content delivery to designing the system that coaches reps online, in real deals.