Sales Doesn't Have a Coaching Technology Problem

It has a coaching consistency problem. The State of Sales Coaching in 2026 report from MySalesCoach highlights the size of the gap:

41%
of salespeople are never or rarely coached
28%
receive weekly coaching — the level that actually moves the needle
34%
of sales leaders have ever received training to become a better coach

Leaders themselves identified lack of time as the biggest barrier, alongside difficulties measuring coaching impact and a lack of confidence in coaching effectively. That is important — because when we talk about AI coaching, we tend to assume the starting point is a functioning coaching system that AI can improve. Often it isn't.

The manager has a packed diary. The 1:1 becomes a pipeline review. The pipeline review becomes a forecast interrogation. The forecast interrogation becomes a discussion about the three deals most likely to close this month. Everyone leaves having discussed the numbers. Very little actual coaching has taken place. AI doesn't automatically solve that. Give a poor coaching process better technology and you may simply create a faster poor coaching process.

Coaching and Deal Inspection Are Not the Same Thing

This is one of the most important distinctions sales leaders need to make.

Deal inspection — helps the manager

  • "When is the deal closing?"
  • "What's the next step?"
  • "Have you spoken to the decision maker?"
  • "Where does this sit in the forecast?"
  • "What's the competition doing?"

Coaching — helps the salesperson

  • "Why haven't you reached the economic buyer yet?"
  • "What's stopping you from securing a committed next step?"
  • "What do your discovery calls sound like — product or outcomes?"
  • "What pattern do you notice in the deals you lose?"
  • "What would you do differently in that conversation?"

The objective of coaching isn't simply to improve this deal. It is to improve the behaviour that affects this deal and the next twenty deals. That distinction is precisely where AI becomes interesting.

AI Can See What Managers Can't Possibly See

Imagine managing eight salespeople. Each person has ten meaningful customer conversations a week. That's 80 conversations — 40 hours of customer interaction every week. No sales manager can listen to all of it. So they coach from a tiny sample. Perhaps they join an important call. More commonly, they rely on the salesperson's description of what happened. AI changes that equation entirely. It can observe every conversation — not emotionally, not politically, not selectively. Continuously.

Patterns AI can identify that managers currently miss
One salesperson talks for 68% of every discovery meeting — leaving no space for the customer to reveal their real challenge
Another regularly discusses pricing before establishing value — creating unnecessary commercial pressure too early
Another asks excellent questions in the first call but rarely follows them up in subsequent meetings
Another has twenty contacts across their pipeline but almost every relationship sits at the same organisational level
Another repeatedly leaves meetings with vague commitments rather than mutually agreed specific actions

These aren't isolated call observations. They are patterns of behaviour. And patterns are where coaching gets powerful.

What AI-Enabled Coaching Actually Looks Like

The most useful model isn't AI as coach — it's AI as the intelligence layer that makes human coaching dramatically more effective.

AI Observes
Captures what is actually happening across meetings, calls, CRM activity, and accounts. Not what we think is happening. Not what someone remembers. What actually happened.
AI Identifies
Looks for repeated behaviours, missing behaviours, differences between successful and unsuccessful opportunities, and changes over time. Surfaces potential blind spots a manager would never see from weekly check-ins.
AI Recommends
Instead of giving the manager fifty dashboards, surfaces two or three specific coaching priorities. "Chris has had twelve first-stage meetings this month but only four have progressed to an agreed next action." Now we have something useful to discuss.
Manager Coaches
The human part. Why is this pattern happening? What does the salesperson think? What have they noticed? What could they try differently? What will they experiment with? The manager provides context, experience, challenge, and judgement. AI cannot provide any of this.
AI Measures
Watches what happens next. Did the behaviour change? Did meeting outcomes improve? Did opportunities progress differently? Suddenly coaching becomes something we can continuously improve rather than simply hope is working.

What the Data Says About Coaching Frequency

The MySalesCoach research found a striking relationship between how often people are coached and whether they hit quota. The relationship is difficult to ignore.

Quota attainment by coaching frequency
Weekly or more
76%
Monthly
56%
Quarterly or less
47%

This doesn't prove coaching frequency alone caused the difference — high-performing organisations may simply do more things well. But it points towards perhaps the biggest contribution AI can make. Not replacing expertise. Creating consistency. The best sales leaders probably already coach well. The problem is doing that consistently across every person, every week, while simultaneously recruiting, forecasting, reporting, and dealing with whatever crisis appeared that morning. AI can reduce that cognitive load.

Salespeople Already Know What They Want

In the MySalesCoach research, salespeople pointed clearly towards a combined model — not AI replacing the human coach.

48%
rated human coaching as extremely useful
39%
gave the same rating to human + AI combined
13%
rated AI-only coaching as extremely useful
"Data is not judgement. AI can tell me that I interrupted the customer six times. A good coach can help me understand why I do it."

There Is a Trap to Avoid

The temptation will be to build AI systems that provide sellers with answers. That is dangerous. Great coaching develops judgement. Poor coaching creates dependency. AI should help salespeople think — not remove the requirement to think.

AI that creates dependency

"You should contact the CFO."

AI that develops judgement

"You currently have relationships with five stakeholders, but none has budget authority. How confident are you that you understand the financial approval process?"

AI that creates dependency

"Ask about the customer's priorities."

AI that develops judgement

"During your last three discovery calls, most of your questions focused on requirements rather than business outcomes. What might you want to understand differently in your next conversation?"

That is a very different type of AI. It is not trying to become the salesperson. It is helping the salesperson become better.

"Perhaps the future isn't one where every salesperson gets an AI sales coach."

"Perhaps it's better than that — every salesperson gets a manager who suddenly has the time, evidence, and insight to coach them properly."

AI won't replace the sales coach. Used properly, it could help create millions of better ones.

Coaching is the C in the COACH Framework — Confidence Through Coaching is built on the principle that AI provides the intelligence layer while humans provide the judgement, challenge, and accountability that actually changes behaviour. Explore the complete AI for Sales guide →