I keep seeing the same story, told with minor variations, across my feeds. A team deploys an AI tool that captures every activity: every call, email, meeting. The tool reviews everything, analyzes everything, generates feedback and recommendations on each activity.
The author, usually the developer of the tool, brags: “Setup took minutes. No engineers, no integration project….”
And then they go through the math. They portray a manager, perhaps with 8-10 sellers, each sending hundreds of emails a day, making dozens of calls, sitting in several meetings. No manager could possibly review everything that’s being done and do the complete job. They simply don’t have the time.
We can all relate to that: monitoring everything our people are doing, providing feedback and coaching on all of it, is an impossible burden.
Often, they talk about even broader spans of control.
But AI always comes to the rescue, it can do it all. The math around manager versus AI capacity is presented to allow only one conclusion.
But these scare pieces miss a critical point: Is reviewing every single activity from each person ever the goal?
In my experience, it isn’t and it never was. Through all my years of selling and leading sales organizations, no competent leader ever attempted to review everything. The reason wasn’t because we lacked the capacity AI now provides.
We didn’t review everything because we didn’t need to. We reviewed selected things: the deals with the highest impact, the most important calls or meetings, activities where something important was at stake.
Experienced leaders know the behaviors we see in selected reviews are likely to be those we’d see in everything else the seller is doing. Sellers don’t choose to display those behaviors in select calls, meetings, or deals, then behave differently in everything else. If a seller isn’t doing the discovery work in an important deal, they probably aren’t doing it in other deals. If they are unprepared for one call, they are likely to be unprepared for every other call. Sampling a few deals or calls always verifies this. Experienced leaders know it’s unnecessary to look at everything.
What managers choose to review is also critical and communicates something important. When a manager can only look deeply at a few things, what they focus on communicates what matters. The team learns the priorities based on where the manager’s attention is, and what they deprioritize. But the AI tool that looks at everything means the choosing disappears. And with it, the signal about what matters most is gone.
When we treat everything as equally important, nothing is important.
The fact that a manager doesn’t need to inspect every activity frees up time to invest in the most impactful areas. Doing a deep dive into a critical deal, working with the seller to help them think it through and develop a much stronger strategy, has an impact not only on that deal. It enables them to look at every deal differently.
But the manager’s side of this is only half the problem, and it’s the half everyone argues about. Almost nobody asks about the recipient.
If the manager or an AI tool could review and coach on everything, what happens to the person receiving the coaching? At the end of the day, the impact of coaching on minutiae is that we’ve told the salesperson: “Just focus on these 25 things and your performance will improve!” And the next day, 10 of those 25 have changed, and the seller gets coaching on those.
Coaching on everything means we are coaching on nothing!
We have decades of research on how people learn and develop. Feedback and coaching that is prioritized, focusing on the top 1-2 things, has impact. Attention to those 1-2 things drives greater improvement in shorter time than any other approach. And after the person has mastered them, we move to the next two, then the next two.
Flood the person with feedback on everything and they have no idea where to focus, they can’t see their own improvement, and they don’t change. What happens instead is the same thing we see with activity metrics: the seller starts gaming the process rather than actually changing and improving in the areas that matter most.
There’s a deeper cost as we look at performance over a longer period of time. The act of taking deep feedback and coaching on 1-2 priority areas helps people develop their judgment. They learn to self-assess. At the end of each meeting, they think: What just happened? Did I miss anything? What would I do differently? It’s this friction, this focus in a few areas, that drives sustained growth and change. It’s this self-reflection that drives learning.
AI providing feedback before the seller has had a chance to reflect preempts this ability to learn and grow. It eliminates the need for reflection. If AI is doing all of this, the seller no longer has to pay attention. Through this, we are demolishing the desire and capability to develop. We are eliminating the very thing that makes coaching so impactful. Through these tools, through the ability to provide feedback on everything—we are displacing human judgment with the AI.
There’s one final aspect. An organization that is reviewing everything, making recommendations on each activity, is an organization that trusts nothing. Sellers know the difference between a manager coaching the select things that count and a manager using the system for surveillance. And people under surveillance don’t get better. They get careful. They get compliant.
None of this is an argument against the technology. Capturing comprehensive data is fine—possibly even valuable. Collectively, it surfaces patterns, enabling the manager to more quickly determine the 1-2 focus areas for each person’s growth.
But capturing everything, reviewing everything, and providing feedback on everything are three different things. The data existing doesn’t mean the data demands a response. Just because we can capture everything doesn’t mean we need to use everything. And this is where human judgment is so critical.
Suppliers of these tools revel in deployment speed: “rules in, tested, working in minutes…” But we live with these systems for months and years. The relevant question was never how fast the system deployed. It’s what the system does to the people inside it. If we’re constantly changing these tools or creating new ones, something else is broken, and the tools aren’t fixing it.
There’s a “tell” in all these stories the vendors publish. It’s what’s left out. There’s never a customer. There’s no outcome. It’s impossible to say what the seller learned or how they developed their judgment.
What we see are systems that overwhelm sellers and managers with data and feedback, systems that end up reviewing everything because they don’t know what’s most important, stripping managers and sellers of the ability to think critically and apply judgment because the machine is doing it for them.
And we call this progress.
Afterword: The opening sentence on this AI Generated discussion is hilarious! Then the discussion goes on. It’s a great discussion, really illustrating the ideas in this post vividly. Enjoy!
