Giving a weak seller 10 more hours a week gives you 10 more hours of weak selling.
AI agents are the wrong answer. Hear me out.
AI agents can now help salespeople with account research, data entry, follow-ups, and other administrative tasks. Vendors claim these tools can free up a significant portion of a seller’s week. The promise is appealing: automate the repetitive work, give salespeople more time to sell, and watch productivity improve.
But there’s a question we don’t seem to ask often enough.
What if the problem isn’t that your salespeople don’t have enough time to sell? What if they don’t know how to sell effectively?
More time for your top performers. What about everyone else?
For your top performers, time can genuinely be a constraint. They prepare well for discovery calls, uncover business problems, quantify their impact, navigate buying committees, build business cases, and know how to move deals forward. Give them more time, and they might be able to work more opportunities without compromising the quality of their conversations.
That’s a legitimate productivity gain.
But what about the rest of the team?
Do we really believe most underperforming sellers miss quota because the day is too short? I don’t.
Some struggle because discovery is shallow. They fail to uncover the business problem, pitch features instead of business outcomes, and send proposals before establishing a compelling reason to buy. Others can’t navigate a complex buying committee or build consensus among stakeholders. Their pipelines are full of activity, but their deals don’t progress.
Give these sellers 10 extra hours a week, and you might simply get the same mistakes repeated across more calls.
This is also why I believe sales leaders need to look beyond the obvious explanations for missed targets. In my article, Why Your Sales Reps Miss Targets?, I explore how the problem may lie in the customers they’re targeting, the quality of their sales process, or other underlying issues rather than just effort. Adding more activity won’t fix a problem you’ve misdiagnosed.
More sales activity doesn’t necessarily mean more revenue
Consider a salesperson selling a customer experience platform to a retail business.
AI helps them research accounts, prepare meeting briefs, draft emails, and update the CRM. They can now spend more time engaging prospects and conducting discovery calls.
During a call, the salesperson asks about the retailer’s existing systems, the number of support agents, and the channels through which customers reach the business. They demonstrate the platform, explain its features, and send a proposal.
The meeting goes well. The prospect is interested. The opportunity moves forward in the CRM.
But the salesperson never uncovers the real business problem.
Perhaps the retailer is struggling with rising support costs, poor first-contact resolution, and declining customer satisfaction. Perhaps the priority is to reduce the cost per resolution without compromising customer experience. Or perhaps these issues aren’t significant enough to justify an investment right now.
Without understanding the business problem and its impact, the salesperson has little basis for building a compelling business case. The proposal becomes a list of features and prices rather than a solution tied to measurable business outcomes.
A few weeks later, the deal stalls. The customer doesn’t see a compelling enough reason to act.
Now imagine that AI gives this salesperson another 10 hours every week. They can research more accounts, conduct more discovery calls, send more follow-ups, and prepare more proposals.
But if their discovery remains shallow, they’ll repeat the same mistakes across more opportunities.
The problem was never the time available to sell. It was the ability to sell effectively.
This distinction matters because B2B sales isn’t simply a sequence of calls, demos, and proposals. As I explain in What Is B2B Sales? A Complete Guide to How It Works, different stakeholders play different roles in a buying decision. One person might identify the problem, another evaluate the solution, someone else control the budget, and another have final approval.
A salesperson can have a great conversation with one stakeholder and still lose the deal. The seller is trying to win a deal; the buyer is often trying to make a decision that the entire organisation can defend.
Automating the work around that conversation doesn’t automatically improve the salesperson’s ability to navigate it.
But can’t AI coach salespeople and improve their skills?
Fair point. And this is where the conversation becomes more interesting.
AI isn’t limited to automating administrative work. It can help salespeople prepare for meetings, analyse conversations, identify missed questions, practise objection handling, and receive feedback on their performance.
Imagine a sales manager reviewing a discovery call and noticing that the salesperson asked about the customer’s existing process but never explored the financial impact of the problem.
An AI call assistant could flag that gap and suggest questions the salesperson might have asked. A coaching tool could help the rep practise a similar conversation before the next customer meeting.
These are valuable use cases.
But identifying a skill gap isn’t the same as fixing it.
A salesperson might receive feedback that their discovery questions are too superficial. They might understand what they should have asked. That doesn’t mean they’ll be able to apply the learning in a live conversation, especially when a senior executive challenges their assumptions or a buyer is reluctant to share information.
Developing selling skills requires practice, feedback, repetition, and the ability to adapt to different customer situations. AI can support that process, but better selling isn’t an automatic consequence of deploying a coaching tool.
And the responsibility doesn’t sit with the salesperson alone. Sales leaders need to create an environment where coaching actually happens.
Promoting your best performers, for example, doesn’t automatically make them effective managers. A top AE may know how to close deals but still need to learn how to diagnose performance problems, coach discovery, and develop others. I’ve explored this in Stop Promoting Your Best AEs.
The same principle applies to AI. A tool can surface patterns and suggest improvements, but someone still needs to turn those insights into better habits and stronger execution.
Before buying another AI tool, diagnose the problem
If you’re a sales leader evaluating AI agents for your team, I’d suggest starting somewhere other than the product demo.
Listen to ten recent discovery calls.
Don’t just evaluate whether the salesperson followed the call structure or asked the standard qualification questions. Listen for the things that actually influence the quality of a sales conversation.
Are your reps uncovering the business problem, or simply documenting the customer’s existing process? Can they quantify the cost of the problem and establish its business impact? Do they understand why the customer needs to act now? Can they challenge assumptions without damaging trust? Are they connecting their solution to measurable business outcomes, or defaulting to a feature-led pitch?
Then look at what happens after discovery.
Are opportunities progressing because the buyer has agreed on the problem and the desired outcome? Are salespeople engaging the right stakeholders? Can they build consensus and establish a compelling business case? Or are deals moving through the pipeline because a demo was completed and a proposal was sent?
These questions matter because the stages of a sales process should reflect how customers make decisions, not simply what the salesperson has completed. If you’re interested in this distinction, my article Buying Is More Complex Than Selling explores the buyer’s side of the equation.
The answers to these questions will tell you far more about your team’s actual constraints than a productivity dashboard alone.
If the fundamentals are strong, AI can help your sellers become more productive. They can spend less time on repetitive work and more time doing what they’re already good at.
If the fundamentals are weak, you have a different problem to solve. Your team needs better coaching, more deliberate skill development, and a clear understanding of what effective selling looks like.
AI may be part of that solution. But it shouldn’t be used as a substitute for diagnosing and addressing the underlying problem.
Fix the selling before you scale the activity
I’m not arguing against AI agents in sales. Quite the opposite. I believe they can remove a lot of unnecessary work and help sales teams operate more effectively.
But we need to stop treating time saved as proof of sales improvement.
A salesperson who doesn’t know how to uncover business problems won’t suddenly become effective because an AI agent gives them more time. A rep who struggles to qualify opportunities won’t necessarily make better decisions because their account research is automated. And a seller who cannot articulate business value won’t close more deals simply because they can send proposals faster.
This also has implications for onboarding. If new salespeople aren’t taught how to conduct discovery, qualify opportunities, understand buying dynamics, and articulate business value, giving them access to a sophisticated sales tech stack won’t fill those gaps. I’ve written more about this in Sales Onboarding Is a Joke!.
Technology can support a good sales process. It can’t compensate indefinitely for the absence of one.
Before asking how AI can help your team do more, ask whether your team knows how to do the right things in the first place.
Because scaling an effective sales process can improve results. Scaling a broken one can simply make the problem bigger.
So, what’s really holding your sales team back: a shortage of time, or a shortage of the skills needed to turn conversations into revenue?
I’d start there before buying the next AI tool.