We've become very good at imagining how AI can help people work faster. I'm more interested in what it could help them accomplish.
Imagine an Account Executive starting her Monday morning.
Her calendar is full. She has meetings to prepare for, customers to follow up with, and a growing list of people she hasn't spoken to recently. She enjoys her work, particularly the moments when she can help a customer solve a difficult problem or accomplish something they weren't sure was possible.
Unfortunately, much of her week is consumed just trying to keep up.
Somewhere in her portfolio, a customer is struggling with a problem she could help solve. Another is approaching a decision where her experience would be particularly valuable. Several are progressing just fine on their own. She has no reliable way to distinguish them without spending considerable time investigating each account.
Now imagine her organization introduces AI.
She can prepare for meetings faster, summarize account histories, draft emails, and automate CRM updates. These are useful improvements, and I would happily take them.
But what happens to the time she saves?
In a May 2026 Gartner report, the firm reported that AI tools were saving sellers an average of 4.8 hours per week, yet 72% of sales organizations reported low reinvestment of those savings into high-value sales activities.
We can give people time back without necessarily making them more effective.
My question is: Did we improve her ability to help customers, or did we just create room for more activity?
What are we actually trying to improve?
I understand why sales organizations measure calls, emails, meetings, and account coverage. These activities are visible, easy to count, and available long before business outcomes become clear. They can be useful indicators of whether something is working.
But it's easy to start treating the things we can count as though they're the things that matter most.
An Account Executive who sends a hundred emails appears more productive than one who sends twenty. Yet those numbers tell us little about which customers received meaningful help.
AI could make that problem worse. If we use it primarily to increase activity, we may become extraordinarily efficient at doing things without becoming much better at accomplishing what matters.
I think there's a more interesting opportunity.
A different Monday morning
Imagine our Account Executive begins her week with an AI research partner that helps her understand what's happening across her customer portfolio.
It identifies three situations worth considering.
One customer has completed extensive training but isn't progressing toward the business objective that motivated the investment. Another has made significant progress and might be ready for a more ambitious goal. A third has shown an unusual decline in engagement.
Rather than simply generating tasks, the system provides context, explains what it knows, and identifies where its conclusions are uncertain.
Our Account Executive recognizes the first problem from experience. She arranges a conversation to understand what's preventing the customer from applying what they've learned.
With the second, she sees an opportunity to help the customer take the next step.
The third recommendation doesn't convince her. She knows the customer is undergoing a planned transition, and reduced engagement is expected. She dismisses it.
By the time her first meeting begins, she's made three informed decisions about where her attention might be useful, including one decision not to intervene.
The AI hasn't simply accelerated her research. It's helped her recognize opportunities to apply her expertise. Her judgment still matters: the system can identify a customer need without knowing whether her involvement will help. It can also be wrong.
This isn't about giving professionals instructions they must follow. It's about giving them an extraordinarily capable research partner they can question and challenge.
That changes the question from how many customers she can contact to where her involvement has the greatest chance of making a meaningful difference.
This is a hypothetical illustration, not a description of a specific professional implementation.
What might that mean for customer value?
Pieces of this vision are already becoming practical.
In a Gainsight customer case, PartsSource describes bringing fragmented customer information together and using AI-assisted sentiment monitoring to surface emerging issues and prompt follow-up. Gainsight reports higher gross revenue retention alongside a broader transformation of its customer management practices. The case illustrates what's possible; it doesn't isolate AI's contribution to the improvement.
There's also a plausible commercial reason to pursue this. McKinsey's 2025 research found that B2B technology companies with sophisticated customer value-realization and adoption practices had net revenue retention roughly seven percentage points higher than peers with basic practices.
That's an association, not evidence that individual AI-assisted interventions caused better retention.
Still, it raises a worthwhile question: If we help professionals recognize the moments when their expertise can improve customer outcomes, might stronger commercial relationships follow?
I suspect they could. But I'd want us to test that relationship rather than assume it.
What kind of work are we trying to create?
There's a human ambition behind this that matters just as much to me.
I want Account Executives who delight in helping customers. People who enjoy untangling difficult problems, building trust, and helping others accomplish things that matter.
AI should make those strengths more valuable, not bury them under a larger volume of activity. It could also help people develop new capabilities and approach unfamiliar problems with greater confidence.
None of that is guaranteed simply by introducing technology. It depends on what we design the technology to do and what we choose to reward.
Productivity still matters. Saving time can create room for more valuable work. But I don't think the ultimate goal should be maximizing how much activity we can extract from someone's working day.
What if we evaluated AI by how much more value people can create through their expertise, rather than how much more activity they can generate?
That's a much more interesting future of work to me.