AI in B2B sales: key takeaways
Two salespeople use the same AI tool. They get the same brief. An hour later, each sends out a flawless proposal: well structured, well written, not a single mistake.
On paper, you can’t tell them apart.
And yet, in the meeting, one of them truly grasps what is at stake for the client, explains their choices and handles the unexpected objection. The other one recites. The document was the same. The skill was not.
This reflection came to us from outside our own field. Miguël Dhyne, a lecturer and researcher specialising in pedagogical and digital innovation, recently shared on LinkedIn an editorial from the International Journal of STEM Education on assessment in the age of generative AI.
The starting point is simple. Good assessment aligns three things: what we want to measure, the evidence we collect and the conclusions we draw from it. With AI, that triangle becomes fragile. Correct work no longer necessarily proves that the student has understood.
The answer proposed is neither to ban AI nor to try to detect it at all costs. It is to shift the focus: assess the final product less, and the reasoning that leads to it more.
Now apply this to a B2B sales team. The parallel is striking.
Today, AI helps write a prospecting email, a sales proposal, a pitch or an account plan. That is great news for productivity. But it has a direct consequence: a polished deliverable is no longer enough to make the difference.
If everyone can produce a great proposal, what will the client base their choice on? On what the tool cannot do for the salesperson: real knowledge of the client, the way they lead a demanding conversation, their ability to respond with good judgement and creativity when something unexpected happens.
With the same starting instruction, the quality of the result depends heavily on the thinking skills of the person using the tool. Asking the right questions upfront. Checking what the AI suggests. Adapting a generic text to a specific client’s context. Defending a pricing or scope decision in front of a buyer.
None of this shows in the final document. And that is exactly where a salesperson’s value lies.
The ideas put forward for schools translate quite naturally to business:
Make the reasoning visible. Ask salespeople to explain the logic behind their proposal, the steps of their preparation and the assumptions they made about the client.
Use spoken formats. Role plays, objection-handling simulations, meeting debriefs. A live conversation remains one of the best ways to reveal real skill.
Work on authentic situations. Real accounts, real stakes, rather than textbook cases.
Critique an AI output. Hand over an email or proposal generated by the tool and ask: what doesn’t work for this particular client? What should be corrected?
In practice, this means combining several types of exercises: some without AI, to check the fundamentals (listening, questioning, rephrasing); some with AI, to see whether people use it with discernment; and others about AI, where the task is to analyse, correct or improve what it has produced.
The editorial ends with a question: what evidence do we want before we can say that a student has learned?
For a sales leadership team, it becomes: which situations give us reliable evidence that a salesperson can really sell? Not just produce a good document. Understand, adapt, co-create with the client.
At Valorize Solutions, we support B2B sales teams in their development, including in how they use their AI tools. Our conviction is simple: the tool speeds things up, but it is the salesperson’s reasoning that creates value for the client. So that is what needs training, and that is what needs watching.
Elevate the sales conversation. AI can help prepare a meeting, even support it. But the conversation itself happens between people. That is the whole point of our Elevate Conversations module: taking the exchange with the client to another level.
Use AI with discernment. With our AI for Sales module, we help teams use their AI tools in service of their sales practice. We explored this idea in our article on the real question behind AI for Sales.
Look at commercial excellence as a whole. People, methods, processes and tools work together. And AI can gradually listen to sales interactions to fuel continuous improvement in skills and team dynamics. Work that then continues with a coach. Discover our approach to commercial excellence.
No. The point is neither to ban AI nor to detect it, but to check that the salesperson understands, adapts and can defend what they propose. Used well, AI remains a useful accelerator for preparation.
By looking at the reasoning rather than the document alone: explaining the choices behind a proposal, role plays and objection-handling simulations, work on real accounts, critiquing a text generated by AI.
It can help prepare meetings and, gradually, listen to sales interactions to fuel continuous improvement in skills and team dynamics. That work then continues with a coach.
From an editorial published in September 2026 in the International Journal of STEM Education, “The shifting landscape of assessment in STEM education in the age of generative AI”, shared on LinkedIn by Miguël Dhyne.