The Shift: What 25 Revenue Leaders Said Has to Change About AI in B2B Sales
Three weeks ago we named the lies. Last week we landed the invoice. This week the conversation moved somewhere harder: if we know what’s wrong, what actually has to change?
Twenty-five B2B SaaS revenue leaders gathered in London for a Cuvama breakfast debate to examine exactly that. The debate was candid in a way that vendor-sponsored research rarely is. Nobody was selling a solution. And the conclusions weren’t tidy. What came out of the room were three shifts – each distinct, each uncomfortable, and none of them fully solved by anyone present.
Here’s what they said.
Design the Sales Motion for AI, or Watch it Work Against You
The conversation kept coming back to one word: structured.
Not “use AI more.” Not “use AI less.” The unanimous conclusion was that the teams suffering most from AI aren’t the ones who adopted it slowly – they’re the ones who handed it to sellers without redesigning anything around it.
When AI operates without guardrails, sellers default to the path of least resistance. Activity goes up. Thinking goes down. And buyers who needed someone to build a case for change got a well-researched summary instead. That’s where no decisions come from.
The shift is designing the sales motion so that AI sharpens judgment rather than replacing it – and then training sellers on exactly how that works in practice. The key words, as one group put it, are structured and augmentation. Both matter. Neither alone is enough.
Stop Writing Business Cases. Start Building Shared Value Plans.
The second shift sounds like a terminology change. It isn’t.
A business case is a point-in-time document. It arrives near the end of the deal, it reads as a closing tactic, and buyers know exactly what it’s for. One CRO in the room described the moment perfectly: “Why are you suddenly talking to me about money? You’ve been talking about technology all along.”
That suspicion – earned or not – kills collaboration on the very document that’s supposed to build it.
The shift is from business case to shared value plan. Not a rename. A structural change in when the value conversation starts and who owns it. A shared value plan is built with the buyer, begins after qualification but well before the final stages, and frames outcomes the buyer can take internally to justify the decision themselves.
As one attendee put it: “We’re not building a business case for you. We’re giving you the content, the data, and the process to take internally for your own justification.”
That reframe changes everything – the buyer’s posture, the seller’s role, and the deal size at close. The group cited smaller deals as the most consistent commercial cost of getting the timing wrong. Attendees also named what happens post-sale: teams that sell on features reach the three-year renewal at “red button, blue button” level, and the C-suite that signed the cheque has no idea what value was delivered.
Mathieu Charles, VP of Value Engineering at Temenos, framed the ambition clearly during the panel. When the moderator asked each panellist to complete the sentence “By 2027, the best B2B sales teams will use AI in discovery primarily to…” his answer was: form a provocative point of view. Not to summarise. Not to personalise. To walk in with something worth saying – before the conversation about money has even started.
Nobody in the room had fully cracked how to execute this at scale. That honesty matters. This is emerging practice, not a playbook.
Measure What Moves Deals, Not What Fills Systems
The third shift sounds operational. Its consequences are strategic: management is measuring the wrong things.
AI has made it easier than ever to look busy. Fields get filled. Account plans get generated. Activity metrics go up. And from a system perspective, the lazy seller and the excellent one are increasingly hard to tell apart — because both have full CRMs and well-written summaries.
The shift is from activity metrics to outcome metrics. Not “did the seller do the prep” but “did the conversation move the deal forward.” Not “are the fields complete” but “do we have the champion, the budget event, the economic buyer engaged.” One attendee described using Gong not just to capture calls but to play back individual sentences and ask: what did this buyer actually mean? That’s analysis, not activity – and it’s where the real signal lives.
Training is where this starts. The group was clear that training has to shift from product knowledge to judgment development: how to prompt AI correctly, how to read what AI surfaces, how to verify outputs before they reach a buyer. The differentiator between teams that improve with AI and teams that stagnate, one attendee noted, comes down entirely to how well they’ve been trained to use it — not whether they have access to it.
The room reached a shared conclusion on this: the question isn’t whether your team is using AI. It’s whether the person managing them can tell the difference between a seller who is using it well and one who is using it to avoid the work entirely.
Three Shifts. None of Them at the Finish Line.
The teams moving earliest on structured augmentation are already seeing fewer no-decisions. The teams starting the value conversation before the final stages are already seeing it in deal size and close rate. The teams shifting management focus from activity to outcomes are starting to surface the real picture in their pipeline – which is the first step to fixing it.
None of this is comfortable. Redesigning a sales motion takes longer than deploying a tool. Shifting how managers measure performance means admitting that current dashboards are telling a partial story. And building shared value plans from early in the cycle requires sellers to have a point of view worth sharing before a buyer has given them much to work with.
But the alternative – continuing to optimise for activity while judgment atrophies – is a price we’ve already named.
Is your sales motion designed for AI, or just exposed to it?
The full event report here, including what the room said the lies and prices are of AI in B2B Sales.
