Does B2B sales still need salespeople when customers research with AI?

Yes — but in a different role. The salesperson is no longer the source of information, because the customer has already obtained that from AI beforehand. The salesperson becomes a validation authority: confirming, correcting and completing the picture the customer brings with them.

This isn't a hypothesis — it's measurable. According to Gartner, 69 per cent of B2B buyers deliberately use salespeople to verify AI insights at decision points. Conversations with a salesperson make buyers 32 percentage points more confident in their decision. For the feeling that 'this provider understands my needs', it's 39 percentage points.

The reason behind this is interesting. Buyers don't blindly trust either AI or the salesperson: 51 per cent expect misleading information from generative AI, 49 per cent expect misleading information from salespeople. A stalemate. The consequence: buyers use both sources to check each other. The AI checks the salesperson, the salesperson checks the AI. TrustRadius confirms the pattern too — 94 per cent of AI users in B2B purchasing verify AI results, 72 per cent do so very often or always.

In practice, this means three things.

Discovery now means something different. Classic needs analysis asks what the customer needs. The new discovery additionally asks: what picture has AI painted for the customer — and where is it wrong or incomplete? Whoever finds and closes the gaps in the customer's AI knowledge creates the value no machine can deliver.

When a customer discloses that they've researched with AI, that's a gift. It shows trust, and it shows the starting point of the conversation. Salespeople who respond to this with confidence rather than defensiveness earn that openness.

And: every AI-informed deal is a competitive deal. If the customer has compared providers via AI, competitors are sitting invisibly at the table — from minute one.

So field sales isn't dying. But field sales that only delivers information is dying. What remains and becomes more important: judgement, industry understanding, the ability to check and complete an AI-generated picture. That's a question of roles, skills and leadership — not software.