Categories
We’re Still Selling to People, Not Agents
As AI becomes embedded across marketing, sales and customer experience, we’re getting much better at answering one question: What can AI do?
It can analyze thousands of customer interactions. It can research an account before a sales call. It can draft messaging, identify competitive patterns, recommend the next best action and increasingly carry on an entire conversation with a customer.
The more interesting question may be a different one: Where do we rely solely on AI vs. AI and humans?
Because for all the discussion about AI transforming go-to-market, one fairly basic fact hasn’t changed. We’re still selling to people, not agents.
When the agent gets it exactly right
There are plenty of customer interactions where AI may actually provide a better experience.
Consider a customer who needs to make a straightforward change to an account. An AI agent recognizes the customer, understands the request, makes the change and confirms it within seconds. No hold music, transfers between departments or repeating an account number three times.
The same principle applies inside a B2B technology company. A salesperson preparing for an enterprise meeting can use AI to pull together recent company announcements, previous interactions, likely competitors and relevant customer stories. A product marketer can analyze dozens of competitor websites, calls, customer reviews and internal win/loss notes and identify patterns that would have taken days to uncover manually.
These are good applications of AI. The technology can process enormous amounts of information quickly and make that information useful to the person doing the work.
When agents can get it wrong
Now consider a customer contacting a telecom provider because of an unexpected charge. The AI agent explains the charge correctly, but the customer has already contacted support twice and was previously told it would be removed. They explain this, and the agent responds with another explanation of the billing policy.
The answer may be technically correct, but by this point the customer is trying to resolve a broken commitment and the frustration that comes with it. Recognizing that context requires something beyond retrieving the right answer.
That distinction matters in B2B technology just as much as it does in customer service.
Consider a product marketer conducting customer interviews. Across 30 conversations, AI identifies implementation as a recurring topic, summarizes how frequently it appeared and pulls representative comments. That’s enormously valuable.
But during one interview, a customer says, “Implementation isn’t really a concern for us,” and then spends the next ten minutes asking who needs to be involved, how long deployment takes, what resources they’ll need internally and what happens if adoption is slower than expected.
A strong human product marketer notices the disconnect and asks why.
Maybe the customer has been through a difficult implementation before. Maybe they championed it internally and it failed. Maybe the concern isn’t implementation at all, but what happens to their credibility if this one goes badly too.
AI can help us understand what customers are saying at a scale that wasn’t previously possible. There is still enormous value in a person understanding what a customer means.
B2B buying is still remarkably human
We tend to describe enterprise buying in fairly clinical terms: buying committees, personas, ICPs, intent signals, pipeline stages and economic buyers.
Behind those terms are people.
The CIO is evaluating technical risk. The CFO wants to understand the financial return. The business leader may be wondering whether their team will adopt another new system. And the person championing the purchase may have a question that will never appear in the CRM: If I recommend this and it fails, what does that mean for me?
AI can help a go-to-market team understand those stakeholders better than ever before. It can research their priorities, summarize previous interactions, surface relevant content and prepare a salesperson before the conversation begins.
What happens during the conversation still matters.
Someone needs to notice when the CFO keeps coming back to the same concern, when the champion hesitates, or when a customer says they’re satisfied but their behavior suggests otherwise. B2B buying may be rational and highly considered, but that doesn’t make it emotionless.
The opportunity for AI
Perhaps the best application of AI in go-to-market isn’t to remove people from more interactions. It’s to make people substantially better prepared for the interactions that matter.
Use AI to synthesize 30 customer interviews, then give the product marketer more time to have the 31st. Use it to prepare the salesperson before the meeting so they can spend the meeting listening. Use it to surface years of customer history before someone picks up the phone.
That balance will become more important as AI gets better.
There will be plenty of moments where an agent is faster, easier and frankly better than dealing with another person. There will also be moments where the value comes from curiosity, judgment, empathy and recognizing something that wasn’t explicitly said.
We’re still selling to people, not agents. The companies that remember that while adopting AI may end up understanding their customers better than they ever have before.
