AI Salesperson
Aaron Ross, author of "Predictable Revenue" and creator of the outbound sales model Salesforce popularized, defined something almost two decades ago that's still central today: predictable revenue requires separating whoever prospects from whoever closes. Generative AI is reigniting that discussion, only now with a new question — how much of that prospecting can a machine actually do?
The honest answer, looking at what already works in the United States: a good chunk of initial qualification and first-touch outreach already runs well on AI. Tools can read a lead's profile, cross-reference it with company history, and write a personalized first approach — work that used to eat hours of a junior SDR's time.
What still doesn't work, despite what a lot of startup marketing promises, is complex negotiation. Closing a high-ticket B2B deal involves reading hesitation in a voice, adapting the pitch to an objection that wasn't in the script, and building trust across several conversations — exactly the kind of relational work that Ross's methods, and other names like Jordan Belfort (author of "The Way of the Wolf," on sales persuasion), have always treated as the core of selling, not a replaceable detail.
The common mistake is treating "AI salesperson" as a single category. In practice there are two different things: AI that qualifies and writes the first contact (works, saves real time) and AI that closes deals on its own (still a promise, especially in complex sales). Separating these two avoids both silly skepticism and empty hype.