Ask AI to Argue the Buyer's Side Before Your Next Sales Call

The rehearsal you're doing is the wrong rehearsal

You know the ritual. Notes open, objections listed, closing line memorized. You rehearse your side of the call until it's smooth.

Here's the problem: smooth on your side changes nothing about theirs. You can deliver a flawless pitch to a mind that is closed and lose every time. The call was never yours to win with better words. It was theirs to open, or not, before you said anything.

That's the first of Atiba's Selling Principles: state precedes message. You cannot move a mind whose state is closed. Attention and openness are physiological before they're ever intellectual. A buyer who is defensive, distracted, or already decided doesn't hear your pitch better because you polished it. They don't hear it at all.

So the rehearsal that actually matters isn't your pitch. It's their state. And the fastest way to find it before you're in the room is to make AI fight you for it.

The rule underneath the rule

The Selling Principles aren't a script. The order is: lead with the principle, then the framework, then the tactic. The tactic is disposable. It's just today's expression of the principle. When one tactic isn't practical, a different tactic that meets the same human need works exactly as well.

That's the license to do something that looks unrelated: build the buyer who argues with you before the call, the same way you'd build anything else meant to expose you to a real mind instead of your own assumptions. Different domain, same principle. Watch what happens when you feed a machine enough real detail about someone else that it stops being your mirror.

Build the buyer who argues, not the buyer who agrees

Most people who "use AI to prep for a call" ask it to play the buyer, get an agreeable simulation, and walk in more confident than they should be. That's worse than no prep. An AI fed a thin sketch of your buyer produces exactly what thin sketches produce: agreeable slop. It nods. It never once contradicts you. You leave the rehearsal believing you've stress-tested your pitch when you've only flattered it.

Atiba ran into the other version of this by accident, writing his own book. Before he wrote a word, he built a coach inside ChatGPT and fed it a thirty-page profile of his ideal reader. Not a paragraph. Thirty pages. The coach got good enough that it stopped assisting and started existing as itself, took on its own personality, and later pushed back on his own prose, telling him flatly it didn't speak in the reader's language.

The tool didn't get better at agreeing with him. It got good enough to disagree with him, on behalf of someone who wasn't in the room, using detail a thin sketch never could have carried. That is the same tactic you need for a call, dressed in different clothes: not a buyer simulator that says yes, one built with enough real detail that it's willing to tell you no, and mean it.

State: the buyer's readiness to hear you, physiological before it's intellectual, closed and guarded or open and curious, and set before you speak a word.

The call Ariana opened by asking to leave

Ariana Demers emailed Atiba to pause her subscription. That was the literal ask. Answering it would have been the easy, correct-looking move: confirm the pause, end the call, move on.

Instead he walked her through her emotion first, then her current reality, then her actual constraint, before he answered the ask at all. Only then did he put seven pages of LinkedIn prospects his team had been generating for her, every month, in front of her.

She had never seen them. She believed the only way she could sell was from a stage, and she'd been quietly betting her business on a brand deal to get there. She was defending that bet by asking to pause, not stating it outright. Diagnosing her state instead of answering her ask surfaced the actual split she was already standing on: stage-selling versus everything already arriving that nobody was working. She'd cast her vote for the stage a long time ago. The call didn't sell her something new. It named the side she was already on and showed her the other one.

She was waiting on a stage deal for a fraction of the leads a month that were already landing in her inbox, untouched.

The call she opened by asking to pause ended with her asking to be helped more, and she came back Monday with a proposal to keep paying. That's what happens when you diagnose state before you answer the message, not a save, a diagnosis that got there first.

Answer the ask, or diagnose the state

What you're actually unlearning

Name the habit plainly, because naming it is the only thing that breaks it: you were trained, by every sales script you've ever read, to answer the ask. Someone says "I want to pause," you process a pause. Someone raises an objection, you rebut the objection. That reflex feels like competence. It is the thing this entire approach asks you to stop doing.

That's a real rewiring, not a checklist add-on. It costs you time up front: a thirty-page profile isn't a prompt, it's an afternoon of pulling out detail you don't usually write down, the buyer's history, their stated objections, the thing they said last time that you filed away instead of using. It costs your ego harder: you're asking a machine to poke holes in the pitch you built specifically to feel ready, and it will. You will read "no" from a simulator before you've walked in the door, and it will not feel good.

Don't do this for every call. It's overbuilt for a transactional sale where the buyer's state barely matters and the decision is near-instant. Save it for the calls where the reflex to just answer the ask is exactly what will lose you the deal: the renewal that's really a cancellation, the deal that's gone quiet, the buyer who's already decided something and hasn't told you what.

Monday

Pick the one call on your calendar where you already suspect the buyer's request isn't their real problem. Build the profile deep enough that a machine could argue with you about it. Instruct it to disagree, not represent. Run your opening on it before you run it on them.

If it agrees with everything you say, you haven't built a buyer. You've built a mirror that flatters you, and the reflex you were supposed to break is still running the call. Go deeper before you get on the phone.