The gap is what you are actually selling
The phrase comes from Keenan's book of the same name, but the idea is older than the label and every good rep half knows it already. A buying decision has three parts: where the buyer is now, where they want to be, and the cost of the distance between the two. Nobody buys the current state, because they already own it. Nobody buys the future state either, because you cannot hand it over. What gets bought is the removal of the distance.
This is why product knowledge alone makes a mediocre seller. A rep can describe every feature correctly and still lose, because features attach to a gap that was never measured. It is also why the same demo lands beautifully in one call and dies in the next. The demo did not change. The diagnosis behind it did.
The pattern is the same well outside B2B software. An acquisitions rep calling a property owner is doing exactly this work. The current state is not "owns a house". It might be "owns a house four states away with a tenant who stopped paying in the spring". The desired state is rarely "cash". It is more often "out of this before the winter, without a conversation about the roof". A rep who leads with a number is negotiating against a situation they never measured, and the number will be wrong in both directions at once.
Why the pitch arrives early anyway
Reps do not skip diagnosis because they never learned it. Almost everyone has heard the advice. They skip it because five separate forces push in the other direction, and none of them feel like a mistake in the moment.
- The pitch is the rehearsed part. It is the only section of the conversation a rep has said out loud more than twice. Under pressure, people fall back to the material they have practised, and for most reps that material is the presentation.
- Silence reads as losing control. A question hands the floor to someone else. A statement keeps it. New reps in particular experience discovery as a slow loss of ground, so they take the floor back by talking about the product.
- The buyer invites it. "So what do you guys do?" arrives in the first minute of most calls. It sounds like permission. It is usually just politeness, and answering it for four minutes is the most common way a discovery call quietly ends.
- Product knowledge feels like safety. A rep who is unsure of themselves reaches for the thing they are certain about. Being right about the product is comforting in a way that being curious about a stranger's problem is not.
- Activity metrics reward it. Nobody counts diagnoses. Plenty of teams count demos, proposals sent, and offers made. A rep optimising for the scoreboard on the wall will present early, and they are not being lazy. They are being rational.
What one more question actually finds
The gap between a pitched call and a diagnosed one is usually a single follow-up. Not a new topic, and not a longer list of questions. One layer down on the answer the buyer just gave.
| What they said | What the pitch-first rep hears | What one more question finds |
|---|---|---|
| "Our reps are not hitting quota." | A training need. Start the demo. | Two people carry the number and the rest were hired in the spring |
| "We are looking at a few options." | Competition. List the differentiators. | They have to present a single recommendation to a committee, and they have not picked the criteria yet |
| "I might sell the property." | An offer conversation. Give a number. | It was inherited, two siblings disagree, and one of them wants it kept |
| "Your price is higher than we expected." | A discount request. Ask for approval. | Nobody in the room has ever priced what the current process costs them |
Notice that none of the right-hand answers are things the buyer was hiding. They were one question away the whole time. That is the uncomfortable part of the gap selling method: the information that would have changed the call was available, and the rep traded it for airtime.
Roleplay that builds the patience
Telling a rep to "do better discovery" changes nothing, because the failure is not a knowledge failure. It is a nerve failure, and nerve is trained by repetition under mild pressure. These four roleplay drills are narrow enough to run in a few minutes.
- The one-more-question rule. Whatever the buyer answers, ask one follow-up on that same answer before you move on. Not a new subject. The rule is mechanical on purpose, so it survives adrenaline.
- Name the urge out loud. In the debrief, ask the rep to point at the exact moment they wanted to start presenting. Reps who can name the urge afterwards start catching it live within a handful of sessions.
- Count a beat. Two seconds after the buyer stops talking. Most of the sentence that matters arrives inside that pause, and reps who fill it never hear it.
- Roleplay against bait. Run a scenario where the AI customer demands the pitch in the first minute. The skill is answering in two sentences and returning to the diagnosis without sounding evasive. That specific move is worth more practice than any other single thing in a discovery call.
Before you present, state the gap in one sentence the buyer would agree with word for word. If you cannot, you are about to present to yourself.
When diagnosis is the wrong move
Gap selling gets taught as an absolute, and that is where teams overcorrect into calls that feel like interrogations. There are real cases where more questions cost you the deal.
- The buyer already diagnosed themselves. They arrive with a spec, a shortlist, and a date. Re-running discovery on them signals that you were not listening, and the honest move is to confirm the gap in one question and get to the point.
- The deal is smaller than the discovery. On low-ticket, high-volume sales, ten minutes of diagnosis costs more than the margin. Two questions, then price.
- You are disqualifying. When the mismatch is already visible, naming it fast is a service to both sides. A short call that ends cleanly is a good call.
- It is late in the cycle. Deep diagnosis on the fourth conversation reads as starting over. If the gap is still unclear at that point, that is a problem to fix in the notes, not on the call.
What a manager can watch
Diagnosis is easier to observe than most sales behaviours, because it leaves a timestamp. Three things are visible without a rubric of fifteen criteria. First, the minute of the first product claim. Second, the ratio of questions to statements in the opening third of the call. Third, and most telling, whether the rep can state the buyer's gap from memory afterwards without opening their notes.
That last one separates reps who asked questions from reps who listened to the answers. Plenty of calls contain a dozen questions and no diagnosis, because the rep was collecting fields rather than building a picture. If the rep can say the gap back in one sentence, the discovery worked. If they produce a list of facts instead, it did not.
This is the kind of thing roleplay fixes faster than feedback does. Inside pichi.ai a rep runs a voice role play with an AI customer who does not volunteer the important detail until someone digs for it, and the debrief afterwards shows where the pitch started. Doing that against a simulated buyer costs nothing when it goes badly, which is the entire point of doing it somewhere other than a live pipeline.