Walking Away as a Skill, and the Pilot-Project Trap

The lost deal is not the costliest thing in this profession. The costliest is the deal that neither dies nor closes: it keeps you in meetings every three weeks, appears in your forecast every quarter, and consumes enough time for three real deals.

Whoever exits a deal like that loses nothing. He recovers an asset.

The number that makes walking away commercial rather than emotional

Return to the figure that opened this path: in The JOLT Effect by Matthew Dixon and Ted McKenna (2022, more than 2.5 million recorded calls), 40% to 60% of qualified deals end in “no decision”.

The practical meaning: if you are working ten deals, four to six of them will not be bought from anybody. Whoever does not know which ones spreads his time evenly across all ten, and gives dead deals half his week.

Walking away is not an admission of failure. It is the reallocation of a scarce asset.

Five signals that are enough to leave

The five are my own judgement from fifteen years in the Egyptian and Emirati markets, and I state plainly that this is collected experience rather than published statistics. One of them is not enough; two together are.

1. No event. Station one covered this: someone who cannot say what changed this year, nor what happens if he does nothing, is not yet a buyer. He is interested, and interest does not sign.

2. No access to the signer after two meetings. Not that you did not ask — that you asked and it did not happen. A man who cannot get you in front of whoever signs either lacks the influence he implies, or knows the project will not pass and does not want to say so.

3. The competitor wrote the criteria. When an RFP reaches you containing a requirement only one product satisfies, you have been invited to be the third bid that makes the tender look honest. We saw this in the two-markets station: whoever arrives after the RFP is written is competing on somebody else’s ground.

4. Price is the first and last question. Someone who did not ask about implementation, support or data, and asked about the number in minute three, is building a comparison file or leaning on his current supplier. That is a legitimate use of your time from his point of view, and not a project from yours.

5. The previous failure has not been dissected. A company where a systems project collapsed two years ago and which refuses to discuss why will repeat it — because the cause was probably managerial rather than technical, and it is still in place. Why ERP projects fail explains it: four of seven causes are management decisions.

How to leave without burning anything

A bad exit severs the relationship. A good exit leaves the door honourably ajar, because the event may occur seven months later and you want yours to be the number reached for then.

The formulation I use:

“From what I’ve seen, the current setup works for you, and changing now will cost you more than it saves. I don’t think this is the moment. If [the specific event] happens, call me the same day.”

Three things happen at once: the drain ends; your credibility rises further than ten meetings would have raised it — because nobody has ever told him don’t buy; and you have planted a specific trigger in his mind attached to you.

Deals have come back to me on that sentence more than a year later, uncontested, because the man remembered the one seller who sent him away.

The other trap: “let’s start with a small pilot”

The sentence sounds like a win. A hesitant customer offers to start with one module or one branch, and the seller accepts gratefully.

Most of these never grow. The reason is structural: an ERP does not show its value in one isolated unit. Its entire value is in the linkage — that a purchase order becomes a journal entry and a stock movement in the same instant. Install the procurement cycle alone and leave finance and inventory outside it, and you have built the customer an island requiring double entry, and demonstrated to him in practice that the system increased his work.

Three months later the pilot is measured and judged to have achieved nothing — a correct judgement on a flawed design that you agreed to.

When a pilot is right

It is right when it is cut vertically, not horizontally.

• Wrong: “we’ll implement procurement only, across the whole company.”
• Right: “we’ll implement the full cycle — from request to journal entry — on one branch or one product line.”

The vertical cut shows real value because the loop closes. The horizontal cut shows burden without value.

Three conditions go with it: a fixed decision date after the trial, a written numeric success criterion the customer signs before it starts, and a known price for the full phase from now — otherwise the pilot becomes a second negotiation starting from a weaker position.

The story: six months against one sentence

The story is composited from real events, and every identifying detail has been changed.

A trading group, three companies under one umbrella. The meetings were cordial and regular, and my contact was an intelligent operations director who genuinely enjoyed the discussion. I visited him eleven times in six months. I revised the proposal three times. I built detailed scenarios.

Every time the answer was: very good, I’ll put it to the boss.

I never met “the boss” once.

What ended it was not a fourth proposal. It was that I finally asked the question I should have asked in the second meeting: “Does the boss think there is a problem at all?”

And he answered with an honesty I owe him:

“Honestly… no. I’m the one who sees it. He thinks the work is running fine.”

Six months went to a deal that ended in its first sentence, and I never asked. The director had not deceived me; he was trying to sell his idea internally and was using me — in good faith — as material for that internal sale, a legitimate role whose price fell entirely on me.

The lesson: a warm relationship is not evidence of progress. And a deal that has not reached the signer after two meetings is not a slow deal; it is an entirely different deal from the one you think you have.

Checklist before you move on

• How many deals in your pipeline are older than six months with no new step?
• In how many have you still not met the signer?
• Do you have written exit criteria, or do you decide by feel each time?
• For every pilot on the table: is the cut vertical or horizontal?
• Does every pilot have a decision date, a numeric success criterion, and a known price for the full phase?

Next station

The deals worth having remain. Ahead of them is the meeting most sellers treat as a chance to perform, and which is in fact a test of discipline: the demo that sells.