All lessonsBeliefs & Identity

Why Two Founders Agree On The Goal And Still Fight About Every Step

Two people agree on the target and fight about every step, because the decision is being made by a criterion neither of them has said out loud. Here is how criteria work, what the values research supports, and a six-step protocol for putting both rulers on one page.

2026-09-29 2,695 words 14 min read

The Argument That Was Never About the Hours

Between 2008 and 2013 my career climbed while my marriage came apart, and the two kept pace with each other. My work collected accolades. Going home collected strain.

She would ask why I was working so hard. I heard an accusation, and I answered it the way a man answers an accusation. From where I stood I was building something that fed people, paid salaries and carried a message that changed lives, and the long hours were the price of that. From where she stood the long hours were the thing that removed me from the house.

The argument ran for five years without moving an inch. We were not disputing the facts. Same week, same hours, same man. We were running different tests for what counts as a life going well. My test put contribution at the top of the list. Hers put presence there.

One evening she stopped mid-sentence and went quiet, and the quiet landed harder than the shouting had. I sat in it and understood nothing at the time. Years later, reading back through what happened, I saw the mechanism. Each of us had handed ourselves the context for the marriage and assumed the other one was holding the same page. Experts on relationships make this point often, and Esther Perel puts it in the plainest form: you write your own context, and you rarely hand it over.

That is the pattern this lesson is about. Two people can share a goal down to the decimal point and still fight about every step toward it, because the decision is being made further upstream, by a test neither of them has said out loud.

What a Criterion Is, Stated Plainly

A criterion is the test you run to decide whether something counts. It answers questions like these. Does this count as a good outcome? Does this count as fair? Does this count as worth the money? Does this count as me staying true to who I am?

You run those tests constantly, and out loud you rarely name them. When a founder says a deal feels wrong, a criterion has fired. When a manager calls a quarter a disaster on the same numbers the board calls a solid result, two criteria have fired at once on identical data.

A criterion sits one level above a goal, in the way a ruler sits above a measurement. The goal says where you intend to arrive. The criterion decides which routes count as arriving properly, and how much the trip is allowed to cost.

Here is the shape of it in ordinary business. Two directors both want the company past twelve million in revenue. One measures the year by margin. The other measures it by the strength of the team that will carry the following year. Same target, different ruler, and every budget line turns into a fight.

This is why the phrase priority problem misleads. A priority is a ranking of things the team already agrees to measure. When the measuring itself differs, the ranking was never going to hold.

Where Criteria Sit in the NLP Model

Robert Dilts set out the level structure most practitioners in this field work from: environment, behaviour, capability, beliefs and values, identity, and purpose. Criteria live in the beliefs and values band, and that placement explains their force. A criterion audits everything beneath it, quietly, before you get a say in the matter.

Dilts developed this material in Changing Belief Systems with NLP in 1990, and again with Tim Hallbom and Suzi Smith in Beliefs: Pathways to Health and Well-Being the same year. Richard Bandler and John Grinder had already placed values work at the centre of the methodology in The Structure of Magic in 1975 and in Frogs into Princes in 1979.

An honesty note belongs here. Criterion elicitation comes from clinical observation and from modelling skilled communicators. It carries no controlled trial behind it, and I will not dress it as experimental psychology. What the values research supports is the underlying claim, that people rank the same values differently and that the ranking shapes behaviour. The technique for hearing the ranking is modelled craft, and it earns its place through repetition in real rooms.

Two People Who Agree Can Still Fight

Milton Rokeach published The Nature of Human Values in 1973 and argued that values form a hierarchy inside each person, and that behaviour follows the ranking rather than the list. Two people can both hold honesty and kindness, and still make opposite choices in the same room, because one of them keeps honesty higher up the stack.

Shalom Schwartz tested the structure across more than twenty countries, publishing the findings in Advances in Experimental Social Psychology in 1992. He found a stable set of value types arranged in a circle, where neighbouring values sit comfortably together and opposite values pull against each other. The word that matters most in Schwartz's work for this conversation is priority. People hold similar values and order them differently, and the order travels with them into the office.

Karen Jehn's 1995 study in Administrative Science Quarterly examined conflict inside working groups and separated task conflict from relationship conflict, with process conflict added in later work. Her distinction is useful here. Disagreement about how work happens behaves differently from dislike between people, and a criteria clash arrives dressed as the second while measuring the first.

Two qualifications, stated plainly. The values research measures individuals rather than boardrooms, so transfer from survey data to a leadership team is an inference. Jehn studied work groups in real organisations, which sits closer to the target, and her design did not test NLP criteria elicitation as a tool. The mechanism travels; the protocol remains modelled.

The Five Criteria Founders Run Without Noticing

Across a long run of these conversations, five criteria surface more than the rest in business settings.

  • Velocity. Speed carries the verdict. A good result delivered slowly reads as a bad result.
  • Certainty. What counts is how much is known before committing. Ambiguity itself registers as failure.
  • Fairness. The effort and reward split carries the judgement. A brilliant outcome can feel rotten when the split looks crooked.
  • Contribution. The question is whether the work fed something larger. Money becomes a scoreboard rather than the point.
  • Recognition. Being seen carries the judgement. A win nobody acknowledged lands as a loss.

These five rarely turn up alone. The trouble starts when two people inside one decision are holding different ones at the top of the stack, and neither can see the other's arrangement. Each then reads the other's position as stubbornness, laziness or politics.

How to Hear a Criterion in a Live Conversation

The question that opens the door is short and uncomfortable in effect: what has to be true for you to call this a win?

People answer with a tactic on the first pass. "I want the launch out in March." That is a date, and a date is not a criterion. Move one level up with a means-end question. What does March get you? The answer arrives. The ability to sign the enterprise client in the second quarter.

Ask again. What does that client give you? Now the value word surfaces. Stability. Credibility. Proof to myself that the last two years were not wasted.

This laddering move has a formal version in consumer research. Jonathan Gutman described the means-end chain in the Journal of Marketing in 1982, running from attributes to consequences to values, and he built it to surface what a person is genuinely pursuing. In a boardroom the same ladder works on internal decisions.

Two questions carry the whole conversation. First: what has to be true for you to call this a win? Second: what does that give you? Ask the second question twice and a tactic usually collapses into a value. Three rounds is often enough, and by the fifth answer people start repeating themselves.

Listen for the point where the person's voice slows down. That is where the value sits, and it is usually the moment they stop arguing about the calendar and start telling you what the calendar was protecting.

The Meeting Where Two Rulers Collide

A founder and her operations lead sat on a decision that needed ten minutes and consumed six weeks. A product line was losing money. He wanted it cut that week. She refused to sign off until the two clients running it had heard the news in person, with a plan for the team attached.

The argument travelled through six meetings with nobody moving, and the language on both sides drifted toward judgement and respect. Neither of them could see the other's ruler, so each assumed the other was being difficult on purpose.

One question broke it open. I asked each of them what the year would have to look like for them to call it a success once this call was behind them. He answered first: no drag on cash past the end of the quarter. She answered second: nobody on that team leaves believing they were expendable.

Both criteria were sound. Cash discipline keeps a company breathing. Reputation with people decides whether the next team signs up at all. The fight came from each criterion being invisible to the person who did not hold it.

They cut the line on the Friday and told both clients face to face on the Thursday. Two extra days bought both criteria. Nothing about the underlying disagreement was irrational. It had been unmapped, and unmapped disagreements grow teeth.