LIFE & LOGIC
← Back to projects
Project

An organization can learn what its people know

Nexa Intelligence exists so that a leader can ask the people who do the work, and so that those people matter to the decisions that shape their work.

Andrea Doyon

Every organization knows less than its people do. A driver notices that a delivery window no longer fits how customers live. A nurse sees that one form slows every admission. A sales associate hears the same objection again and again. Each of them holds something useful, and most of it never reaches the people who decide.

Leaders have tools for this. Surveys reach many people and ask each of them a little. Interviews reach a few people, and each interview asks a great deal of them. Dashboards count what the systems already record. None of these answers a plain question: what do the thousands of people who do the work actually know?

Nexa Intelligence exists to answer that question. A leader should be able to ask them, and what they know should shape the decision. An organization that can do both makes better decisions and keeps the trust of the people who carry them out.

The engineers’ objection stopped at one meeting.

On the evening of January 27, 1986, engineers at the company that built the solid rocket boosters for the space shuttle joined a teleconference with NASA managers. The next morning’s launch of Challenger was forecast for about 26 degrees Fahrenheit, which is roughly minus 3 degrees Celsius. The engineers had data on the rubber O-ring seals in the booster joints. Their chart recommended no launch below 53 degrees Fahrenheit, the lowest seal temperature of any earlier flight. One of them, Roger Boisjoly, argued that cold slows the seal, and he said plainly that he could not put a number on the risk.

The managers on the call pushed back. One argued that the data did not change the case for the primary seal, because a secondary seal sat behind it. Another said he was appalled by the recommendation and added that he would not recommend a launch over the manufacturer’s objection. During a break, a senior manager at the manufacturer asked the head of engineering to think as a manager would. The company then returned with a recommendation to launch, and a signed copy was faxed that night. Challenger launched at 11:38 the next morning and broke apart soon after liftoff.

The Rogers Commission, which investigated the accident, asked who had known. The launch director at Kennedy Space Center testified that he had not known of the objections, and that he would not have launched had he known the 53 degree criterion applied. A senior official at NASA headquarters testified that he had not known either. The engineers had a chart, data and their names on both. They lost the argument that night, and the people above that meeting never heard it.

Nexa Intelligence exists for that launch director. He should be able to ask what the people closest to the seal knew, on any day before the night, and read their account with its data open beneath the answer. The engineers should be able to count on being heard by the person who decides, without needing a seat on the call. Challenger is the extreme case. The ordinary case is the nurse and the form, a thing known on the floor that never reaches a meeting. Nexa Intelligence serves both.

A great change succeeds when the people living it understand it, and it needs what they know.

Life & Logic holds that a great change succeeds when people understand the story they are living. Nexa Intelligence adds the other half. A change also needs the knowledge of the people living it, because they are the ones who will make it work or watch it fail. The nurse who enters the form twice knows where a new admission process will break before anyone has drawn it.

An organization cannot explain a change to people it has not heard. The explanation has to start from what they are living: which step costs them time, which rule they work around, which customer they are afraid of losing. A leader who knows these things can tell the story of a change in words that people recognize as theirs, and the people hear a leader who has been where they work. Without them, people hear a decision from far away. That is why listening to thousands of people belongs to the work of leading them.

A conversation about last Tuesday tells more than a question about an opinion.

The conversations are held by voice with an AI interviewer, in the language the person works in. The interviewer opens with a recent, concrete event: the last difficult customer, the last task that went wrong, the last time a rule was bent to get the job done. An event produces names, places and a sequence. Ask a person whether they are satisfied with a new tool and you receive a sentence they have said before. Ask what happened the last time they used it and you receive a story with the detail the question never thought to request.

Voice matters because most of the knowledge in an organization is spoken. A driver will say in two minutes what she would never type into a form, and a nurse who is tired at the end of a shift will still talk. A conversation in the person’s own language puts the effort on the interviewer. The person has only to remember what happened.

Imagine a sales associate asked about last Tuesday. She tells how a customer wanted delivery on Saturday, how the screen showed nothing available, and how she phoned a driver she knew and found a slot by hand. She never mentioned a problem with the booking system, and she described one precisely. The interviewer follows the event with questions that go deeper into it: what the customer said, what she tried first, what she would have wanted on the screen.

The interviewer stays neutral. It uses no words that suggest the answer it hopes to hear, and it reacts with neither approval nor surprise. The person is the expert on what happened, and the interviewer is there to get the account complete. Because the interviewer is software, the first person and the last person are asked the same questions in the same way, and no person in the organization has to be the one who asks.

A person’s account keeps her words and leaves out her name.

A leader who can read the account behind an answer raises a fair question. The nurse is then speaking to her employer, and she has every reason to say less than she knows. Her name is never asked for, and it does not enter the report. Because her account is anonymous, she can say that the form is a waste of an hour without wondering who will read it. A person who fears being identified tells less, so the anonymity has to be kept.

A minister or a chief executive who opens an account finds a person with a job and a way of speaking, and the account is easier to believe than any total. The labels tell the leader whom the account speaks for.

The conversation is then broken into statements. A statement is one claim by one person. It keeps the words the person used and a few labels: what it concerns, whose view it is, which group the person belongs to and whether the person lived the thing or heard of it. Every finding a leader reads links back to the statements beneath it. A leader can check each finding against its source and trust the answer enough to act on it.

One nurse’s Thursday becomes a manager’s answer.

Follow an invented example from start to finish. A nurse describes last Thursday. A patient arrived on the ward, and admission ran long because the form had to be entered twice, once on paper and once on screen. The interviewer asks what she did next, and she says she wrote the paper copy at the end of her shift. This nurse is a composite, and no real person said these words.

A statement is kept: the admission form is entered twice. It concerns admissions, it is held by a ward nurse, and she lived it. Other nurses on other wards describe the same step in their own events, and their statements have the same label. Together they form one finding, and the report says that it describes nurses and not patients.

A hospital manager asks what slows admissions on the wards. The organization has decided what this manager may read, and the answer is built from the statements inside that permission. It names the double entry as the most consistent cause, says which group describes it, and points to each statement it rests on. The manager opens one account and reads about Thursday, without learning which nurse it was.

A leader can ask a question across everyone and hear where people agree.

The easiest summary of a thousand conversations is a ranked list of what people talked about most. That list is true, and it hides two things. The first is who is speaking. Employees and customers see the same work from opposite sides, and a blended figure describes neither of them. Each finding is reported for its own group, and when two groups disagree the report shows both.

The second is the gap between what people discuss and what they agree on. Suppose a subject comes up in one conversation in ten, and everyone who raises it describes the same fix. By volume it ranks near the bottom of the list. By agreement it may be the strongest signal in the whole set. A leader who reads the two separately sees the one thing the people had already settled among themselves, and can act on it with their backing.

People who are heard begin to matter to the decisions that shape their work.

Picture the leader of an organization asking, before a decision, what the people who will live with it already know. The answer arrives with the accounts behind it. The driver, the nurse and the sales associate are in it, each in their own account and none of them named. The decision that follows has their knowledge in it from the start.

A leader who works this way changes what a meeting is for. The meeting no longer settles whose version of the facts to trust, because the facts are already on the table with their sources. The people on the call can spend their time on the decision. The engineers, the nurses and the drivers are in the room through their accounts.

The nurse who described Thursday then finds the form entered once. She also knows why it changed: someone asked, and the answer reached the person who could act on it. A person who sees that connection between speaking and a changed form has a reason to speak again.