The provider's discovery team arrived on a Monday morning. There were four of them. The team lead was a woman named Diane who had managed transitions of this kind eleven times in the previous four years and who carried the particular efficiency of someone for whom the unfamiliar has become familiar through repetition. With Diane were two analysts whose names the CEO did not immediately retain and a technical specialist named Kwame who was there to assess the integration between the provider's systems and the operational infrastructure Joel had built. They arrived at nine. By nine fifteen they were at a table in the meeting room with the whiteboard and the four chairs, which had been supplemented for the occasion by two additional chairs brought in from the kitchen area, and Sandra was explaining the structure of the support operation in the clear and methodical way that Sandra explained everything, starting with the overview and moving to the detail in a sequence that made the complexity navigable without making it seem simpler than it was. The CEO watched from the corridor for a few minutes without going in. Diane was taking notes on a laptop. The two analysts had notebooks open. Kwame had a tablet and was making occasional entries with a stylus. The note-taking was thorough and professional and captured, accurately, the words Sandra was saying. The CEO thought about the green notebooks. Not the green notebooks specifically. The thing the green notebooks represented. The difference between the words someone says when they are explaining something and the understanding that the words are trying to convey. Sandra was explaining the support operation accurately. The provider's team was recording the explanation accurately. At the end of the four weeks they would have an accurate record of what Sandra had said. What they would not have was what Sandra knew. On the provider side of the operation, in an office building twenty minutes from the company by car, a team of twelve agents was being assembled for the transition. The team had been recruited and trained according to the provider's standard processes, supplemented by the product knowledge materials that had been shared during the due diligence phase. Among the twelve was a team lead named Gil. The CEO would never meet Gil. The company would never know his name. He would not appear in any report or any quarterly business review or any of the formal channels through which the provider communicated with the company. He existed, from the company's perspective, in the aggregate, in the numbers that described the team's performance without describing the team. Gil was forty-four years old. He had been with the provider for six years, in three different client assignments, in each of which he had been described in his performance reviews as thorough and reliable and outcomes-focused, which were the words the provider used for people who did the work correctly, and which were accurate and insufficient in the same way that Appendix C was accurate and insufficient. He had read the transition materials. All of them. Not because he was required to read all of them. Because he read everything. He had read the forty-three page script and the knowledge transfer assessment and Appendix C and the Cultural Alignment Obligations clause on page eighty-nine. He had read the Cultural Alignment Obligations clause three times. He had then gone to his manager and asked a question. His manager's name was not important. His manager was a regional operations director who managed four client accounts and whose primary accountability was to the contract renewal metrics. He was not unkind. He was oriented in a specific direction and Gil's question was not in that direction. "What does this mean in practice?" asked Gil. He had the contract on his desk, open to page eighty-nine. "Cultural alignment. How do we actually do this?" "That's a legal protection for the client," said his manager. "We demonstrate it through the training programme and the quality framework." "But the training programme is the standard module," said Gil. "And the quality framework is Appendix C. The clause is asking for something more specific than that." "The clause is asking for reasonable steps," said his manager. "The training and the framework are reasonable steps." Gil looked at the clause again. "I think they want Priya," said Gil. His manager did not know who Priya was. "Focus on the script and the KPIs," said his manager. "The client will measure us against the SLA." Gil went back to his desk. He did not raise it again through that channel. But he did not stop thinking about it.

Iris and Marco had been colleagues for three years. They had met in the second week of the new twelve-desk office when they had been assigned to adjacent desks in the operational team and had discovered, over the course of a Monday morning, that they had both grown up in port cities, both played an instrument badly and with great enthusiasm, and both had a strong and specific opinion about the correct way to make coffee that turned out to be identical. The friendship had developed from there. It was genuine and specific and had the comfortable texture of a friendship between two people who genuinely like each other and who have the particular ease that comes from not wanting anything from each other that the friendship does not already provide. Marco was thirty-four. Iris was thirty-two. They had lunch together two or three times a week. They talked about everything. About work and about their lives outside work and about the company and about the things they found funny and the things they found frustrating and about what they were going to do in the future, separately and occasionally in the abstract plural. The friendship was real. What happened next was also real, and the two realities did not know how to coexist with each other when they arrived in the same space. Marco told Iris how he felt on a Wednesday evening in the fifth month of the transition period. They had stayed late after a planning meeting and were walking toward the lift together when Marco said, without preamble and without the performance of having planned it, that he had been in love with her for about eight months and that he wanted her to know that and that he was not telling her because he expected anything but because not telling her felt like a dishonesty he could not sustain. Iris stopped walking. She looked at him for a moment. She could see that he meant it. She could also see, with the clarity of someone who has not been thinking about a person in the way they are telling her they have been thinking about her, that the friendship she valued and the thing he was describing were not the same thing. "I am not interested in that," said Iris. Her voice was even. Not unkind. "I want to be honest with you the same way you've been honest with me." "I know," said Marco. "I didn't think you would be. I just needed to say it." "The friendship matters to me," said Iris. "It matters to me too," said Marco. They stood there for a moment, in the corridor outside the lift, in the particular awkwardness of two people who have been honest with each other and who are now figuring out what honesty costs. Then Iris pressed the lift button. "Same time for lunch on Thursday?" said Iris. "Yes," said Marco. They rode down in the lift and went home in separate directions. For four weeks it seemed like the friendship had absorbed the conversation and continued. The lunches happened. The ease was mostly there. Marco did not repeat what he had said. Then, in the sixth week, he said it again. Differently. Not as a declaration. As a plea for understanding. He said that he was finding it hard and that he wanted her to know he was finding it hard, and that he was asking her to acknowledge that and to give him some indication of how they navigated this together. Iris said that she had nothing to add to what she had already said. Marco said that he understood but that understanding and accepting were different things. The conversation ended there. But it did not end there. It continued in the specific way that conversations continue when one person is not finished with them, in the peripheral looks and the slight adjustments of behaviour and the one additional message sent at eleven at night that said only that he was sorry and that he did not know how to stop feeling this way. Iris read the message. She did not reply. She sat with it for a long time.

Maya had prepared three sessions for the discovery team. The first session was on the knowledge base. Maya walked Diane and one of the analysts through the structure of the documentation, explaining not just what was in it but why it was organised the way it was, what decisions had been made about what to include and what to leave out, what the evolution of the documentation revealed about the evolution of the product and the customer base and the nature of the problems the support team most commonly encountered. The session lasted four hours. Afterward Maya came to find the CEO. "They are thorough," said Maya. "They asked good questions." "But?" said the CEO. "They are treating it like a document review," said Maya. "They are trying to understand the structure so they can replicate it. They are not trying to understand the decisions behind the structure." "Is that a problem?" said the CEO. "The structure exists because of the decisions," said Maya. "If you replicate the structure without understanding the decisions you get something that looks like the knowledge base and works like a filing system. When a new situation comes up that is not in the documentation, which happens every week, you need to understand the decisions to know how to handle it. If you only understand the structure you have to escalate it or leave it unresolved or handle it in a way that feels right but is not aligned with what we would do." "What did you tell them about that?" said the CEO. "I told them," said Maya. "Diane made a note. I am not sure it landed." "What makes you say that?" said the CEO. "Because the question she asked after I told her was about the escalation process," said Maya. "Which is a good question. But it was the wrong next question." The CEO thought about this. "I am going to try again tomorrow," said Maya. "I have the second session on the customer types. I will try to build it in through the examples." "Thank you Maya," said the CEO. "I just want them to understand it properly," said Maya. "I know," said the CEO. It was the same exchange they had had three weeks earlier, when Maya had been preparing the notebooks for the discovery phase. The CEO noticed this. The repetition of it felt like something. Not a warning exactly. More like a note played twice in a piece of music, the second time drawing attention to itself by virtue of having been heard before. On the provider side, Gil had spent the first week with his twelve agents, running them through the script, watching them practise the flows, noting the places where the script did not quite cover what the customer was likely to ask. He had made a list. He had taken the list to his manager. "These are edge cases," said his manager. "They are going to come up," said Gil. "I have read the product documentation. These are not edge cases for this product and this customer base. They are the second most common call type." His manager said he would pass the list to Jonathan. The script amendment process took three weeks and produced two changes. The other items on Gil's list were described in the response as adequately covered by existing pathways. Gil looked at the response. Then went back to his agents and built in additional coaching for the situations the script did not cover, drawing on the knowledge transfer documentation and on his own reading of the product materials and on the conversations he had been having with Kwame, who had the technical knowledge Gil did not have and who shared it without the friction that technical knowledge sometimes comes with. He did this on his own time. Not because he was required to. Because it needed to be done. Nobody above him in the provider structure noticed this. Nobody in the company knew it was happening. Kwame spent the first week with Joel. Joel showed him the systems with the same walk-through approach he had used with the CEO in the early days. At the whiteboard, drawing as he talked, thinking made visible rather than a finished product being demonstrated. Kwame was technically competent. More than competent. Joel told the CEO, at the end of the first week, that Kwame understood the architecture quickly and asked the right questions about the integration points and had identified two dependencies that Joel had not explicitly documented and that would have caused problems if they had not been caught during the discovery phase. "That sounds positive," said the CEO. "It is," said Joel. "The technical integration is going to be fine. Kwame knows what he is doing and his team knows what they are doing. The systems will talk to each other. The data will flow correctly. The reporting will work." "But?" said the CEO. "The systems are not the problem," said Joel. "The systems are the visible part. The invisible part is the decisions that happen between the system outputs and the customer interaction. The system tells the agent what the customer's account looks like. The agent then has to decide what to do with that information. That decision is where the knowledge lives. Not in the system." "And that is what the knowledge transfer sessions are for," said the CEO. "Yes," said Joel. "I just hope they are working." "Maya thinks they are not landing the way they should," said the CEO. "I know," said Joel. "I had lunch with her yesterday. She is trying." "I know she is," said the CEO. "The problem is that you cannot transfer context in a session," said Joel. "Context accumulates. It builds over months and years of doing the work and making mistakes and seeing what happens when you get it wrong and adjusting and seeing what happens when you get it right and adjusting again. Four weeks of sessions cannot replicate that. It was never going to be able to." "I know," said the CEO. "Then why?" said Joel. "Because the alternative was not doing this at all and the board had decided we were doing this and I had decided we were doing this and here we are," said the CEO. Joel looked at the CEO for a moment. "I am not criticising the decision," said Joel. "I am trying to understand what we do about it." "We make the transition as good as we can make it and we watch what happens and we fix what breaks," said the CEO. "And if what breaks is not fixable from the outside?" said Joel. The CEO did not answer this question.

The second week of the discovery phase produced a document. Diane presented it to the CEO and Sandra on the Wednesday of the second week. It was called the Knowledge Transfer Assessment and it was thirty-one pages long and it was organised into five sections covering process knowledge, product knowledge, customer knowledge, system knowledge and what Diane called cultural knowledge, which was the section the CEO turned to first. The cultural knowledge section was four pages long. It described the company's customer service values using language that had been taken, largely, from the company's own internal documents. It described the expected behaviours of support team members in customer interactions. It described the escalation culture, meaning the circumstances under which issues were escalated and the expected response times. It described the feedback culture, meaning the process by which customer feedback was collected and shared with the team. The CEO read all four pages. "Where is Priya in this?" said the CEO. "I am sorry?" said Diane. "There was a member of the support team called Priya," said the CEO. "She is no longer with the company. She left before the transition. She represented something about the culture that I would describe as the most important thing we have. I am not seeing it in these four pages." "Can you describe what you mean specifically?" said Diane. "A customer called," said the CEO. "The problem was partly the company's responsibility and partly not. Priya stayed with it for forty minutes until it was sorted. Not because the process required her to. Because leaving it unsorted was not something she considered. That instinct is what I am looking for in these four pages." "I understand," said Diane. "I think what you are describing maps to what we have documented under expected customer ownership behaviours in section three." The CEO looked at section three of the document. Section three described expected customer ownership behaviours as follows. Support agents are expected to take ownership of customer issues and to pursue resolution through available channels before escalating. Agents are expected to communicate clearly with customers about the status of their issue and expected resolution timelines. Agents are expected to follow up on escalated issues to ensure resolution has been achieved. The CEO read this three times. "This is the process version of what Priya did," said the CEO. "It is not Priya." "I understand the distinction you are making," said Diane. "I want to be honest with you that the distinction you are making is one that is very difficult to codify in a knowledge transfer document. What we can do is build it into the training and the quality assurance framework and the coaching approach that our team leads will use with the agents." "And how well does that work?" said the CEO. "It depends on the team," said Diane. "We have had transitions where the culture transferred very effectively and transitions where it was more challenging. The quality of the original documentation and the commitment of both sides to the process are the biggest variables." "And in the cases where it was more challenging, what happened?" said the CEO. "There was usually a period of adjustment," said Diane. "Customer satisfaction scores would dip during the stabilisation phase and then recover as the team built familiarity with the product and the customer base." "How long is the period of adjustment typically?" said the CEO. "It varies," said Diane. "Somewhere between three and nine months in most cases." The CEO thought about three to nine months of customer satisfaction scores dipping. About the customers who would call during those three to nine months and have an experience that was different from the one they had been having. About whether those customers would still be customers at the end of the nine months or whether they would have made the quiet decision to look elsewhere that customers make without announcing, the same way that the team members who had decided to leave were sitting differently at their desks. "Thank you Diane," said the CEO. "This is useful." "We are committed to making this work," said Diane. "I know you are," said the CEO. Sandra resigned on the last Friday of the discovery phase. Not dramatically. Sandra did not do things dramatically. She came to the CEO's office at four in the afternoon and sat down and said, in the clear and methodical way that Sandra said everything, that she had given the situation a great deal of thought and had concluded that she was not the right person to manage the support function through and beyond the transition. "I am sorry to hear that," said the CEO. "I have enormous respect for this company and for what has been built here," said Sandra. "I am not leaving because I am angry. I am leaving because I am honest. I cannot manage a team to a standard I do not believe the structure will support. That is not a fair thing to ask of myself or of the team." "Is there anything I can do to change your mind?" said the CEO. "No," said Sandra. "I have thought about it carefully and this is the right decision." "How much notice are you giving?" said the CEO. "Four weeks," said Sandra. "I will do everything I can to support the handover." "I know you will," said the CEO. "Thank you Sandra. For everything." "Thank you for building something worth working for," said Sandra. "I mean that." Then she said something she had not planned to say. "I want you to think about Owen," said Sandra. "What about Owen?" asked the CEO. "He has been managing his team through all of this better than anyone has managed anything," said Sandra. "His team is calm and productive and doing good work in a situation that is genuinely difficult. He has done that. Nobody above him has noticed." The CEO was quiet for a moment. "He should have been in the room more," said the CEO. "Yes," said Sandra. "He should have been." She left the office and the CEO sat for a long time looking at the space where she had been sitting. Sandra was the third person who had defined what the company was in the places the metrics did not reach. Maya had built the knowledge. Joel had built the systems. Sandra had built the people. The three of them together had created the conditions in which someone like Priya could exist and thrive and produce the kind of customer experience that made a customer write a paragraph about a forty-minute phone call. Priya was gone. Sandra was leaving. Maya was still here. Joel was still here. But they were product and operations, not support. The support function was being handed to a team who had spent four weeks conducting a thorough and professional discovery process and who had produced a thirty-one page Knowledge Transfer Assessment with a four-page section on cultural knowledge that described the process version of Priya and not Priya. The CEO opened the laptop. The board presentation was on the screen from the previous session. The four bullets. The fourth one deleted. Added it back. Cultural alignment obligations included in contract. Then deleted it again. Then closed the laptop and went home and did not think about the four-page cultural knowledge section for the rest of the weekend, which was not the same as not thinking about it but was the closest available approximation. On the Monday of the third week Diane's team began the joint operation phase. The provider's agents, twelve of them, had been selected and trained and briefed and had completed the knowledge transfer assessment with scores that met the defined threshold. They were, by every measurable standard, ready. They sat at temporary workstations set up alongside the existing team. The existing team, reduced now by the departures, by Priya and by six others who had left in the weeks since the announcement, handled the cases as they always had. The provider's agents observed and then, gradually, began handling cases themselves, with the existing team available for questions. The CEO walked through the support area on that Monday morning and stood for a moment at the edge of the room. The provider's agents were at their stations. They were professional. They were attentive. They were doing the work in the way they had been trained to do it, which was the correct way as described in the Knowledge Transfer Assessment and the training materials and the Appendix C rubric. They were not Priya. Not one of them was Priya. Not because they were inadequate. Because Priya was not a level of skill. Priya was a combination of instinct and context and genuine ownership that had developed over three years in a specific environment that no longer existed in the form that had produced her. The CEO watched the room for a few minutes and then went back to the office. On the desk was a note from Maya. The note said: one of the provider's agents asked me this morning what to do when a customer's problem is partly the company's responsibility and partly not. I told them to stay with it until it was sorted. They asked what the escalation process was for the part that was not the company's responsibility. I said there wasn't one. They asked how they were supposed to handle it then. I said: you just stay with it. They asked: for how long? Maya had written at the bottom of the note, in smaller letters than the rest: I did not know how to answer that.

End of Chapter 6

Writer's Thought:

For how long? Maya does not know how to answer that. Neither do I.

Here is What is Broken. The CEO. The Transfer.

MarvinPro | March 2026

marvinpro.com

Think Simple.