Then came a text: “Brooke, call me immediately. ”
I stared at it longer than I should have. I knew exactly why Thomas was reaching out. Lumicore had launched Project Relay without closing four red items, and the system had failed exactly the way I had warned it would.
But I wasn’t their employee anymore. Management had decided someone else should run the launch, and that Monday, someone else was running it. I had completed every required handoff session before my departure. I hadn’t deleted anything, disabled anything, hidden any password, or touched a Lumicore system after my access was formally closed.
I was finally doing my job instead of everybody else’s. I had spent years at Lumicore as the person who bridged gaps no one else wanted to own. Hospital procurement teams imposed their own vendor requirements. Compliance interpreted regulations one way, finance another, and my department sat in the middle.
When two groups read the same rule differently, I was usually the one asked to figure out which interpretation would still work once an actual hospital order hit the system. None of that made me famous. Management loved what Project Relay promised to produce—a faster path from hospital order to revenue—but they didn’t love my warnings about what could go wrong if we skipped the boring parts. Thomas ran the program.
He had built his career in commercial leadership, where hesitation could lose a customer and complicated explanations usually meant somebody was avoiding a decision. He was personable, confident, quick with customers, and very comfortable in rooms where other people were still reading the agenda. Brent, the consultant brought in to support the launch, was the same type. They had known each other for years through industry events and charity golf tournaments.
Brent had been a successful regional sales executive before moving into consulting. He was exactly the kind of person Thomas wanted beside him when things got tense. I had raised my concerns during the planning phase. “Why do we need two separate validation windows here?
” Brent asked one afternoon. “Because customer credential activation and bank authorization are controlled by different organizations,” I said. “If one finishes late, we need time to validate the downstream transaction before production. Statistically, how often do both fail?
”
“They don’t both have to fail. One is enough. ”
Finance had to complete production approval before revenue could post normally. The risk register I maintained included a line written in language that required no engineering degree: “Failure to complete production authorization before legacy termination may prevent invoice recognition.
” I flagged it repeatedly. But Thomas had already mentioned the accelerated date during a board discussion, and once an executive turns an internal target into a public commitment, schedule becomes more important than risk. My warnings had become inconvenient to the story Thomas wanted to tell. So they moved me aside.
In a meeting I wasn’t expecting, Thomas introduced Brent as the new launch lead. “You’ll retain your director title, but for the remainder of the launch, you’ll report into him on Relay and complete a full knowledge transfer. ”
“I’ll need you to help me translate the technical stuff,” Brent said. “Then we should clarify authority before the transition,” I said.
“Who owns escalation if authorization is still incomplete at cutover? Who makes the go or no-go decision Friday night if red dependencies remain? ”
“Use your judgment during handoff,” Thomas said. That wasn’t an answer, and I knew it.
I could have pushed harder. I could have argued that ambiguity was exactly how failed launches happen. But I was tired of being the person who absorbed risk while others took credit for the wins. So I nodded, completed the knowledge transfer sessions, and added one sentence to my final written status report: “Launch should not proceed until all four items are closed.
”
Nothing failed immediately. The days after my departure were quiet. I started a new role at Clearbrook Payments Group, where the work was cleaner and the ownership lines were drawn in ink. I thought about Lumicore less and less.
But I had been bridging those gaps for so long that old habits followed me. When a manager at Lumicore sent me a production acceptance form because historically I had reviewed them before launch, I reviewed it. When Treasury asked whether I could check a banking partner status, I checked it. Not because I had any formal responsibility—because letting those gaps remain open would have hurt customers.
That isn’t the same thing as being responsible. I told myself that repeatedly. I was helping out former colleagues, not running their launch. Every instruction I gave them was already in the shared repository.
I was just making sure the obvious things didn’t get missed. Thomas called me into his office before lunch one Friday. The tone was friendly, but the message was clear: I needed to stop answering questions. “We’ve got it under control,” he said.
“Brent is handling the technical coordination now. We need you to fully step back. ”
“I understand,” I said. I let that sentence sit between us.
Then I walked out. The following Friday was the launch. I didn’t know the details until later. The implementation team believed Brent had already handled a critical item because it was marked complete in the program summary.
But a hospital test transaction failed authentication because one credential set was still in staging status. Instead of stopping, the team kept retrying—nobody panicked because orders could still enter the environment. The new certificate had to be synchronized with the banking endpoint before financial authorization could complete. The route stopped accepting transactions because Lumicore had told the legacy system it was no longer needed.
Each attempted workaround exposed another dependency. The controls worked exactly as designed. Monday morning, Lumicore recorded zero revenue. Finance could not accept a revenue path that had never completed an end-to-end transaction.
The items that should have been closed before launch were exactly the items that brought the whole system down. I saw the first missed call while sitting in my first planning session at Clearbrook. Then came the text: “Brooke, call me immediately. ” Then another.
Then a message that reached me indirectly—Thomas explaining that they needed help and hoping I could step in unofficially. I read it twice. Then I answered: “Thomas, I completed the formal transition and all required handoff sessions before my departure. I no longer have access, and I no longer have authority to intervene.
If Lumicore needs assistance, the proper channel is through Clearbrook. ”
Twenty minutes later, my phone rang from a number I didn’t recognize. It was a senior executive at Lumicore, not Thomas. She explained that the company wanted to request temporary professional assistance through proper channels and understood that I was now employed by Clearbrook Payments Group.
I walked into the Clearbrook COO’s office and told her exactly what was happening. She didn’t blink. “If we do this, we do it cleanly,” she said. Lumicore would receive limited recovery support under a professional services agreement.
I would have written authority for the work, explicit access boundaries, and no responsibility for decisions outside the engagement. The consulting rate was substantial. By Tuesday morning, I was back inside Lumicore for the first time since leaving. I wasn’t returning as Thomas’s employee.
I was returning because the company had formally purchased expertise it had recently decided was unnecessary. My badge was temporary, my scope was documented, and my name was on a contract instead of an org chart. The first page of the launch status still showed all four items in red. I didn’t ask who approved launch with those open.
I already knew. Instead, I walked through each item and gave the team clear direction: no shortcuts, no one changes status because the color is uncomfortable. The vendor synchronized the new certificate with the banking endpoint, then ran controlled transaction tests instead of assuming connectivity because a dashboard icon had turned green. Two items were completed quickly.
The others required solutions that involve waiting for organizations they don’t control—which is exactly why they should have been resolved before cutover. I watched the team work through the recovery. Transactions started posting again. Customer service confirmed priority orders, and distribution released shipments in controlled waves.
By Thursday, the system was stable. By Friday, most delayed revenue was either posted or scheduled. The internal review found no sabotage, no deleted information, no hidden credentials, and no employee misconduct. The company did not collapse because of one failed launch weekend.
Most of the delayed revenue was eventually recovered. One midsized account moved its business elsewhere after its procurement team concluded that the failed launch exposed too much operational risk. Thomas left Lumicore a few months later. Brent returned to consulting with a quieter résumé.
The company revised its launch governance to require that go or no-go criteria be mandatory and that status changes be approved outside the program team. The lessons were documented, but the experience left a mark on me that went deeper than process. I thought about how easy it had been to become indispensable by absorbing everyone else’s ambiguity. I thought about how long I had told myself I was helping when I was actually carrying weight that was never mine to carry.
I thought about the satisfaction of that Monday morning when I heard Lumicore had recorded zero revenue. It wasn’t satisfying because people panicked. It was satisfying because for once I had not created the emergency, and I had not volunteered to absorb it. I had warned them clearly, and then I had let them own the consequence.
At Clearbrook, the rules were different. If someone owned a dependency, their name was written beside it. No one asked me to handle something because I had “always handled it. ” When the COO asked me one afternoon whether I wanted to take on a broader coordination role, I considered the offer carefully.
I wasn’t going to pretend compensation was spiritually irrelevant. But I also knew I never wanted to be the person who became necessary by keeping things unclear. So I said, “Only if we enjoy having revenue on Mondays. ”
She laughed, but she also understood what I meant.
Mason, a colleague who had heard enough of the story over the following months, asked me once why I had changed jobs. I told him the simplest version: I had finally learned that unclear ownership always sends an invoice eventually. I still think about those four red items sometimes. I think about how many people saw them and decided someone else would handle it.
I think about how close I came to being that someone else again. But I didn’t. I let them own it. And I went back only when they were willing to pay for my expertise, not because I felt obligated to save them from themselves.
That was the lesson I carried forward: being helpful is not the same as being responsible, and being responsible is not the same as being everything to everyone.
