Situation
Two mission-driven financial institutions were merging into one organization. No name, no brand architecture, no shared identity. A new CEO arrived as the work began. We were engaged for the whole arc: discovery, naming, brand architecture, visual identity, business systems, guidelines, launch toolkits. And, throughout, educating the client team on what branding actually is — so the decisions in front of them would be understood as decisions, not preferences. Research ran as a ladder, not a survey: a short written survey to a defined tier of employees; a long-form session with people outside the stakeholder set; a session with the stakeholders; then a session with the incoming CEO. Running the non-stakeholder group before the stakeholders was the deliberate call. We begin every engagement by naming who the stakeholders are and will be for the duration of the work. Approval only means something if the approvers stay the same. Before committing, I also vet whether a client is actually ready and able to receive the work they're asking for. This client screened as ready. They had the mandate, the budget, the appetite, and a new CEO with a clear vision.
Decision
Every round, the real question wasn't whether the work was right — we had approved briefs and approved deliverables proving it was. It was how hard we can push before defending the work costs us the relationship. That's the call I was making, over and over, with incomplete information. Push too little and the work erodes. Push too hard and you're the difficult vendor, and the difficult vendor loses the next argument before it starts. I kept choosing to spend relationship capital to protect the work, and I kept running lower on it, and eventually there wasn't enough left to win with. I don't think I chose wrong in any individual round. I think I was solving the wrong problem — playing each round as a design argument when the failure was structural and no amount of winning individual rounds could fix it.
Outcome
The naming and architecture held. Internal discovery, naming exploration, brand architecture, a securities sub-brand. The structure survived. The visual identity didn't. The brief asked for bold, direct, and distinct. We delivered that. Then we went roughly ten to twelve internal rounds, three or four of them presented to the client, defending it each time with the strategy, the design rationale, and everything the discovery process had surfaced. We usually won those arguments. It didn't matter — because the people in the room kept changing. A merger rewrites the org chart while you're working inside it. New people arrived mid-process holding real decision-making weight, having attended none of the four discovery sessions and read none of the briefs. Evidence doesn't compound if the audience resets. Both safeguards existed and both had been voided by the same thing. What shipped was substantially more conventional than what we were briefed to make.
Reflection
Treat a broken stakeholder list as a stop-the-work event, not a headwind to manage. The moment new approvers began arriving mid-process, the right move was to halt and renegotiate the terms — not keep presenting into a process that no longer had a closed door. Ten rounds of being right is worth less than one round of stopping. The deeper lesson is about the readiness check itself. I was testing readiness as a fixed property, something true or not true at kickoff. It isn't. Readiness is a state an organization can lose, and a merger is exactly the condition that takes it away. A brand can't be braver than the organization carrying it. I'd still take the work. I'd just build the check to run continuously instead of once.
