When Selecting the Wrong Client Kills Your Agency
Most conversations about vendor selection focus on one direction: how organizations choose their partners. Less often do we ask the inverse question — how vendors choose their clients — and what happens when both decisions go wrong at the same time.
What follows is a story about exactly that. I watched it unfold from the inside, as a participant in the institutional process that set its conditions. I’ve thought about it many times since. It doesn’t have a villain. It has two parties who each made understandable decisions that, in combination, produced a catastrophic outcome. One of them absorbed the consequences and moved on. The other ceased to exist.
The Setup
A large, well-resourced research institution decided to undertake a significant website redesign. After an extended — and expensive — internal requirements-gathering process, they ran an RFP and eventually narrowed their selection to two vendors. The final decision fell to a senior administrator who, faced with two proposals she may not have had the technical context to differentiate on merit, selected the one with the lower price. Whether cost genuinely seemed like the right criterion, or whether a defensible budget story held appeal in its own right, is something I can only speculate about. What is clear is that price became the deciding factor, and the process offered no mechanism for questioning whether it should be.
The selected agency was small. Scrappy in the best sense. They had brought genuine intelligence to the discovery process — the kind of careful listening and sharp problem reframing that larger firms, padded with business development overhead, often can’t match. Their work in the early phases of the project was good. They understood the problem. They had earned the contract.
The project that followed would eventually end them.
The Collision
What neither party had adequately reckoned with was the institutional environment itself.
Large research institutions — universities, hospitals, foundations — operate on timelines that are largely incomprehensible to outside vendors until they’ve worked inside one. Approvals move through multiple layers of stakeholders whose competing priorities have nothing to do with the project. Feedback cycles that should take days take weeks. Key decision-makers go on sabbatical, get pulled into other crises, defer to committees that haven’t been convened yet. This is not dysfunction in the pathological sense. It is the normal operating rhythm of a complex institution, and it is entirely predictable — if you know to look for it.
The agency hadn’t built that institutional drag into their timeline or their budget. They were ostensibly working on a time-and-materials basis, but their original estimate had been written into the final contract as a strict not-to-exceed (NTE) cap. This meant that every delayed approval, every re-opened decision, and every week spent waiting for a stakeholder to return a round of feedback was a week of runway quietly burning. The old website still worked. The project would get done eventually. There was always something more pressing.
By the time the project reached roughly two-thirds completion, the agency had hit that NTE cap and exhausted their budget. They requested an extension — a standard ask in this situation, and a reasonable one given the circumstances. The institution refused. Equipped with more sophisticated legal and procurement resources than the agency could mobilize, they used the capped contract to their advantage, compelling the agency to complete the remaining work without additional compensation.
The agency complied. The work got finished. And then, ground down by the financial and organizational toll of that final third, they eventually closed.
Two Failures, One Outcome
It would be easy to frame this as institutional malfeasance — a powerful client exploiting a small vendor. That reading is not entirely wrong. The institution had the resources to act differently and chose not to. The contract became a weapon rather than a framework for a working relationship.
But the full accounting is more complicated than that, and more useful.
The institution entered a major technology project without having established the internal decision-making infrastructure to execute it on a reasonable timeline. They hadn’t designated clear decision rights. They hadn’t protected the project from the ordinary chaos of institutional life. They hadn’t thought carefully about what capped time-and-materials contracting would actually mean in their particular environment — an environment where time, reliably, costs more than anyone plans for. These are not exotic oversights. They are common ones. But they have consequences, and in this case the consequences were exported almost entirely onto the vendor.
The agency, for their part, took on a client that was operating at the outer edge of what they could safely absorb. The engagement was a significant reach — in budget, in institutional complexity, in the kind of political navigation that large organizations require. Their discovery work was excellent. Their craft was solid. But excellent craft and institutional fluency are different competencies, and the gap between them, in this context, was fatal. A more experienced institutional vendor might have built timeline contingencies into the contract. Might have flagged the decision-making bottlenecks earlier and in writing. Might have had the leverage to push back when the extension was denied. The agency didn’t have those tools, and the RFP process that selected them had no mechanism for surfacing whether they did.
Evaluating this kind of fit requires asking hard questions about how vendors manage complexity. Our guide to evaluating technology partners includes a section on “Behavioral Signals During the Sales Process” that covers exactly this.
What This Story Is About
Client selection and vendor selection are mirror problems. The questions run in both directions: Is this the right partner for our challenge? Are we the right partner for theirs?
The institution never seriously asked whether its own internal processes were mature enough to support the kind of engagement it was procuring. The agency never seriously asked whether this client’s institutional complexity was something they were equipped to navigate. Both assumptions went untested until the project collapsed under their combined weight.
This is what I mean when I say the RFP process produces the appearance of due diligence more reliably than the substance of it. A rigorous selection process would have surfaced some of these questions before contracts were signed. It would have asked vendors not just what they could deliver, but how they manage institutional timelines and decision bottlenecks. It would have required the institution to articulate its own internal governance structure as part of the brief. It would have created the conditions for an honest conversation about fit — in both directions.
Instead, the process optimized for a signed contract. The signed contract is not the goal. The working relationship is the goal. And working relationships require both parties to understand, in advance, what they’re actually getting into.
The institution got a website. The agency got a lesson they didn’t survive long enough to apply.
Launch Day Advisors is a buyer-side advisory that helps organizations select technology and design partners without traditional RFP processes — and helps both sides ask the right questions before anything gets signed.
