Writing
After a decade with one supplier, a tender stops being a statement of need and becomes a description of what the client already has. One line gave the whole thing away.
A client had used the same supplier for the better part of a decade. The contract reached the end of its term, the work went out to tender, and we bid against the supplier already doing it. The same situation turns up wherever services are bought on long contracts: software, facilities, logistics, payroll.
Their requirements document said two things that do not sit together. They are not particularly unhappy with the system they have, which controls who comes on site, when, and for how long. And, several pages later, they would like the ability to add a bay: one more space for arriving delivery vehicles to park and unload.
Adding a bay is trivial. That it had to be requested is not. Ten years with one supplier had turned the requirements document into a portrait of that supplier, and the bay was the single line that gave away its limits.
When one supplier has held the work for ten years, the document that goes out for the retender is not a neutral statement of need. It mostly describes what the client already has. The people drafting it have only ever seen the problem solved one way, so that system's assumptions have turned into the requirements, and its limits have set the boundary of what anyone thinks to ask for.
It happens in plainer categories too. Payroll retenders routinely specify approval steps and cut-off times that exist because the incumbent runs batches overnight, not because the business needs them. The constraint has been quoted back as a requirement for long enough that nobody remembers it was ever a constraint.
If the document largely describes the current supplier, a strong response to it mostly proves the current supplier is doing a good job. There is no version of competing on their own description that ends well.
So the document is more useful read for what it reveals than for what it asks. Two things carry most of that signal.
The small request that should not need making. A client should not have to ask a supplier for the ability to add a bay. Asking for it indicates that adding capacity currently means a change request, a quote and a wait, and that this has been the arrangement long enough to seem normal. One constraint was irritating enough to be written down. The others have been worked around for years without being recorded anywhere.
What the document does not mention. It treats congestion and overstaying as things to be flagged: an alert fires, someone responds. That is genuine capability, and it sits downstream of the cause. If the slot allocation was optimistic, or the booking process pushed people into the wrong window because that window was easier to select, the congestion existed before anyone arrived. The document also says nothing about helping the person at the gate decide anything. None of that is an oversight.
If the task is to change what the requirement is understood to be, the next question is where that language comes from. It does not come from inside the category. Every supplier bidding is reading the same document, and the category's own vocabulary has been shaped by the same generation of systems, so it will only produce variations on what already exists.
It comes from an adjacent field that solved the same human problem under harder commercial pressure. Pharmaceutical sales CRM has spent years on next best action: instead of showing a rep the account history, the system proposes the action most likely to work in that moment. The investment followed the shape of the decision. A rep has a few minutes, incomplete information, one objective, and a measurable loss if they miss a step. Fail to raise a medical inquiry, or fail to bring a medical science liaison into the next conversation, and the opportunity is gone.
An officer at a gate has the same shape of decision. A few minutes, incomplete information, keeping traffic moving as the primary duty, and a measurable cost to hesitating. What they do not have is the investment, because their category never described the job in those terms. More data does not help them. A recommended action with the reasoning attached is a different proposition, and it appears nowhere in the requirements because it could not have.
The transfer holds because the decisions have the same structure, not because both run on software. Borrowing on any looser basis is decoration.
Seeing this much is the easy part. Everything that makes a reframe fail happens after you have one.
Someone started this process. A ten-year supplier with a reasonably content client does not go to tender by accident, and the likeliest explanation is the dullest one: the term ended and procurement policy required a competition. A tender driven by policy rather than dissatisfaction has no internal sponsor for change. It usually also means the budget is under review rather than under expansion, because a content client following policy has little reason to keep a seven-figure commitment at its previous level.
That puts everything above under pressure. A wider set of requirements is more scope, and more scope normally costs more. To a panel briefed to hold or reduce spend, that reads as inflation, so the more convincing the expansion, the more expensive it looks.
The temptation is to answer by claiming the same envelope buys more. That does not survive contact with a procurement team, and it contradicts the argument that got you here. If the client cannot see past the system they have, they cannot tell you which parts of their current spend have become unnecessary, and from outside the building you do not have their numbers either.
What can be done is narrower. Each addition has to be tied to a cost the operation is already absorbing somewhere other than the contract: time at the gate, congestion, staff chasing vendors who booked approximately, small changes that require a quote because they should have been settings. Those costs are invisible for the same reason the requirements are narrow, which is that nobody has had a reason to itemise them for a decade. They sit in the operation, not in the line being defended.
From outside you can only size them with stated assumptions and offer to confirm them together. That is weaker than a precise saving and it is the only honest version, which is also the better position: the first thing a competent panel does with an unsourced number is ask where it came from.
If those figures do not hold up, that is not a gap to write around. It is information about whether to bid at all.
The panel is scoring against a matrix built from the original requirements. Introduce anything outside it and there is nowhere to record the answer. At best it scores nothing. More often it is treated as unrequested scope, which reads as a supplier who did not follow instructions. So anything additional has to arrive with the means to assess it: the criterion named, the reason it belongs, and a way to compare suppliers against it. You are asking people to widen their scoring model partway through a process, and they need something defensible to do it with.
The second is the size of the request. Changing how staff work, how vendors book and how decisions get made at the gate is a bigger ask than replacing a system with a similar system. The current supplier does not have to make that ask at all.
The benefit of switching is spread across the organisation, while the consequences of it going badly land on whoever signed off. That person will be defending the decision in meetings you do not attend. What they need is not an assurance that it will be smooth, but something detailed enough to hold up when a gate backs up in month three. A bid that promises a painless transition gives them nothing to use.
It is worth granting what the incumbent has earned. Ten years without the client becoming unhappy is an achievement, and a panel leaning towards continuity is usually making a reasonable judgement about risk rather than a lazy one. Sometimes the right answer is to renew, and a challenger who reads every long tenure as complacency will misjudge the room.
This approach fits when the existing supplier is performing, the client is content, and the requirements have settled around one way of working.
It does not fit in two cases. If the client is genuinely unhappy, the dissatisfaction is the opening and reframing wastes it. And if the tender is purely a term-expiry exercise with a reduced budget, a fixed scoring model, and no access to anyone able to widen the criteria, then the requirements really are the requirements. There is no sponsor to carry a different idea and no money to pay for one. In that situation the right decision is not to bid.
The bid is still live, so I do not know whether this was the right read. The panel may have wanted the direct product comparison I chose not to make.
What I would say is that the answer to the bay request was never going to be that we would add their bay. It was that they should not have to ask. That is the whole argument in one line, built from a detail the client thought was minor.
If they do not buy, they will still have language for something they were describing differently a few months ago, and some of it will appear in the next tender. Losing a competition after changing how the problem is framed is a different outcome from simply losing.
Details kept general while the bid is open.