Deciding What Not to Localize Has a Deadline
A brand with a Japan launch on the calendar arrives at the same point in the same order. The market is chosen. The budget exists. What is unsettled is how much of the brand has to change for it — the copy, the imagery, the product presentation, the site, the retail idea — and the sensible next move is to ask someone who does this for a living.
That conversation is usually free, and it usually happens with the party who would produce whatever ends up being changed.
This article is not about whether that party is honest. It is about where the conversation sits in the sequence, and what its position does to one particular answer: change less than you expected, or leave this part alone.
We have written separately about why a guideline that works at home stops persuading once it arrives in Japan. This is the step immediately after that one. Not why the document loses its grip, but when you can still decide what to do about it and have the answer go either way.
The decision that comes before the brief
Our own services page puts this step ahead of everything else:
Every engagement starts by deciding what to change and what to protect.
Read that as a sequence rather than as a slogan. Deciding is a step. It has inputs — market context, a reading of the brand, a view on which parts of it are load-bearing — and it produces a conclusion. The conclusion is a document. Nothing gets built by it.
Which means it does not look like work in the way the things after it look like work. A campaign is visible. A store fit-out is visible. A page that says “we looked at the product line and concluded that three items need new photography and the rest can ship as they are” does not photograph well and has no launch date.
It is also the step at which change nothing here is still an ordinary thing to conclude rather than a reversal. Everything downstream of it has already accepted that something is being made.
Free scoping is paid for somewhere
We give a reason on our own services page for why our first program is billed rather than given away:
Deciding what to change is the work, not the preamble. Agencies who give this away recover the cost inside production — which is where the incentive to keep scope honest disappears.
That sentence is about accounting before it is about anyone’s character. The decision step takes real time from experienced people. If it is not billed, that time still has to be covered from somewhere in the same business — and the sentence above names where it gets covered from.
This is not a hidden arrangement. It is the standard shape of a free proposal in any industry that sells production, and both sides usually understand it: the unpaid work is an investment in winning the contract, and the contract is where it comes back. Nothing about that requires anyone to behave badly.
What it does mean is that the answer “you need less than you think” is being drafted by the party who would be paid for more. Not falsified — drafted. The person writing it may believe every word of it.
And the counterweight is weaker than it looks. The case for changing less has to be argued by the side that knows the market least well — which is the side that went looking for advice, because that is what it lacked. Being suspicious of the proposal does not supply the knowledge needed to shorten it.
After the contract, the question changes shape
Suppose the decision is deferred. The scope is agreed, the contract signed, production begins, and somewhere in the third week it becomes clear that a part of the plan is not right for the market after all.
The conclusion available at that point is not the same conclusion. Before the contract, “change less” is an outcome — the result of the step doing its job. After the contract, the identical words arrive as a change request, and a change request means something has gone wrong. Someone scoped it incorrectly. Someone approved the wrong plan. There is a schedule built on the assumption, and people are working to it.
The difference is not that anyone’s motives have shifted. It is that the decision has moved seats. In the first position it is the deliverable. In the second it is a defect report filed against a plan that a room full of people has already been paid to produce and has already defended internally.
Reaching the same conclusion now requires somebody to volunteer to be the person who says the plan was wrong. That is a different act, with a different cost attached, and it is being asked of people who are mid-delivery.
Keeping “stop” available is a design choice
If the difficulty is structural, the remedy is structural too, and it is duller than it sounds: make the negative conclusion one of the named outputs of the step, so that reaching it does not require anyone to interrupt.
Our own digital pilot is specified that way. Among the six things it names as what you get:
Requirements and explicit non-requirements.
A decision document: build, revise, or stop.
The second of those is worth reading twice. A negative outcome is not an unstated possibility hovering behind the program; it is one of the named things the program hands over.
That quotation establishes a design, not a record. It says where the word “stop” sits in the definition of the deliverable. It does not tell you how often the pilots end there, and you should not read it as though it did.
But the design property is the transferable part, and it does not belong to us. Where a decision step lists its outputs as a plan and a scope, “stop” can only arrive as an objection, and objections need a person willing to make one. Where “stop” is named among them, reaching it is somebody completing the assignment. The same conclusion is an act of dissent in one case and a filled-in field in the other.
This is worth setting up whoever runs the step, and it costs a sentence in a brief.
We are inside this, not outside it
Charging for the decision removes one tilt. It does not make us neutral, and it would be convenient for us if you assumed otherwise.
Our own page names, among what a readiness sprint leaves you with, a description of the work that comes after it:
A phased roadmap with indicative budgets. A written scope for whatever comes next.
That is a document describing work we could then be contracted to do. Being paid for the decision means we do not need to recover its cost inside production. It does not make us indifferent to the size of what the decision recommends, and no arrangement we could describe on a web page would make us indifferent to that.
So apply the same test to us that this article is asking you to apply to a free proposal. The test is not whether the person deciding is being paid. It is whether their next several months look the same whichever way the decision goes.
Ours do not.
What we can offer against that is a step where the negative answer is one of the named outputs rather than an interruption, and an early statement of the parts we do not take on at all — the services page names those before a first meeting rather than after a contract. Neither of those makes us disinterested. They make the tilt visible enough to argue with.
The gap between deciding and contracting
The practical form of all this is a date.
Somewhere in the plan there should be a point at which the decision is finished and nothing has been commissioned — where “less than we assumed” is still available as an answer without anything having to be unwound. The interval does not have to be long, and it does not have to be bought. What it has to do is exist, and be held by someone whose work looks the same whichever way it resolves.
That person does not have to come from outside. A Japan lead whose remit is the market rather than the launch. A brand owner in another region whose year does not change based on how much gets rebuilt here. A finance or legal reviewer with no production to win. Where such a person exists, has the context, and has the time, none of this is a problem you need to buy your way out of, and a paid decision program would be duplicating a job that is already being done.
Where the interval is zero — where the meeting that settles what to change is the same meeting that scopes the work to change it — the answer has been shaped by who happened to be in the room, however carefully the people in it behaved.
None of this is specific to Japan. The sequence works the same way for any market a brand has not operated in. What Japan supplies is the condition that makes the sequence bite: enough unfamiliarity that the decision genuinely cannot be made from the home office, and enough distance that the people who could correct it later are not the ones in the room.