Your Japan team wants its own pipeline. Give them four fields instead.

The request arrives politely, usually in the second quarter after launch. The Japan team would like its own pipeline. The global stages, they explain, do not fit how deals work here.

Headquarters hears this as a localisation request and either approves it to keep the region happy, or refuses it to protect global reporting. Both answers are wrong, because both accept the framing.

Ask one question first: do deals in Japan exit each stage on different conditions, or do they sit in the same stages for longer?

Almost always, it is the second one

Global pipelines are named after seller actions. Demo delivered. Proposal sent. Negotiation. Those names work in markets where the deal moves when the seller moves.

In Japan the deal moves when a document moves. Research by IDEATECH and Hiroyasu Kitagawa, published 16 April 2026, surveying 307 people in Japan who took part in a B2B purchase of at least ¥3 million alongside at least one colleague, found that 60.9% of purchases required two approval stages and more than 80% required two or more. Consideration ran three to eight months in 63.6% of cases, with roughly 40% taking six months or more.

So a Japanese deal parks at “Proposal sent” for four months. Nothing in your CRM changes, because in your CRM nothing has happened. The proposal was already sent.

That is not a stage-definition problem. The exit criteria are identical: the customer says yes. What is missing is any representation of the three months between the proposal and the yes.

A separate pipeline does not add that representation. It just moves the blank space somewhere headquarters no longer looks.

What splitting actually costs you

The moment Japan has its own pipeline, you lose the ability to answer the only question that matters about the region.

Headquarters will keep saying Japan is slow. With one pipeline, that claim is testable: you can compare stage-by-stage duration against other regions and see whether Japan is slower in discovery, slower in evaluation, or slower only after the proposal. Each of those has a different fix, and two of them are your fault, not the market’s.

With two pipelines and different stage names, the claim becomes unfalsifiable. “Japan is slow” turns into a permanent line in the board deck, and no one can show which part is slow. The regional plan then gets built on a description instead of a diagnosis.

Four fields, not a second pipeline

Add these to the deal record, globally or as Japan-visible fields. Each one turns a month of silence into a value someone can filter on.

1. Approval stage reached. Not your stage. Theirs. Section head, division head, executive, board. Blank means your rep has not asked, which is itself the finding.
2. Number of approvers remaining. With more than 80% of purchases requiring two or more stages, this is the single best predictor of remaining time you can collect in one question.
3. Customer decision deadline. The date driven by the buyer’s budget cycle and their internal submission cut-off, not by your fiscal quarter. Reps almost never ask for this, and it is usually the date that decides the outcome.
4. Internal document handed over: yes or no. Whether your team has given the champion Japanese-language text they can paste into their own approval document.

Four fields cost an afternoon of configuration. A second pipeline costs you the comparison forever.

When splitting is correct

There is one case where a second pipeline in Japan is the right call: partner and reseller deals, where the exit criteria genuinely differ because the partner, not your rep, controls the steps.

Even then, split the pipeline but keep two things common with the global one: the closed-won and closed-lost definitions, and the loss-reason picklist. If those two are shared, five pipelines still roll up into one forecast. If they are not, one pipeline will not save you.

Write the exit condition of each stage as a single sentence, for the global pipeline and for the proposed Japan pipeline. Who does what, and then the deal advances.

Split a pipeline only when the exit criteria differ. In Japan they usually do not. What differs is that the deal advances through approvals your stages do not model, and that belongs in fields, not in a second pipeline.

Related reading

The same question comes up when headquarters decides how to enter Japan at all. distributor or direct: choosing your Japan entry model covers what stops reaching headquarters when a partner owns the account.

Other notes on the same problem are collected under CRM adoption.