Japan deal cycle length can be estimated at the first meeting

A Japan deal is in its fourth month. HQ looks at the global average cycle, sees a gap, and asks what the Japan team is doing wrong.

The follow-up was prompt and the demo went well, so the team has no answer that sounds convincing.

Often there is one, and much of it was visible at the first meeting. That is our hypothesis, and how to test it in your own data is at the end.

The buyer is already well into the process when your rep first sits down

A joint survey by IDEATECH and Hiroyasu Kitagawa (Demand Gen Research Institute), fielded 23-26 March 2026 and published 16 April, asked 307 people in Japan who had taken part, with at least one other person, in buying or renewing B2B products costing JPY 3 million or more.

By the time they began serious exchanges with a vendor’s sales rep, the purchase process was on average about 40% complete.

About 90% had already shortlisted several vendors before talking to sales.

This is a panel survey run by a firm that sells survey PR, and it covers only large, multi-person purchases, so read 40% as an order of magnitude.

The direction is consistent with what Japan teams report: someone inside the buyer has already named the problem, listed candidates and, sometimes, started looking for budget. Your rep meets that person, partway through.

But 40% is an average. The survey says nothing about how that figure varies from deal to deal, or how the variation relates to cycle length.

The working assumption from here on is that deals differ in how much was already done, and that a deal arriving with most of the internal work finished tends to run short while one arriving with none of it tends to run long.

The test for that assumption comes at the end.

The floor on cycle length is the number of rooms the deal must pass through

In the same survey, the most common consideration period was three to five months, and 63.6% of buyers fell between three and eight. Under one month was 2.3%.

Short cycles are rare in Japan not because reps are slow, but because more than 80% of the companies surveyed have two or more approval stages. The number of rooms sets the floor.

What separates a deal near the floor from one that runs well past the ceiling is rarely rep behaviour. When you line up the long ones, at the first meeting one of three things was missing.

Three things to record at the first meeting

Who brought the deal in. A manager who found the problem and went looking has a reason to move it internally. Someone told to “look into options” may finish when the report is filed.

Whether a budget line exists. If this year’s budget table already carries a line, approval signs off a row that is there.

If not, the champion first has to create the reason for a new approval, and that takes months regardless of the amount.

Where it is approved next, and when that forum meets. A department meeting, a management committee, an IT review.

A deal with a date can be timed. A deal without one has no destination yet.

Two of the three overlap with the questions in The first Japan sales call: check what the buyer has already decided.

Who brought the deal in comes out when you ask question two (what each person is responsible for), and the next forum comes out with question four (whose approval is needed).

The only question this note adds to the first meeting is the budget line.

None of these change with sales skill. What changes is whether they are asked and recorded at the first meeting.

Deals run without them tend to surface “actually, we still need to talk internally” at meeting two or three, and the six months start there.

Place long deals as long deals from the start

Sort the first meeting into groups. If the person who brought it has a reason to move it, there is a budget line, and the next forum is visible, the deal is on the short side.

If two or more of the three are unknown, it is on the long side. If only one is unknown, ask that one at the next contact and let the answer move the deal to one side or the other.

Then change how the deal is placed, not how the rep behaves.

Putting a long-side deal in the same “in progress” column as a short-side deal means it is chased every month as stalled when it was never going to be quick.

Place it as a deal waiting for a forum to be set, and act when the forum becomes visible. The forecast error and the rep fatigue both shrink.

Do this in the forecast, not by adding a stage. Long-side deals carry a lower weight and stay out of commit, and a saved view lists every deal with no next forum date.

Waiting does not mean leaving the deal alone. Once a month the rep checks with the buyer whether the next forum has been scheduled. When a date goes in, the deal goes back to being handled like any other.

Make the owner of the forecast call responsible for reviewing that view. Chasing a deal in a panic after it has slipped costs far more rep time than tracking one date from the start.

Global data points the same way, with a caveat. Ebsta’s 2025 GTM Benchmarks, covering 655,000 deals across 387 companies and not drawn from Japan, show win rates falling with the length of a slip: 18% after a one-week slip, 13% after one month, 8% after three, 5% after six, and 3% beyond six.

The figures do not transfer to Japan. They do support spotting slow deals early over trying to rescue them late.

Two approaches considered and rejected

Penalising reps for long cycles, or auto-closing deals past a threshold.

Cycle length in Japan measures the buyer’s internal path, not rep neglect. Auto-closing makes the reps with the most long-side deals look worst and teaches them to stop taking those deals.

Adding more stages to track it finely. Stages describe the seller’s progress. Mixing the buyer’s internal progress into the same column makes both unreadable.

Keep the three points as fields separate from stage

Add three deal properties: the role of the person who brought it, budget line (yes, no, unknown), and the date of the next forum.

Which approval stage that forum is goes in the approval-progress field from the five-questions note, so this field holds only the date. The same approval held in two fields means one of them goes stale.

Keep each to three options or fewer so a rep can fill them in a minute after the first meeting.

Leave unknown as a valid answer and do not make the fields mandatory, because forcing an answer invites a guess entered as fact.

After about three months, check whether deals with more unknowns ran longer in your own data. This is where the hypothesis in this note gets tested in your business.

If they did run longer, keep placing them this way, and you have an internal benchmark for Japan that no global average can give you. If there was no difference, look at which of the three points is not doing any work.

The wider model, forecasting from the buyer’s internal process rather than from seller activity, is in Japan Market GTM and Messaging.

In Japan, deal length depends less on what your rep does than on how far the buyer’s organisation had already moved at the first meeting. Record who brought the deal in, whether a budget line exists, and where it must be approved next, estimate the cycle from day one, and check it against your own data after three months.

This week, open five live Japan deals and check whether the budget line and the next forum date can be read from the CRM. The number you cannot read is the number of deals with no basis yet for a cycle estimate.

Common mistakes

Judging a Japan deal against the global average cycle. The floor is set by the number of approval rooms, not by rep pace.

Reading a slow deal as a rep problem. Check which of the three first-meeting points were missing before looking at activity.

Making the three fields mandatory. A forced answer becomes a guess recorded as a fact, and the forecast inherits it.

Mixing the buyer’s progress into the stage column. Keep the seller’s stage and the buyer’s internal position as separate fields.

Related reading

Once a buyer says they will take it back, how to learn where it went is in In Japan, ‘we’ll take it back’ is a handoff, not an answer. That note covers what to ask after the meeting; this one covers what to record at the start.

Other notes on the same problem are collected under Sales process.