Japan headcount: split the deal days before you hire more

Planning season. The Japan country manager asks for two more account executives.

“Our cycles run three months. Each rep is carrying as many deals as they can.”

That is probably true. Pipeline is up, win rate is flat, the team is busy. The numbers and the people both point to hiring.

What none of it tells you is whether more reps would make those four months shorter.

The same 90 days can call for opposite investments

Say the average Japan cycle is 90 days.

If the rep was the one who owed the next move for 70 of those days and the buyer for 20, your side is the bottleneck. Proposals are queuing, calendars are full. More capacity helps.

If the rep owed the next move for 20 days and the other 70 were the buyer’s approval document moving between a department head, IT, procurement and an executive, more reps barely move the number.

What helps there is material the buyer can take into their own approval process: an ROI model they can fill in with their numbers, a one-page summary for their manager, a security review started early, a conversation with whoever signs. The investment goes to buyer enablement, not hiring.

This is a point Consilegy’s founder, Seiko Yamaguchi, has made in her newsletter. The problem is that most CRMs do not hold the data to split those 90 days. They hold stages and dates. They do not record who was waiting on whom.

Without that split, headcount usually wins, because it is the easiest request to understand. Next year, the same meeting has the same conversation.

Some of your own waiting will not shrink with more reps

From here on, this is our recommendation, not something the newsletter or the data shows.

Split the seller side once more: days spent waiting for a rep to have time, and days spent waiting on your own company’s approvals. A manager’s discount approval, a quote that needs sign-off, legal’s answer on a contract redline, your security team’s answers to the buyer’s questionnaire.

For a Japan team, many of those approvals sit at HQ. If your deal desk and legal team work in another time zone, a question asked in the Tokyo morning waits for HQ to start its day, so the answer lands the next Tokyo morning, and every question back adds another day. Hiring a third AE in Tokyo does not change that. More reps can even lengthen it, because more deals reach the same approvers.

On a professional message board, one poster wrote that everyone talks about efficiency but no one measures how much revenue is lost while deals wait on internal approvals. A reply described customers left waiting on discounts and implementation, and the impact showing up in Q4 numbers. That is one company, outside Japan. It is still a reason to count this bucket on its own.

So sort each idle day into one of three:

• Waiting on the rep: your rep owed the next move and needed no one else to decide or answer
• Waiting on your approvals: your side owed the next move, but someone other than the rep had to decide or answer (manager, deal desk, legal, security), in Tokyo or at HQ
• Waiting on the buyer: the buyer owed the next move, and the text shows someone inside the buyer was moving it (“I’ll take this to my director,” “it goes to next month’s budget meeting”)

A gap where the buyer simply went quiet, with no sign of anything moving inside, does not go in the third bucket. Count it separately as stalled. As we wrote in “Stalled deals in Japan: is the deal dead or just queued?,” stalled and queued deals call for opposite responses.

For a security questionnaire, the time from receiving it to sending your answers is waiting on your approvals. The buyer’s review after you send it is waiting on the buyer.

On large Japanese purchases, the buyer’s process alone takes months

Published research gives an order of magnitude for the buyer’s side, nothing more.

A joint survey by IDEATECH and Hiroyasu Kitagawa of Demagen Research Institute (published April 2026, fielded March 23 to 26, 2026, n=307) asked people involved in B2B purchases worth ¥3 million or more a year, with at least two people involved. The most common time from starting to evaluate to choosing a vendor was three to five months (34.9%), and 41.4% took six months or longer. More than 80% had two or more approval levels.

IDEATECH sells survey-based PR, and this was an online panel survey, so read it as a rough scale for large deals, not as a benchmark. It also did not ask who was waiting during that time. It says nothing on whether Japanese buyers are slower than buyers elsewhere. Your own split has to come from your own deals.

Open 20 closed Japan deals and mark the idle days

If the headcount decision is due this quarter, reading closed deals is faster than changing your CRM. It takes time, because someone has to read each deal, but it changes no fields and no automation.

Have RevOps or sales operations do the sorting, not the Japan sales team that asked for headcount. The team making the request will tend to see gaps as rep waiting. Pair that person with a Japanese reader if the notes and emails are in Japanese, and record who did the count so the approver can ask.

1. Gather the records. Beyond the deal’s emails, meetings and notes, pull your own approval trail: approval history if you use approval processes in Salesforce, quote or CPQ approvals, deal desk tickets, the dates legal and security received and answered requests, and the discount threads in internal chat. Approvals rarely show up in the deal’s activity, so skipping this step makes the second bucket look small. (HubSpot quote approvals are limited to certain subscriptions; check yours.)
2. Pick the deals. List this fiscal year’s closed Japan deals, won and lost, in order of close date, and take every nth one until you have 20. Won deals alone skew toward buyers who got their approval through. If you closed fewer than 20, use all of them.
3. Sort the gaps. Open each deal’s records in date order. For each gap between activities, assign it to one of the three buckets or to stalled. If both sides were working at once, assign it to whichever side owed the next reply.
4. Mark what you cannot tell. If the text does not tell you, mark the gap “unknown.” Do not guess. Ask the deal owner about the unknown rows only.
5. Stop lost deals at the last real reply. For lost deals, count up to the buyer’s last substantive reply. The days after that, until the deal was marked lost, go to stalled.
6. Add it up. Total the days in the three buckets, stalled and unknown across all 20 deals. Write the sorting rules on the sheet and use the same rules next time.

If the new hires would cover one segment, such as enterprise accounts, count only that segment. Approval levels differ by deal size, and mixing segments blurs the answer.

This is an estimate, not a measurement. It is enough to see where the days went.

Put next year’s budget where the longest wait is

If waiting on the rep is longest, you have a case for headcount. Check first whether the waiting is concentrated on one or two reps. If some are overloaded and others are not, territory and assignment come before hiring.

The hire itself is not easy to make. In JETRO’s FY2024 survey of foreign-affiliated companies in Japan (published March 25, 2025, 1,427 valid responses), 42.4% said they were struggling to secure the people they need, and among them sales and marketing talent was the hardest category, at 57.3% (n=583, multiple response). The survey does not say how long a hire takes. It does say this is the role foreign companies in Japan find hardest to secure, which makes a hire against the wrong bucket an expensive mistake.

If waiting on your approvals is longest, hold the headcount. Name the approvers for discounts, quotes and Japan contract redlines, and agree how fast they answer. Decide in advance what the Japan team can approve locally. Shortening that path works faster than adding people.

If waiting on the buyer is longest, fund buyer enablement. Give the buyer material they can carry into their own approval process before they ask for it. We wrote about the ROI piece in “Give Japanese Buyers an ROI Model They Can Edit Themselves.” If that material never went out because reps or presales had no time to build it, those days will already show up as waiting on the rep.

If stalled days are large, the question before headcount is which deals the team chooses to chase. How to handle stalled deals is in the stalled-or-queued article linked above.

If the unknown days exceed the gap between the top two buckets, you cannot rank them yet, because the unknown days could flip the order. The first fix is the record, not the hire. At minimum, Japan’s activity notes need to say who owes the next move.

Check the same way after you decide

After the decision, count again with the same rules, but change which gaps you count.

If you pick deals by close date, most of their gaps started before the change. Count only gaps that began after the change started, including gaps in deals that are still open. For buyers on an April-to-March fiscal year, approvals can bunch up before year end, so compare against the same months a year earlier. If you hired, leave out the new rep’s ramp period.

If the bucket you funded got shorter, the call was right. If it has not moved after two quarters, revisit it. A new hire with no drop in rep waiting means the constraint was somewhere else.

Compare only the days in the same bucket. Win rate and pipeline move for other reasons too, so they are not evidence for this decision.

What we decided against

Adding “In approval” or “Legal review” stages. That mixes deal progress with waiting in one field and breaks comparisons with past stage reports. The stalled-or-queued article above keeps that distinction in fields instead of stages.

Calculating the split automatically from activity dates. Dates and email direction tell you who spoke last, not who owed the next move or whether anything was moving inside the buyer. That is only in the text.

Before approving more Japan headcount, sort the idle days in this year’s closed Japan deals into waiting on the rep, waiting on your own approvals, and waiting on the buyer. Hiring directly shortens only the first, and can lengthen the second.

Before the planning meeting, open 20 closed deals. Whether the headcount is approved or not, you will be able to explain why on one page.

Related reading

Why hiring is often the first constraint when entering Japan is covered in “Don’t leave hiring until after translation and pricing in Japan.”

Other notes on the same problem are collected under Data & KPIs.