Japan pipeline grew, revenue flat: the stall is downstream

Japan came in under plan, so the instruction was the obvious one: more meetings per rep. The next quarterly review looked encouraging. Meeting count up, pipeline value at a record, activity compliance green. Bookings within a rounding error of the quarter before.

HQ has two standard readings for that shape. The local reps are not good enough, or Japanese demand is softer than the business case assumed. The first leads to a performance conversation, the second to discount authority. Neither touches what happened, which is that every added deal queued at the same point and nobody was counting that queue.

The added deals joined one queue, and it was not on your side of the table

The stall sits after the proposal, while the buyer works the decision through their own organisation.

Before the push, nobody was counting how many open Japan deals were sitting there. What was counted was meeting volume, pipeline value and win rate. All three are worth having. None of them answers the only question that matters here: right now, how many deals are waiting, and whose desk are they on.

So the added deals entered at the top, settled in the same place, and the rollup showed growth all the way through the quarter. What arrived at the end were two lines that do not belong next to each other: record pipeline, flat bookings.

Queue is discovered after you add to it. Seen beforehand, it changes whether you add at all.

What the deal is waiting for is a Japanese executive, not your rep

There is Japanese data on where the wait happens, and the ranking is the whole argument.

IDEATECH, with Hiroyasu Kitagawa of Demagen Research Institute, surveyed 330 people between 25 and 26 May 2026 who had been involved, with at least one colleague, in buying, replacing or renewing B2B software, SaaS or IT infrastructure and security products at annual contract values of 5 million yen or more. Of them, 73.5% said the evaluation ran longer than originally scheduled (n=325 for that question).

Those respondents were then asked why (n=239, multiple answers). Top of the list: “it took time for management to reach a decision,” 63.6%. Then information gathering 45.2%, more departments and people joining 38.1%, budget approval 31.0%. “The vendors did not give us enough information” comes in at 16.7%, near the bottom.

Read that ranking against the instruction your Japan team received. What dominates the delay is time spent waiting for the buyer’s own approvers to reach the item. A shortage of vendor contact sits four places below it. The constraint is downstream of everything a meeting can influence, and it is inside an organisation you do not manage.

The caveat belongs in the open: IDEATECH sells research-led PR, so this survey doubles as a demonstration of its own service, and it is an online panel. Treat the percentages as order of magnitude and ranking. The gap between 63.6% and 16.7% is not a gap that reverses.

One more number sets the scale of the circuit the deal is moving through. In the same survey, 58.5% of deals involved seven or more people in the decision, and the specialist functions taking part were IT 42.2%, finance and accounting 38.2%, security 37.2%, legal 33.2%, procurement 32.0% (n=325, multiple answers). Adding a rep does not add capacity to any of them.

More meetings crowd out the work that actually clears the queue

The sharper cost is not the wasted activity. It is what the activity displaces.

Clearing a Japanese deal through its approval circuit is unglamorous, per-deal work. Writing text the internal champion can paste directly into the 稟議 (ringi) document that carries the decision. Producing an operating-load estimate the IT department will accept. Answering finance and legal before they ask. Creating an occasion where the approver is actually in the room. None of it scales, none of it is predictable in duration, and none of it has a field in your global CRM.

Meetings have all three. They sit in a calendar, they count, they appear in the weekly rollup you read.

When one person’s week is contested by work that is recorded and work that is not, the unrecorded work loses. Anyone would make the same trade when one half of the job is visible to headquarters and the other half is not. Calling it a discipline problem in your Japan office gets the causality backwards.

So the quarter you push activity is the quarter per-deal approval support drops. You raised throughput at the front of the process and slowed the stage that was already the constraint.

That a single stage governs the whole rate is not a new idea. What makes it awkward in Japan is that the governing stage sits outside your company and leaves no trace in your CRM, so the decision to add volume never shows up in any report as the thing that slowed the process down. What shows up is pipeline growth and flat bookings, which reads like a Japan problem.

The buyer has a capacity ceiling too. In the same survey, 13.4% named “our own staff were too busy to make time for the evaluation” as a cause of delay (n=239). The queue you are feeding is congested on the other side as well.

Before approving more activity, ask how many open Japan deals have a known stall point

Change the order of the decision. Before signing off more meetings or more headcount for Japan, count how many open deals your team can state a stall point for.

If most cannot, the increase is not yet a decision you can make, because you will not be able to tell afterwards whether what you bought was bookings or queue.

A rough read takes half an hour and no new tooling. Open ten recent Japan losses and ten open deals untouched for more than three months. For each, look only at the last activity: did your side act last, or was your side waiting on the buyer.

Where your side acted last, the delay is yours. Proposal turnaround, internal quote approval, or simply nobody having time. That distribution is the one where more meetings and more headcount convert, and the instruction was right.

Where the buyer’s response is what is outstanding, the same spend extends the queue. What converts instead is material the buyer can carry through their own organisation. Same budget, opposite destination: buyer-enablement content rather than sales capacity.

This is a direction rather than a measurement, and it is enough to decide with. Making it repeatable is where CRM design comes in, and it starts with holding stall location as a field rather than reconstructing it from activity history each quarter.

When losses cluster at the shortlist and the stalls are yours, add the meetings

Two conditions. Losses concentrate in “never made the shortlist” or “never reached anyone,” and the stall points on open deals are known and mostly on your side. With both true, activity converts and the instruction needs no revision.

One option not taken: capping the number of open deals a rep may carry at once. It does reduce queue, but once the cap binds, someone has to decline new conversations, and that judgment lands on a rep who cannot yet see value or probability. What survives a cap is the easy deal. Making stall location visible gets the same result without handing the local team a decision they are not equipped to make, and without HQ discovering the cap in a QBR.

The full model for designing revenue around the buyer’s internal process, rather than around seller activity, sits in Japan Market GTM and Messaging.

Activity converts to revenue only where deals are stalled on your side of the table. When the stall sits inside the Japanese buyer’s approval circuit, the meetings you add become queue rather than bookings.

This week, ask your Japan team for one column against every open deal: is the ball on our side or theirs. Do not ask for confidence, and do not ask in a forecast call, where the answer is social. How many deals come back with an answer tells you whether next year’s Japan budget belongs in sales capacity or in buyer enablement.

Related reading

Stall location only becomes usable when it lives in a column rather than in a rep’s head, and the same column changes which deals belong in this quarter’s number. How to build that second axis, and how to write the exclusion rule before the quarter closes, is in Forecast your Japan pipeline on approval progress, not rep confidence. This note decides whether to add volume; that one decides what to count.

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