Japan cross-sell: choose accounts by need, not usage score

Published Nov 2, 2026

“We are one customer. Why do we have to explain this to your company twice?”

A multi-product SaaS team in Japan can hear this a few months after the first rollout. Product A goes live, the implementation lead rolls off, and the handoff note is three lines long. Then a product B specialist, often working to an HQ attach target, asks for a meeting and starts from the beginning.

The cross-sell sounds like a push, and a better talk track will not fully fix that. The account was chosen, and the meeting booked, by people who do not know how the customer works.

A usage score shows product A is adopted, not that product B is needed

The typical HQ play ranks Japan accounts by usage or health score for product A and hands the top of the list to the specialists. High usage does get you a warmer conversation.

So usage is a necessary condition, but not a sufficient one. It is evidence about product A. A company that uses a time-tracking product every day may have no one in-house managing hiring, and then a recruiting product has no one to talk to. After a few misses, the Japan account owner tends to protect the relationship by being careful about the next introduction.

Holding on to existing revenue is not a given either. In Cloud Circus’s Fullstar customer success survey for 2025 (August 2025, 200 customer success and support staff at Japanese B2B information and communications companies), average net revenue retention was 102.1%, and 38.4% of responses were below 100%.

These are self-reported figures that include churn and downgrades, the survey comes from a vendor of customer success tools, and it says nothing about how cross-sell is run. Use it only for scale: nearly four in ten said they were not holding existing revenue.

The grid shows that the gap is process knowledge, not contract data

Build one table: accounts down the side, products across the top. This is a whitespace analysis, except that the non-contracted cells are split in four.

Process exists: the customer does this work in-house
Outsourced: the work exists but sits with a vendor
No process: the work does not exist
Unknown: nobody on our side knows

A standard whitespace view treats every non-contracted cell as an opportunity. Those cells mix accounts with a reason to buy and accounts with none.

Fill it in for ten Japan accounts. Most non-contracted cells will probably read Unknown. Before you have a pitching problem, you have a knowledge problem.

Fill Unknown from the rollout, and get introduced to the other departments

Nobody hears more about how a customer works than the implementation lead on the first product. Then it disappears into three lines.

Ask each product owner to write one question for their column, such as “Do you manage hiring in-house or through an agency?” Publish it in Japanese with the English alongside. At implementation sign-off, the lead answers the other products’ questions and records three things:

whether the process exists
which department and person own it
what would make it need attention (the trigger)

Record only what came up during the rollout, and let Unknown be a valid answer. If you require people to know, you get plausible No process entries.

That covers work in the same department as product A. Work in other departments does not come up in a rollout, so the current owner uses a regular meeting to ask who looks after it, one column per meeting. The second product may sit with a different department and its own approval path, so that name matters as much as the need.

Add only fields with a known reader and a known moment of use. “What your Japan handoff drops is the reason the contract was approved” covers what protects the current contract. This note is read by the product B specialist.

Pitch when the customer confirms the trigger

Contact every Process exists account and you are back to a ranked list. The trigger is what changes the conversation.

Keep it in the customer’s words: “opening a second site next fiscal year,” “the agency contract ends in March.” Give it a status (not yet, confirmed, gone) and a next check date. Set that date ahead of the customer’s internal review and budget deadline, not on the event itself. Once the agency contract has ended, the decision on what replaces it has been made.

The current owner checks in a regular meeting and marks it confirmed when the customer says it is going ahead. That is when the pitch starts.

Usage keeps a job as an exclusion. Hold the pitch if product A has not been adopted, if a support incident is open, or if the customer is discussing cancellation.

The current owner makes the first contact as an introduction. The specialist reads the row first and does not ask again what is already written. Where the specialist is outside Japan and cannot read the Japanese notes, keep one English line beside the original.

Run thirty accounts in a spreadsheet and compare with the usage list

Do not start by adding CRM fields. Run it in a spreadsheet for the thirty accounts with the nearest renewals. Replace the contracted column in full from contract records each month, and let nobody edit it by hand. When a product churns, set its process cell back to Unknown and keep that account out of pitches for a while. Where a reseller holds the relationship, leave cells Unknown until the reseller agrees to share.

The Japan customer success lead owns the grid. Only “whether the process exists” becomes part of sign-off, agreed with the implementation lead. Owner and trigger are filled in when known.

Keep the usage list running for the quarter. Compare the two on pitches made, share reaching a deal, and share won.

Pitch volume will fall, and an attach target will feel that. The CRO’s decision is to agree now that deal rate and win rate at quarter end settle which list continues. If the usage list wins, keep it. You will have learned how your products get bought.

Tag deals from the grid with type “Existing customer cross-sell” and the introducer, and keep them out of new-logo reports. If an implementation lead’s only target is product A usage, there is no reason to fill in anyone else’s column, so credit the introducer for those deals. When a customer says there is no budget this year, put the deal on hold rather than losing it, and set the next check ahead of their budget round.

If the grid wins and the column habit has stuck, move it into the CRM as a custom object with one record per account and product (check your HubSpot subscription, as custom objects are plan-limited). Even with three products, contract state, process, owner, trigger, status and check date would otherwise be 18 fields. Before moving, review automation that starts at sign-off, automation that runs on company record updates, which system is the master if you run both HubSpot and Salesforce, and whether product B campaigns already reach these accounts.

What we decided against

Scoring accounts from product logs does not work here, because logs describe product A and never show whether another product’s process exists.

Giving one person every product was also rejected. Specialisation is fine; the problem is that the customer’s story vanishes at each team boundary.

A mass survey goes to the contract contact, who cannot answer for other departments. Being introduced to the right person is more reliable.

Usage rank alone does not tell you where to cross-sell in Japan. Keep an account-by-product grid of whether the process exists, who owns it and what would trigger a need, and let the current owner introduce the next product once the customer confirms that trigger.

This week, build the grid by hand for ten Japan accounts with renewals coming up. It will probably look emptier than you expect. Fill that in before you set next quarter’s attach target.

The exclusion checks are faster when deal, renewal and support status already sit at the top of the account. That layout is in “Japan accounts: four lines every team reads before calling.”

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