Your Japan team's conversion rates don't match HQ's? Start with the denominator

Your global dashboard says Japan’s lead-to-opportunity rate is far below every other region, and headquarters wants to know why the Japan team is underperforming. Before you accept that read, check the denominator.

When Japan’s conversion numbers won’t reconcile with HQ’s, the cause is rarely the CRM or the team. It is that “lead” means a different thing in Tokyo than it does in the global model.

Align what you count, and as of when, before you compare.

Why the same conversion rate produces different numbers

When one metric moves depending on who reports it, suspect the counting method, not the effort behind it.

A conversion rate is a simple division: opportunities divided by leads. But that denominator, “lead,” often points to different objects in different places.

Marketing counts everyone who downloaded a whitepaper. Inside sales counts only the ones they reached by phone.

The account executive counts only deals with a visible budget and decision-maker.

Run the same “conversion rate” over three different denominators and the numbers will never agree.

This is not a Japan-specific failure, but it compounds when a Japanese subsidiary inherits a lifecycle model built at a US headquarters.

HubSpot Japan’s 2024 survey (fielded by Macromill, 1,545 seller-side respondents at companies of 51 to 5,000 employees; note that HubSpot sells a CRM, so read its own survey with that in mind) found that 79% of companies report some difficulty using their data, with 28.1% citing that data within the sales organization is not properly managed and 24.4% citing that data does not flow between departments.

Misaligned numbers are the default state, not the exception.

Align the denominator on one shared map first

The first move is not to automate the reporting. It is to put the definitions onto one page.

Draw the path from lead to closed deal as a single horizontal map, and write, in words, the entry condition for each stage.

The stage marketing hands over (fix an attribute threshold and a behavior threshold together, not “high intent”). The stage inside sales takes on (write it as an observable behavior, so a new hire judges it the same way).

The stage that becomes a registered opportunity (define it as budget, authority, need, and timing all present).

Once this one page exists, marketing and the Japan sales team can answer “where is that deal right now” by looking at the same picture. The meeting shifts from trading impressions to confirming positions.

Precise stage definitions also let you decompose the conversion rate itself.

The standard document behind Revenue Architecture, Winning by Design’s “The Bowtie: A Proposed Standard” (v1.0, benchmark period 2016 to 2022, n=868), breaks the traditional lump-sum funnel into separate, defined conversion steps: the move from MQL to SQL is the product of two consecutive rates, and opportunity-to-close is the product of two more.

Instead of chasing one large number, you see which specific step is leaking. That decomposition is only possible because each stage was defined first.

Always fix “as of when”

Even with an aligned denominator, the number moves again if the cutoff date differs between people.

The “June conversion rate” is one number if the denominator is leads created in June, and a different number if it is the originating leads of deals that converted in June, counted backward.

One is a current-month cohort; the other is a close-date lookback.

Neither is more correct, but every region and every function has to pick the same one.

Before you measure, write in a single line “as of when, and which population” and freeze it. This becomes the foundation for measuring the full revenue path end to end, which is what Japan GTM built on Revenue Architecture depends on.

There is no public Japan conversion benchmark to anchor to

With no external yardstick, internal consistency is the only reliable standard you have.

Overseas there are large-scale benchmarks such as Ebsta and Pavilion’s “2024 B2B Sales Benchmarks” (n=4.2 million opportunities, non-Japanese companies).

For Japan, no public benchmark discloses the denominator and method behind lead-to-opportunity and opportunity-to-win rates. So a Japan subsidiary cannot look up whether “45%” is high or low for the market.

That makes it doubly important to align your own definitions and cutoff dates first, so you can at least compare Japan to itself over time and judge whether a change came from the work or from the counting.

The absence of an external comparison raises the value of internal consistency, rather than lowering it.

A conversion rate is decided by its denominator. Before comparing Japan to global, align the definition of what counts, and as of when, on one shared map.

Common mistakes

• Building the dashboard before agreeing on the definitions. Automate on a misaligned denominator and you produce a clean chart of the wrong number every day. Aligning one page comes first.
• Treating “lead” as a single word. “High intent” cannot be operated. Pair an attribute condition with a behavior condition so a different person reaches the same judgment.
• Comparing Japan to HQ without a shared cutoff. If the global model counts by created-date and the Japan team counts by close-date lookback, the regional gap you are escalating may be a difference in arithmetic, not performance.

Related reading

After the denominator is aligned, the numbers still need a rule attached. Write the threshold before adding another dashboard covers that step.

Other notes on the same problem are collected under Revenue data.