The fix isn't a new dataset -- it's a new mandatory pairing rule for the datasets that already exist.

Call it the Paired Metric Standard: any institution that publishes an account-ownership statistic as a financial-inclusion indicator (Global Findex, national financial-inclusion strategies, telco regulatory filings, donor program reports) must publish, in the same table, on the same page, at the same level of prominence, a 90-day active-use rate for the same population.

Ownership without a paired use rate stops counting as a valid inclusion indicator for donor conditionality or regulatory scorecards.

Mechanically, this works because the underlying data mostly already exists -- it's a disclosure and formatting requirement, not a new measurement infrastructure.

GSMA operators already calculate monthly active accounts internally for their own commercial purposes; central banks already receive transaction-volume reporting from licensed mobile money issuers for prudential supervision.

What's missing is a rule forcing that number into the same public artifact as the registration number, and a definition standard (a transaction-based test: at least one person-initiated transaction, not an incoming-only or agent-initiated float top-up, within 90 days) so operators can't quietly define 'active' down to inflate the paired figure.

Implementation runs through three channels simultaneously: statistical (getting Findex and GSMA to adopt a shared active-use definition and publish it as a co-headline figure, not an appendix table), regulatory (central banks in high-mobile-money markets like Kenya, Uganda, Ghana, and Tanzania requiring licensed issuers to report and disclose active-use rates as a license condition), and financial (multilateral and bilateral donors -- World Bank, USAID, FCDO, Gates Foundation-funded inclusion programs -- rewriting grant and loan conditionality to score programs on active-use improvement rather than registration counts).

None of the three requires the others to move first, which matters because past inclusion-metric reform efforts have stalled waiting for a single coordinating body to act.

The mechanism that actually changes behavior is conditionality: once donor disbursements and program renewals are scored against active-use rates instead of registration counts, telcos and NGO implementers lose the incentive to run registration drives that never convert to habitual use, and gain an incentive to fix the frictions (fees, agent liquidity, fraud response time) that are currently suppressing use.

That's a testable, falsifiable claim, and the pilot phase below is designed specifically to test it before asking anyone to commit at scale.