Personetics

The removal test leaves a transaction list, which is precisely the raw material the platform exists to convert. The process is described in stages rather than asserted: enrichment first, cleaning merchant names and adding logos so the record is legible, then behavioural and predictive models reading income stability, spending patterns and upcoming obligations across more than 150 million active monthly customers, then identification of specific needs such as a predicted cash flow shortfall. Nothing beneath the models constitutes a product.

Control sits with the bank's own business teams rather than the vendor, through no-code tools that let them create and manage contextual insights directly, which is the right construction because those teams carry the conduct responsibility for what reaches a customer. Output is prompts, guidance and offers the customer chooses to act on rather than actions taken on their behalf, with savings journeys automated only once configured. Held at B because no review or approval workflow is described for insights before they reach millions of customers, and nothing states what constrains an automated prompt.

No accuracy, validation, false positive rate or evaluation method is published for models predicting income stability and cash flow shortfalls at this scale. Results described as validated across 150 million customers refer to business outcomes such as customer primacy and reduced churn rather than to whether the predictions themselves are right, and a wrong cash flow prediction produces either an unnecessary credit prompt or a missed warning, neither of which is measured publicly.

Scale and named customers both hold up. The company states hundreds of banks and financial institutions across 30 markets reaching over 150 million customers, and names institutions spanning a top-twenty United States bank, a Canadian big five bank, a major Belgian group, a central European group, a Canadian cooperative, a regional bank and a community savings bank, which evidences the range rather than only the headline.

A major core banking provider embedded the platform inside its own digital banking product in 2026, and the company was named among the world's top 50 fintech companies for 2025 by a broadcaster and research firm jointly.

No boundary statement was located, and one product makes the question concrete. The company describes a capability that deciphers a consumer's external banking relationships and uses that to drive a greater share of wallet, which means data a customer shared through open banking for their own benefit is used to move business away from their other providers. Whether models learn across the hundreds of client institutions, several of which compete directly, is separately unaddressed.

No data protection agreement, retention schedule, subprocessor list or deletion commitment was located, and this is among the largest consumer data footprints in the index. The platform processes transaction-level detail for over 150 million people across 30 markets, and extends beyond the bank's own records through open banking to a customer's external accounts, which is the most complete financial picture any vendor here assembles about ordinary individuals.

No attestation, certification, trust centre or enumerated framework was located. Hundreds of institutions including several systemically significant banks have completed supplier assessment before connecting core transaction data, so the underlying assurance is extensive, and none of it is published for a prospective institution to begin from.

The open banking framework the platform operates through is named specifically, which is more than most vendors here manage, and it matters because consumer data sharing rights are what allow the external account picture to be assembled lawfully. Operating across 30 markets implies handling several such regimes.

Held at B because no other regulator, conduct rule or marketing standard is named, and a platform delivering product offers to 150 million banking customers sits squarely inside financial promotion and fair treatment requirements in most of those markets.

The consumer protective outcomes are specific and some run against the bank's own revenue, which is rare enough to be worth crediting: helping customers avoid overdraft fees, save more effectively and prepare for upcoming expenses, framed as equipping them to avoid problems rather than react to them. The central tension is left unaddressed.

The same engine that predicts a cash flow shortfall so a customer can prepare also prompts the bank to offer a timely loan at that exact moment, which is the point of maximum persuasive advantage over someone about to run short. Nothing describes how those two uses are governed against each other, and no analysis of who receives which kind of prompt is published.

No guarantee, indemnity or correction process was located. The banking customer receives predictions and prompts about their own finances with no described route to understand why they were shown, to correct a mis-enriched transaction that skewed the picture, or to decline the analysis while keeping the account. At 150 million customers the absence of any stated opt-out or explanation mechanism is a substantial gap.

Models are described as proprietary enrichment, behavioural and predictive components, which tells a buyer the capability is built rather than wrapped and stops there. No base model, provider, hosting arrangement or subprocessor is named, and the merchant reference data behind the enrichment step, which determines how accurately transactions are categorised in each of 30 markets, has no stated source.

The substantive integration is embedding inside a major core provider's digital banking product, which places the capability in front of that provider's bank and credit union base without each institution integrating separately, and the platform is described as delivering across every digital touchpoint with development kit options and connections to other providers. Open banking connectivity brings external account data in. Held at B because no core banking, data warehouse or campaign management system is named individually and no developer documentation was located.

No hosting provider, region selection, residency commitment or private deployment option was located. Operating across 30 markets with transaction-level consumer data means several jurisdictions with binding localisation rules are involved, and nothing describes where processing occurs for any of them.

No pricing, packaging or basis of charge was located. A customer states publicly that they surpassed their return on investment, which is an outcome claim rather than a cost disclosure, and for a platform sold to institutions ranging from community banks to global groups the basis of charge determines which of them can adopt it.

Coverage spans 30 markets across North America, Europe, Latin America and Asia Pacific, with named deployments at institutions of very different sizes from community savings banks to global groups, which is the harder thing to evidence than a single large logo.

Functional coverage runs from transaction enrichment and insight generation through contextual offers, savings journeys, custom trackers and integrated marketing offers, reaching customers across digital channels rather than one surface.