Termina

The removal test leaves a transaction file. What the product does is read financial and transactional data at a scale no deal team could process manually, one published case involving more than 120 million transactions from a single target, and convert it into a judgement about the quality of a company's growth measured against benchmarks built from prior deals. Strip the models and there is no scan, no benchmark comparison and no finding, only raw data the client already had access to. The company describes models working alongside operators, so a human service layer sits in the delivery, but the analytical work that constitutes the product is model driven.

The boundary is published in plain terms on every page: the platform relies on subscriber input and facilitates subscriber analysis but does not substitute the subscriber's own independent analysis. That is an explicit non substitution statement rather than an implied one, and for a product whose output informs an investment committee it draws the line in the right place.

The description of models working alongside operators reinforces it, since a human is involved in producing a scan rather than only receiving one. What is absent is anything more granular: no confidence indication on a benchmark comparison, no flag when the ingested data was insufficient for a reliable read, and no account of how a subscriber is told the difference between a finding and an artefact of thin data.

Two genuine transparency artifacts exist and neither is a measurement. The company publishes platform documentation and a glossary of metrics, which for a product whose output is a set of computed measures is the equivalent of publishing methodology, and it lets a user establish what a given figure actually means before acting on it.

What is absent is validation of any kind: no accuracy or error rate, no account of how a benchmark cohort is constructed or how many companies sit behind one, and no confidence exposure on a scan run against limited data. The published portfolio case is a strong anecdote in which the scan caught deterioration standard reporting missed, and it is a single self reported instance rather than evidence of a hit rate.

Two case studies are published and both are unusually specific about mechanism while naming nobody. In the first, an institutional investor in a competitive process with bids due in three weeks had a target scanned within an hour of notification and understood its quality of growth inside 24 hours across more than 120 million transactions, detecting several material issues.

In the second, a scan on a portfolio company seeking follow on funding found that top line revenue growth looked sound in standard reporting while the underlying benchmarks showed serious deterioration in customer acquisition cost, payback period and contract value, with revenue retention dropping sharply at five months, and a principal took that to the board. That is a real and checkable account of what the product found. Against it, no customer is named, no client count or scan volume is published, and no funding for this entity was located.

Two stewardship questions sit here and the company gestures at both without answering either. The first is stated as a selling point: the platform recalls insights from thousands of previously observed deals and draws on what it describes as the world's largest pool of operational data, which means one investor's target contributes to the benchmarks another investor's target is later measured against.

The company says it is committed to helping leading firms protect their privacy and edge, which acknowledges the tension without describing a mechanism, and nothing states what is aggregated, at what threshold, or whether a subscriber can decline to contribute.

The second is corporate: the operating entity publishes that it is separate from its affiliates and that their businesses differ, which is an unusually explicit separation statement, and the investor register points toward an active venture firm. A diligence platform seeing deal flow across many investors while sitting inside a group that invests is the QuantumStreet shape, and here the separation is asserted rather than mechanised.

The personal data chain is short, since the payload is a company's transactional and financial records rather than information about individuals, the mitigating property recorded for Nammu21 and Daloopa. The sensitivity is commercial instead, and it is acute: a target's full transaction history handed over during a competitive auction is among the most closely guarded material a private company holds, and in one published case it ran to more than 120 million records. No retention schedule, deletion commitment or subprocessor list was located, and nothing states what happens to a target's data when a bidder loses the deal or when a portfolio company is sold.

A trust centre is published as a standing part of the site alongside a responsible disclosure programme, platform documentation and a documents section, which is a genuine disclosure surface rather than an assertion and is more than most vendors in this index maintain. A responsible disclosure route in particular indicates a company that expects external security researchers to find things and has decided how to receive them.

Held at B because no individual framework or attestation is named in accessible material, so the position resembles Socure's, where a hosted trust centre exists but does not render its framework list publicly, and a buyer still has to request the underlying reports.

No supervisor, statute, rule or admission process is named, and the vendor holds no licence and needs none, which is the correct posture for a technology supplier. What is present is unusually careful securities disclaimer language, stating that nothing on the site constitutes an offer to sell or a solicitation to buy any security and that past performance is not indicative of future results, which shows awareness that analysis feeding investment decisions sits near a regulated boundary.

What is not addressed is the regime the output enters, since limited partners and institutional allocators using these scans do so under fiduciary duties and their advisers under rules governing the diligence supporting a commitment.

No consumer subjects apply, so the axis adapts, and for a benchmarking product the governance question is what the comparison set contains. A company is scored against benchmarks assembled from the deals this platform has previously observed, so the reference distribution reflects the sectors, stages, geographies and business models its subscribers happen to transact in.

A target outside that distribution is measured against norms that may not apply to it, and the consequence is not an abstract one: a scan showing deterioration against benchmarks is what a principal takes to a board. Nothing published describes the cohort behind a benchmark, how comparability is established, or how the platform signals that a comparison set is thin for a given company.

No guarantee, indemnity or falsifiable accuracy commitment was located, and the published position runs the other way, stating explicitly that the platform does not substitute the subscriber's own independent analysis. That statement is honest and it is the correct division of responsibility, but a disclaimer allocating risk to the customer is not recourse.

What keeps this off the floor is that the subscriber retains the underlying data and can reconstruct any finding, and that the responsible disclosure programme provides at least one published channel for raising a defect. Nothing describes correction of an erroneous scan, notification if a benchmark is later found to be wrong, or what a client is owed when a diligence finding proves unfounded after a deal has closed.

No model provider, hosting arrangement or subprocessor is named. One honest structural disclosure does appear, since the company describes models and operators working together rather than presenting the output as fully automated, which tells a buyer that people see their data as part of normal delivery, and that is a disclosure most vendors would rather leave implied.

The data side is described in kind, comprising subscriber supplied transactional records plus the accumulated pool of previously observed deals, without naming any external source. Marloo sets the standard this axis can reach by naming both its model providers and the retention terms governing them.

Ingestion is the whole integration story and it is described as a coordinated exercise rather than a connection, with the vendor working with the deal partner to collect and load financial and transactional data. No accounting system, billing platform, data warehouse, customer relationship system or virtual data room is named as a source, and no developer documentation or interface reference was located beyond general platform documentation. For a product that must consume a target's transaction history under deal timescales, how that data arrives is the operational question, and the published answer is that people arrange it.

No hosting provider, region selection, residency commitment or private deployment option was located. The exposure is commercial rather than regulatory, since the material is company records rather than personal data, but a sponsor handing over a target's full transaction history during a live process will want to know where it rests and for how long, and nothing published answers that.

No pricing, packaging or basis of charge is published. The product is structured as a catalogue of turn key diligence reports alongside bespoke engagements, which implies a productised price exists for the standard scans, and none is stated. The unit matters here because diligence demand is episodic and a firm cannot tell whether it is buying per scan, per portfolio company, per seat or as a subscription.

Four buyer types are named and one of them appears nowhere else in this index: sovereign wealth funds, alongside venture capital and private equity firms, corporate development and merger teams, and company founders using the same analysis on themselves. Use cases span the investment lifecycle rather than a single moment, covering pre investment diligence, portfolio diligence ahead of follow on decisions and board meetings, value creation work and benchmark development.

What holds this at B is that the analysis is confined to growth quality in operating businesses, so it does not reach credit, real assets or public markets, and no international footprint is evidenced.