demonstro

The product makes no artificial intelligence claim anywhere on the pages read, which in a corpus where the category norm is to lead with a model is notable in itself and is the fourth such vendor in this index. The mechanism is set intersection: a member uploads target account domains, the platform compares that list against every other member's list, and a match surfaces where two sellers each want access inside the other's company.

One feature edges toward inference, since recommendations are said to analyse each organisation's ideal customer profile to suggest matches beyond simple overlap, and no method or model is claimed for it. The lowest grade on this axis is reserved for a marketed claim that fails the removal test, not for a competent product that does not make one.

A human decides the last step and the first step happens without one. Matching runs passively and the vendor sells it that way, with the middle stage of its own three step explainer given as chill, and the member's involvement beginning only when a notification arrives and they choose whether to send a meeting link. That final choice is real oversight.

What sits before it is not described: once a target account list is uploaded it is continuously exposed to matching against every other member, with no account level control, no approval before a match is surfaced to a counterparty, no described way to withdraw a specific account, and no log of which members were matched against which entries.

There is nothing claimed to disclose, so the axis is graded on what is knowable. The one capability that implies inference is recommendations, described as analysing an organisation's ideal customer profile to propose matches that go beyond list overlap, with no method, input set, confidence measure or accuracy figure published. A member acting on a recommendation is spending a meeting slot and a colleague's goodwill on a suggestion whose basis is undescribed, which is the practical cost of the silence here.

The proof surface is empty and what fills the space is category research rather than product results. No customer is named, no logo appears, no case study exists and no outcome is quantified for any member. In the slot where results normally sit are three statistics about referrals in general: that 84 percent of B2B buyers start with a referral, that 90 percent of buying decisions are influenced by peer recommendations, and that referrals carry 24 percent higher contract values.

None carries a source. A counter promising the number of companies targeted by users renders empty in the served page, a community is claimed with no size given, and one review exists on an independent platform. Recorded as an early stage fact rather than as concealment, and the substitution of category claims for product evidence is the specific defect.

The regulated surface is unusually thin because the platform sends nothing to a prospect. The approach reaches its target as an introduction made by a colleague inside their own company, using that colleague's own channels, so the obligations that normally attach to commercial email or telephone contact do not arise in the same form. No regime is named anywhere, which costs little here and is still recorded.

The claim that sits closest to this axis is architectural rather than legal: the vendor states the model guarantees complete deliverability because messages arrive as internal referrals rather than through crowded inboxes and spam filters.

A terms of use and a privacy notice are published at stable URLs and neither was read this pass. The question specific to this product goes unanswered on the surfaces that were read, and it is a sharp one: a target account list is competitive intelligence, and here it is uploaded to a third party and continuously compared against the lists of everyone else on the platform, including sellers at companies a member competes with.

What is retained, for how long, what a counterparty can infer from a match, and what happens to the list on cancellation are all undescribed. A data protection officer, a sub processor list and a data subject request route were also absent from the pages read.

The posture is light because almost no data is sourced. There is no contact database, no enrichment, no scraping and no third party supply: the inputs are company domains a member types or uploads, and the people who appear are the members themselves and the colleagues they choose to introduce. Nothing is sold onward. Off the top of the band on the shape that makes this product unusual rather than on a sourcing failure.

The uploaded target account list is the platform's raw material and it is matched against every other member's, so a member's account strategy becomes an input to a shared engine, and no published statement describes retention, isolation between members, or what a competitor on the platform could infer from the pattern of matches they do and do not receive.

Conventional exposure is absent by design. There is no browser extension, no automation running under a user's account on someone else's platform, no credential custody, no scraping and no sending infrastructure, and the integrations named are calendar and CRM connections through official routes.

The exposure that does exist is a different kind and nothing published addresses it: the terms most at risk here belong to the member's own employer rather than to a platform, because the model asks a seller to introduce an outside vendor to a colleague in exchange for consideration, and whether that is permitted sits in an internal policy the platform never mentions.

No model is claimed, so the training question does not arise in its usual form, and the stewardship question that replaces it is unanswered. What accumulates is a map of who is targeting whom across the membership, built from uploaded account lists, plus post meeting feedback scores attached to named individuals through the trust score mechanism.

Whether that aggregate is used for anything beyond matching, how long feedback about a person is retained, and whether a member can see or contest a score recorded about them are all undescribed on the pages read.

A new shape for this axis and it is the finding of the build: the relationship is real and the consideration behind it is not disclosed. A seller at one company introduces a seller from another to a colleague inside their own organisation, and in return receives an introduction going the other way. Every person in the chain is genuine, the introducer really is a colleague, and the vouching really does come from inside.

What the person being introduced does not learn is that access to them was traded, and the platform makes the barter explicit internally through a trust score that monitors reciprocity balance while it stays invisible externally. The vendor describes the arrangement as a double opt in, and the two parties opting in are the two sellers rather than the prospect, who is the only person in the transaction who did not agree to it. Compare the warm introduction routing graded elsewhere in this index, where the recipient's belief that a colleague knows them and is vouching is simply true; here it is true and incomplete.

Three integrations exist and two of them sit behind an unpriced tier. Calendar booking arrives on the paid individual plan, and CRM connection to Salesforce and HubSpot is available only on the enterprise plan, which also carries the only route to importing prospect lists from a CRM rather than by hand or by spreadsheet. No public API, webhook surface, developer documentation, marketplace or agent facing endpoint was located. For a product whose entire output is booked meetings that a seller then needs to log, the absence of a CRM path on every plan below enterprise is the gap that matters most.

The residency question goes unanswered across every page read. Hosting provider, processing region, storage location and sub processor list are all absent. The material held is modest in volume and sensitive in kind, consisting of target account lists, calendar connections on paid plans and CRM connections at the enterprise tier, and a buyer weighing whether to upload their account strategy has nothing published to tell them where it will sit.

A security page, trust centre, certification, audit, penetration test, control set and status page were all absent from the pages read, and the footer legal section carries exactly two items, terms and a privacy notice. Held at the bottom of the band rather than lower on two grounds: no certification is claimed and left unsubstantiated, and the product requests less access than most, with no mailbox, no social account and no credential custody. Against that, the asset it does hold is unusual. A target account list is competitive intelligence, and the platform holds one for every member and matches them against each other.

Two published prices with real limits attached. A free tier carries 100 prospect accounts and two meetings, the individual plan runs at 39.99 US dollars a month or 360 a year with 500 accounts and unlimited meetings, and a comparison table sets the quantified rows side by side, including prospect upload method and integration availability per tier. Off the top of the band on three counts. The enterprise tier is unpriced and is the only route to CRM integration and to unlimited accounts.

Four features are named in that comparison table and defined nowhere on the site, so a buyer is asked to compare plans on terms they cannot look up. And recorded as a retrieval limit rather than a vendor defect: the tick and cross marks did not survive extraction, so per tier availability of the unquantified rows could not be read this pass.

Getting data in is documented and getting it out is not. Prospect lists arrive manually, by spreadsheet, or from a CRM at the enterprise tier, and no export function, download route or portability statement appears anywhere on the pages read. Nor does any post termination position: retention period, deletion timeline and deletion artefact are all absent, which matters more than usual because the material a member would want removed is the account list they uploaded. The published terms of use are the likely home for any of this and went unread this pass.

The platform sends nothing to a prospect, so the axis is graded on what is knowable. The message reaches its target through a colleague's own channel as an internal referral, which means the sending reputation at stake belongs to nobody the vendor controls and the usual control set has no surface to attach to. One claim sits on the record and deserves naming: the vendor states the model guarantees complete deliverability. Read as an architectural point about a referral bypassing filters it is defensible, and stated as an absolute it is the kind of guarantee no channel supports.

Segmentation is asserted and not built. The footer names four audiences, founders, sales leaders, sellers and demand generation, and every one of those links points back to the homepage, so there is no page behind any of them. The market is described in the headline as B2B software sales and the frequently asked questions promise answers on fit for specific roles and company types, though that content did not render on the page as served.

No headcount band, customer count, size distribution or ceiling appears anywhere. Worth noting in fairness: the exchange model has a genuine structural constraint that the vendor never states, since a member only receives matches where another member is targeting their own company, so fit depends on the shape of the network rather than on the buyer alone.