ReactIn

The model layer scores every lead against the customer's ideal profile before a message goes out, segments captured signals, and personalizes outreach to match the customer's tone of voice, with the privacy policy confirming external providers generate the messages. Apply the removal test and the platform survives: signal capture across twenty plus triggers, self updating lists, enrichment, sender rotation with caps, branching campaigns and the unified inbox are all conventional machinery. What is lost is the score and the draft. The signals architecture, not the model, is the product's actual differentiator.

The pipeline is sold as autopilot: signals add prospects to lists without intervention, scoring gates which leads get messaged, campaigns branch and follow up automatically, and the vendor's own welcome flow sends an automated direct message from the founder to every new signup. The stated controls are throughput mechanics, each sender auto capped at what the vendor calls its safe daily limit with managed active hours.

The ideal profile score is a filter, not oversight: no approval step, review queue, role model, permission set or activity record is described anywhere, and nothing states what a customer can inspect or stop before a message leaves in their name.

The privacy policy discloses the dependency at category level, stating data may be shared with artificial intelligence providers for message generation, which confirms message content leaves the platform and is more than most vendors this size volunteer. No provider, model, family or version is named anywhere, and no accuracy or evaluation is published for the ideal profile scoring that decides who gets messaged.

The headline number, a twenty seven percent average reply rate against an eight percent cold list baseline, is repeated across the site with no sample, period or measurement method, and a funnel worked example running from five hundred scraped followers to five clients is illustration rather than a customer report. The scale claim is three hundred plus go to market teams, with not one named, and the single quoted user is identified by first name and initial only.

One genuinely substantive independent review exists in French sales tooling media and describes the product accurately, which is third party validation of what it does rather than evidence of what customers achieved.

The product initiates automated direct messages to people because of something they did elsewhere, commenting on a competitor's post, visiting a profile, attending a webinar, and sends email campaigns beside them, and no statute, consent standard, suppression mechanism or unsubscribe mechanic is stated anywhere read. The European regulation appears in the legal documents only in respect of the customer's own account data. Processing the behavior of third parties into outreach triggers is the product's whole premise, and the lawful basis for it is never addressed.

A real, current, bilingual policy exists and carries two disclosures rarely published at this size: a stated retention outcome, deletion within twelve months of cancellation, and a named complaint route to the French data protection authority, alongside enumerated European rights including portability and a data protection officer contact. Three defects hold it here.

The statutory legal notice was published with unfilled template placeholders where French law requires the registered address, company registration number, share capital, publication director, telephone and consumer mediator to appear, so the accountable entity behind the policy cannot be located. The data protection officer is a generic mailbox. And the prospects the product exists to process, scraped followers, profile visitors and enriched contacts, appear nowhere in the policy, which is scoped to the paying customer.

Every lead is automatically enriched with company data, job title and contact information against included credits, and no source, database, partner or method is stated for any of it. Scraping competitor followers is marketed by name as a capability, so the collection technique is disclosed for one input and the licence position for none. The distinction between signals the customer's own tools emitted, payments, bookings, form fills, and intelligence assembled about third parties from the professional network is never drawn.

The bottom band is taken from the vendor's own marketing, which is the standard this index set on the first application of it. Rotation of dozens of professional network and email senders is the headline scaling mechanism, up to eight senders per organization on standard plans and up to fifty on the agency plan, run from one dashboard. The homepage states the purpose outright: scale your outreach without triggering platform restrictions.

Scraping competitor followers is a named capability on a network that prohibits automated collection, and the question and answer content confirms multi sender rotation for outreach at scale. The auto capped safe daily limit per sender is volume management in service of the same objective. Rotation framed as the way past what one account may do is the exposure itself, sold as the feature, and the account that gets restricted belongs to the buyer.

The policy confirms message content is shared with external artificial intelligence providers for generation and that data is used to provide and improve the services, and nothing states whether any customer's content or any prospect's scraped profile trains anything, whether customers are isolated from one another, or where inference runs. The stated retention window and encryption commitments cover storage, not model data flows.

The product's premise is that a person's behavior elsewhere, a comment on a competitor's post, a profile visit, a webinar attendance, silently becomes the trigger for a personal seeming direct message, drafted by an external model to match the sender's tone of voice.

Nothing marks any message as automated, no position on the European marking obligation appears, and the vendor's own welcome mechanic makes the point: every new signup receives an automated direct message presented as a personal note from the founder. The recipient is not told the attention is manufactured, and the person whose engagement triggered the sequence is not told they were watched.

The programmatic surface is specified in a way most vendors this size never manage: webhooks fire on five named events, lead added, lead enriched, connection accepted, message replied and first message sent, with connections to the two major workflow platforms and a programmatic interface for custom builds. Inbound trigger sources are named individually, a payment platform, a scheduling tool and a form builder among more than ten.

A dedicated integrations feature page, a template library, a changelog and a help centre all exist, and four published comparison pages name competitors directly. Off the top band: no developer documentation was located, there is no marketplace, and system of record delivery runs through the workflow platforms rather than native connectors.

The legal notice names the hosting provider and its United States address, which is a real disclosure most vendors omit. It also creates the question it does not answer: a French company subject to European data protection law hosting on United States infrastructure, with no transfer mechanism, region choice or residency option stated anywhere, and the company's own registered address published as an unfilled placeholder.

The policy enumerates a control set, encryption in transit and at rest, strict access controls, continuous monitoring and regular security audits, which is a real published posture for a platform holding authenticated access to up to fifty sender accounts per customer. The audit claim carries no type, auditor, period or report, and no certification, attestation, trust centre, status page or vulnerability disclosure route was located anywhere. Asserting audits without enumerating them is weaker than it looks, and the sensitivity of the access class here makes the gap between the two words and the evidence wide.

The published path is real and computable: twenty nine dollars per sender per month for the signals only entry tier, sixty nine for the full outreach tier, and nine hundred ninety nine monthly for the agency plan covering up to fifty senders, with the per account arithmetic stated on the page. Trial terms are unusually concrete, seven days with no card and an extension promise if no replies arrive.

The entry tier's exclusions are printed honestly, no outreach and no lead management on an outreach platform's cheapest plan. Off the top band: the agency card lists custom price inside a card headlined with a figure, no annual pricing appears, the two thousand included enrichment credits carry no published top up rate, and the sender rotation the product is sold on makes the real bill a function of account count the page leaves the buyer to multiply.

Portability here is structural before it is contractual, which is the pattern this index has learned to credit. Webhooks push the commercially valuable record outward continuously on five lead events, so pipeline data accumulates in the customer's own tools by default, and the workflow platform connections plus the programmatic interface give two further outbound routes.

The policy then adds the contractual half most small vendors omit: an enumerated portability right and a stated deletion outcome within twelve months of cancellation. Off the top band: no export function or format is documented as a product feature, twelve months is a long tail for a departed customer's data, and the terms of use were not read.

Per sender caps, managed active hours and what the vendor calls safe daily limits are stated as automatic, which is a real governor, and none of the numbers behind them is published, so a buyer cannot know what the cap actually is. The same machinery is then marketed for the opposite of discipline: rotating dozens of senders to scale volume past what one account sustains, which is the settled position on this axis. Nothing addresses warmup, authentication records, bounce handling or complaint monitoring for the email half, and the sending infrastructure is never described.

Three buyer types are addressed by name, go to market teams, software startups and agencies, with the agency segment given its own tier, use cases and economics, and the site runs bilingual in English and French. The scale claim of three hundred plus teams carries no names, and no region, country, industry, seat band or language coverage statement appears beyond the interface languages.

An independent review adds the sharpest segmentation fact the vendor does not state itself: the product is built for structured teams rather than solo operators, and sits in the premium bracket of its category. Nothing published says who it is wrong for.