Linkyfy

Generation carries the positioning entirely. The vendor sells message writing from the prospect's actual role, company and recent activity and defines itself against the alternative in its own words, not a merge tag and not a template. Three of the four steps in the published workflow lean on it: relevance filtering, message generation, and post discovery.

The removal test still leaves a product, a capped connection sender with a profile fetcher and an email finder, and the price card confirms it. Every tier is priced on volume units, invites a day, profile fetches a day, posts a day, emails found and sent, and not one tier prices a model capability. When a vendor's own ladder scales on throughput rather than intelligence, the price card is telling you what the removal test tells you.

One structural control does real work and it is a capacity limit rather than an oversight mechanism: 20 invitations a day and 20 profile fetches a day, holding constant across the free, individual and team tiers, with the top tier reaching 60 only by adding two more accounts at the same per account rate. The oversight claim itself is a sentence without a mechanism, that the platform combines automation with user control to keep outreach intentional.

Whether a generated message is reviewed before it is sent goes undescribed, and no approval queue, escalation threshold or record of what was sent on the buyer's behalf appears anywhere. Automatic follow ups, which would raise the question sharply, are published as roadmap.

Model claims run through every page and the model itself is never identified. Provider, family, version, hosting and boundary are all unstated, and the closest the site comes is a phrase in the frequently asked questions referring to the underlying structures doing the analysis, which discloses nothing at all.

The contrast with the vendor graded immediately before this one in the same alphabetical block is instructive: a smaller competitor with a thinner product publishes a dedicated page naming its provider and model family outright. Naming a model is not expensive and it is not happening here.

The evidence surface is thin but honestly labelled, which is what separates it from the band below. Two testimonials appear under headings reading early user feedback and loved by early users, attributed to a first name and initial with a country and a role, and the section links out to an independent review site where a reader can check for themselves.

Against that sit three headline figures with no basis attached, an 85 percent response rate, 40 hours saved and 1,247 connections, the first of which is far outside any published benchmark for this channel. A separate safety claim of zero accounts blocked to date carries an asterisk whose footnote was not located anywhere on the page. A named customer with a measured result would move this row.

A privacy policy and terms and conditions are published and linked from the footer, and third party indexing of those terms describes them as covering anti spam provisions and account usage guidelines alongside billing and cancellation, which suggests a real acceptable use position sits inside them.

Across every marketing page, the frequently asked questions and the pricing table, no electronic marketing statute is named, no consent position on the outreach appears, and no unsubscribe, suppression or recipient removal route is described, which matters more now that the product sends email as well as platform messages. The terms were not read and this row is flagged for re verification rather than graded as an absence.

A privacy policy is published and linked from the footer alongside terms, and it was not read, so this row records what the public surfaces say and flags the document for re verification. The public surfaces say very little: a footer tagline promising results while maintaining privacy standards, and a frequently asked questions answer stating that sensitive credentials and private data inputs remain securely locked within the infrastructure and are never shared or sold externally.

Neither statement identifies a lawful basis, a retention rule, a sub processor, a transfer mechanism or a data subject route. The prospects whose profiles are fetched and analysed, and whose business email addresses are found and validated, are not mentioned on any public surface read.

Two distinct sourcing questions arise and both get a characterisation rather than an answer. Prospect discovery draws on profiles browsed through the buyer's own connected account, described as analysing publicly available profile snippets such as job titles, summaries and industry sectors.

The native email finder is the sharper question: it is sold as surfacing verified business addresses linked to profiles and as scrubbing business profiles and validating delivery, priced as a daily allowance of 16 addresses on paid tiers and 48 on the top tier. Supplier, database, matching method and licence are unnamed for that function, and the validation step implies contacting mail infrastructure on the recipient's domain, which nothing addresses.

The posture here is unusually conservative for this category and the framing deserves credit for pointing the right way. Sending runs from the vendor's cloud on a dedicated address per account rather than a shared pool, with randomised delays and varied pacing, and the vendor states the purpose plainly as keeping outreach within the platform's own daily limits.

Account rental, address rotation to exceed limits, antidetect browsers and unlimited profile stacking are all absent, and the account ceiling is a stated three rather than open ended. Held at the middle band because human like activity patterns and a zero accounts blocked claim are still detection management by another name, because automating with a connected account is squarely what the platform's agreement addresses, and because the vendor states no conformance position of its own anywhere.

The vendor poses this axis's central question in its own frequently asked questions, asking whether it uses customer data for model training, and then answers a different one. The reply states that profile information is processed to run outreach workflows, that the underlying structures analyse publicly available profile snippets to construct optimised communication notes, and that credentials and private inputs are never shared or sold to external third parties.

Never shared externally and never used internally are different commitments, and only the first is made. Posing the question and declining to answer it is more informative than silence and it is still an unanswered training question, which is what this band records.

The authenticity language is present and comparatively mild: messages are described as human like, context aware and reading authentically, positioned against templates and merge tags rather than against detection. A recipient receives a note referencing their real role, recent post or company news, written by software, from the buyer's own genuine account under their own name.

That last point matters and separates this from two vendors graded earlier in the same block, where the sending identity itself belonged to somebody else. Nothing on any surface read discloses artificial authorship to the recipient, and the European disclosure obligation goes unmentioned.

Live integration depth is close to zero and the vendor is candid about it. Connectors are published under a coming soon heading, described as protocol based and naming two customer relationship systems, one of them apparently misspelled. The only live external attachment is a mailbox connected for sending. A knowledge base and a support surface exist, which is more than several peers maintain.

The frequently asked questions position this deliberately, stating the product complements existing tools rather than replacing a customer relationship system, but complementing a stack requires connecting to it and today nothing connects.

Execution is described as cloud infrastructure with a dedicated address per account and no dependence on the buyer's machine, which answers the architecture question and not the residency one. Processing region, storage location, transfer mechanism and any choice for the buyer are all unstated, and no legal entity or registered address appears anywhere on the site.

The 18 country list published elsewhere is easy to misread here and should not be: it governs which platform accounts the automation supports, not where any data lives. That is the second time in this session a coverage list has been positioned where a residency statement belongs.

The trust surface is empty. A certification, an audit, a penetration test and a trust centre were all sought across the navigation, the footer and every page read, and none was located, from a vendor that holds connected professional network account access and mailbox sending credentials for its customers.

The security claims that do appear are two sentences: that credentials remain securely locked within the infrastructure, and that a dedicated address per account is safer than a shared pool. Neither is a control a buyer can verify. The privacy policy was not read and a documented control set inside it would move this row one band, which is exactly what happened for the vendor graded immediately before this one in this session, so the document sits at the top of the re verification list.

The allowance table is the most complete in this alphabetical block and it is not close. Four tiers at nothing, 39, 69 and 149 dollars a month each publish connected accounts, invitations a day, posts a day, campaign counts, profile fetches a day, and email addresses found and sent a day. A buyer can model throughput precisely before paying anything. Two things hold it off the top band.

The vendor contradicts itself three ways on whether the product is even paid: the price card shows four tiers, the frequently asked questions state the product is currently free while the initial version rolls out, and a waitlist offers to notify the reader when pricing goes live. And every tier button on the price card resolves to a dead anchor.

Worth flagging for any buyer reading the ladder: the 69 dollar tier carries identical accounts, invitations, profile fetches and email allowances to the 39 dollar tier, so 77 percent more money buys post volume and one extra campaign, nothing more of the outreach capacity the product is sold on.

Cancel anytime appears twice in promotional copy and that is the whole of the published exit position on the surfaces read. An export function, a file format, a download path, post termination data rights, a deletion commitment and a retention rule are all undescribed.

The asset at stake is real and growing: campaign history, fetched profiles, generated message copy and found email addresses accumulate in the vendor's cloud, and the unified inbox on the roadmap would add conversation history to that. Terms and conditions are published and were not read, so this row is flagged for re verification rather than graded lower.

This is the strongest sending posture graded in this alphabetical block and the structure of the price card is the evidence. The per account ceiling is 20 invitations and 20 profile fetches a day and it does not rise with spend: the top tier reaches 60 only by permitting three accounts at the same per account rate, and the account ceiling is a stated three rather than open ended. Email is capped at 16 found and 16 sent a day per account.

Randomised delays, varied pacing and a dedicated address per account rather than a shared pool round it out, and the vendor states the purpose as staying within the platform's limits rather than past them, which is the opposite of the posture that earned lower grades elsewhere in this index.

Off the top band on the classic omissions: no warmup ramp for a new or cold account, no complaint or bounce policy on the email side, no authentication guidance for the connected mailbox, no stated response to a platform warning, and a zero accounts blocked claim whose asterisk leads nowhere.

Geographic coverage is published to a level of specificity almost nothing in this index matches. Eighteen supported countries are named individually and grouped by region, with a plain statement that the automation will not function correctly for an account registered outside them. Publishing a restriction that eliminates most of the world from the addressable market is a disclosure against interest and it saves a buyer a wasted trial.

Buyer definition is documented too, with a dedicated page each for founders and senior executives, sales professionals, and marketing agencies, and the agency case is served concretely through multi account workspaces with per sender limits.

Off the top band because size band, industry fit and any scale ceiling beyond the three account cap are absent, no corporate entity or address is published, and several capabilities that shape who this suits, the unified inbox, automatic follow ups and connectors, are roadmap rather than shipped.