LeadHeed

Not an artificial intelligence product and it does not present itself as one. The homepage, the four product pages and every page title carry no model claim at all; the positioning throughout is a simple record for beginners. One section of the pricing comparison table is headed for artificial intelligence powered features and contains exactly two rows: an email writing assistant marked coming soon, and automated task generation, which the homepage itself describes as a rule that assigns a task when a deal enters a chosen stage.

Strip every model from the product and the whole of it remains, because the only model dependent feature has not shipped. Graded at the middle band under the convention that reserves the lower band for a marketed claim failing the removal test rather than for a competent product that makes no claim. The tension worth recording is that a rules trigger is filed under an artificial intelligence heading in the one table a buyer uses to compare tiers.

Automation is rule based and every rule is written by the operator. The three published examples are all stage triggered: send a prepared email when a deal moves to a chosen stage, send a reminder when a deal sits in one stage past a set time, and assign a task to the deal owner when a deal enters a stage. Nothing initiates contact on its own judgement and nothing acts outside a condition somebody configured.

An all activities view records what happened and when, which is an activity log rather than an oversight control. Held at the middle band because approval steps, sending ceilings, escalation paths and an auditable record of automated actions are all unaddressed, and because the stuck deal rule mails a prospect with no person in the loop at the moment of sending.

The single model dependent feature on the whole surface, an email writing assistant, is marked coming soon in the pricing comparison table, and its allowance cell contains an unfinished internal editorial note rather than a value. No model, provider, hosting location or processing boundary is named anywhere on the pages read, and there is no documentation of how generated text would be produced or reviewed.

Held at the middle band rather than lower because very little is claimed and therefore very little is left undisclosed. The gap becomes material the moment the feature ships, because a writing assistant inside a record of correspondence reads customer email content and nothing published says where that content would go.

Evidence of outcomes is absent at every level a buyer could check. Ten customer logos sit under a heading claiming trust and recommendation from more than one hundred industry leaders; not one is linked, not one is named in any case study, and not one corresponds to a company that could be located.

Four statistics counters, for businesses onboarded, automations created, leads captured and pipelines created, rendered at their animation starting values on fetch, so the figures the vendor publishes were not readable and are recorded here as unread rather than absent. A heading stating the product is rated among the best customer relationship management software has no review content beneath it.

There are no case studies, no named customers, no quantified results, no analyst coverage and no located listing on the major software review platforms, while the vendor's own announced telephony partner does carry one. The about page claim of hundreds of teams across countries stands entirely unsupported.

One genuine compliance statement exists and it sits in the privacy policy rather than in any commercial document. The omnichannel messaging section allocates responsibility plainly: messaging is customer initiated unless platform rules permit otherwise, customers are responsible for obtaining lawful consent from end users before messaging them, end users may withdraw consent at any time, and the vendor states that it does not itself send unsolicited promotional messages.

The problem is coverage. That allocation is written for the three social messaging channels, and all three are marked coming soon in the pricing table. The channels that do ship, stage triggered email automation on the middle tier and built in calling and text messaging on both paid tiers, carry nothing at all: no statute named anywhere on the site, no acceptable use or anti spam policy located, no unsubscribe mechanics, no do not call handling and no time of day rules. A fair use policy is listed in the footer and was not read this pass.

Materially stronger than the vendor's size predicts. The controller is named as a specific legal entity incorporated in Hong Kong, the governing statute is named by its ordinance number, four lawful grounds are enumerated against seven stated purposes, and data subject rights including portability are listed with three routes to exercise them.

Five sub processors are named individually with the function each performs, covering application infrastructure, website hosting, the social messaging layer, live chat and payment processing, which is a named supplier register that several far larger vendors in this index do not publish. A data processing agreement is offered with six obligations enumerated, including deletion or return of personal data on termination.

A dedicated deletion request page carries a thirty day commitment after identity verification, and the policy states plainly that personal data is not sold. Off the top band on four items: transfer safeguards are described only as contractual protections with no mechanism named; the processing agreement is available on request rather than published; no data protection officer or European representative is identified; and the cookies policy link inside the policy body points to a working document in an editing view on a third party document service, while a proper cookie policy page exists in the footer.

Clean by absence and that is the honest reading. There is no contact database, no enrichment waterfall, no intent supplier and no scraping mechanism anywhere on the surface. Every record arrives because the customer imported a spreadsheet, built a capture form, or received an inbound message on a channel they connected themselves.

The people whose details end up stored hold a relationship with the customer rather than with the vendor, which is among the least extractive sourcing models in this category. Held at the middle band because those third parties still have no notice that their details sit in this particular system and no route designed for them, though the published deletion request page is a better answer than most vendors of this shape offer and is recorded in the vendor's favour.

The social channels are approached through the platform owner's own business messaging products rather than around them. Integrations are enabled only after the customer connects their own account and grants authorisation, customers may disconnect at any time, and the policy defers explicitly to platform rules on when messaging is permitted.

Nothing on the site offers a browser extension, a scraping mechanism, account rotation, proxy infrastructure, multiple identity sending or any language about avoiding detection, which is the behaviour that puts vendors in the bottom band of this axis.

Off the top band on two counts: the messaging integrations are marked coming soon in the pricing table, so the stated posture currently runs ahead of the shipped mechanism; and built in calling and text messaging do ship on the paid tiers with no carrier registration, sender identity or messaging conformance position published anywhere.

Nothing published addresses the question. There is no statement on whether content held in customer records is used to train or improve any model, no cross tenant boundary commitment, and no model provider named. The closest the privacy policy comes is a platform improvement purpose scoped to analysing usage patterns and improving performance and features, which speaks to telemetry rather than to content.

The gap sharpens against the roadmap: an email writing assistant is announced as coming, and a writing assistant inside a system that stores correspondence, deal notes, attached proposals and social messages has an obvious reason to read that material. A buyer signing today has no published answer to what happens when it ships.

Nothing synthetic operates anywhere in the product today. There is no agent, no generated sending in production, no cloned voice, no manufactured caller identity, no automated social action and no simulated conversation, so every outbound act still originates with a person who wrote it. The privacy policy also states that the vendor does not send unsolicited promotional messages and places the consent obligation on the customer.

Held at the middle band rather than higher because there is no affirmative disclosure mechanism of any kind: stage triggered emails and stuck deal reminders reach a prospect in the operator's name with nothing marking them as automated, the announced email writing assistant has no stated disclosure position, and the European transparency obligation covering artificial intelligence systems that interact with people is nowhere addressed on the site.

Connection is promised broadly and documented barely at all. A third party integration feature page exists and the homepage undertakes to connect the product to the tools a buyer already uses, but the only specific connections named across the pages read are the three social messaging channels and the partner telephony platform, and all four are marked coming soon in the pricing comparison table.

There is no programmatic access, no developer documentation, no webhook surface, no listing on a general automation connector, no marketplace and no agent protocol server. Two way email synchronisation is the practical exception and it does ship on every tier including the free one; calendar connection is itself marked coming soon. This weighs heavier for a system of record than for a point tool, because a record that cannot exchange data with the rest of the stack becomes the thing a growing team has to leave.

Processing locations are disclosed without a commitment attached. The policy states that data may be processed in Hong Kong, in the United States, and in other jurisdictions where infrastructure providers operate, and it names the cloud provider for the application and the separate host for the marketing site. No region is offered as a choice, no data centre or cloud region is identified, and no residency option is sold.

The open ended third clause is what holds this at the middle band rather than one higher: a vendor stating that data is collected globally and stored in one named country has taken a position with no options in it, which is still a position, whereas a clause covering wherever suppliers happen to operate commits to nothing a buyer can hold.

Six controls are enumerated in the privacy policy: encryption in transit, encryption at rest, role based access control, infrastructure on a named major cloud, logging and monitoring, and regular security reviews, followed by a plain acknowledgement that no system is completely secure. Naming specific controls and declining to overclaim is more than most vendors of this size manage.

What is missing is any independent verification: no certification of any kind is claimed, there is no trust centre, no audit report or audit period, no penetration test, no dedicated security page, no vulnerability disclosure route and no status page. Assertion without attestation is the definition of the middle band.

Among the more complete published price pages of any vendor this size. Three tiers carry real figures on both billing periods, zero, fifteen and thirty five dollars per user per month, falling to zero, twelve and twenty eight billed annually, with the saving stated.

An eight section comparison matrix renders every tick and cross, and the hard limits that decide the real bill are published per tier: user counts of three, ten and unlimited, contact allowances of five thousand and fifteen thousand, pipeline counts of one, ten and unlimited, a form builder cap and email template counts. Signup is self serve with a fourteen day trial and no card. Off the top band on three items.

The upper tier contact allowance reads as unlimited or an add on, and no add on price appears anywhere. Built in calling and text messaging ship on both paid tiers with no per minute rate, per message rate or bundled allowance published, which is a consumption line sitting inside an otherwise complete price. And the trial length contradicts across pages: the pricing page and homepage both say fourteen days while the contact page answers the same question with thirty days, twice.

Bulk import and export ship on every tier including the free one, which is more than several paid only competitors offer, and the export ceiling on each tier matches that tier's own contact allowance rather than sitting below it.

The privacy policy enumerates a portability right, names deletion or return of personal data on termination as an obligation inside the offered processing agreement, and publishes a dedicated deletion request page with a thirty day commitment after identity verification. That is the post termination half of this axis at least partly answered, which is uncommon at this size.

Off the top band because the instrument carrying the termination obligation is available only on request rather than published; because no retention window, deletion timeline or export file format is stated for the account itself; because report export is withheld from the free tier; and because there is no programmatic route out at all, so extraction depends entirely on a manual download.

Sending runs through the customer's own connected mailbox rather than shared vendor infrastructure, which keeps the reputation consequences with the party making the sending decisions and avoids the pooled sender problem that catches cheaper platforms. Beyond that the axis is unaddressed.

Nothing is published on authentication guidance, sending volume limits, warmup, bounce handling, complaint thresholds, blocklist monitoring, list hygiene or unsubscribe mechanics, and stage triggered email automation ships on the paid tiers regardless. Open and click tracking is marked coming soon, so there is not yet an engagement signal available to govern sending with even if a buyer wanted to. Built in calling and text messaging ship with no sender registration, number reputation or carrier filtering position published.

The core segment is stated consistently and the product genuinely matches it: small businesses, first time buyers of a record system, and teams outgrowing a spreadsheet, with a free tier capped at three users that gives the claim a numeric floor and a direct spreadsheet comparison in the frequently asked questions. Breadth then undoes it.

The pricing headline widens to businesses of all sizes, an enterprise solutions page sits alongside the small business one from a vendor whose top tier is thirty five dollars per user, and the solutions menu spans fourteen industries, six buyer roles and four internal teams, one of which sits outside the go to market stack entirely. Several of those pages read as search coverage rather than product specialisation. No ceiling and no disqualifying condition is stated anywhere.