PersistIQ

Established platform where the removal test leaves the entire product standing. This company predates the current generation by roughly a decade and the model layer arrived as a single named feature under its current owner, a response generator that drafts tailored replies to inbound prospect mail.

Strip it and the buyer retains prospecting through a browser extension, multi channel campaigns across mail, calls and tasks, split testing, reply tracking, reporting across every message and team member, and bi directional synchronisation with the system of record. The product's own navigation places that feature fourth of four under the product menu, which is a fair reflection of its weight. Reply sentiment classification is the only other model dependent capability located.

One genuinely distinctive control concept and no machinery published around it. The vendor markets safeguards against embarrassing mistakes as a named differentiator, covering misformatted messages, forgotten follow ups and crossed lines of communication, that last one meaning two representatives unknowingly contacting the same prospect.

That is a guardrail aimed at protecting the recipient's experience rather than the sender's throughput, which is a rarer thing to build than it sounds and is creditable. Beyond it nothing is described: no role model, permission structure, approval step, audit trail or escalation rule was located.

The reporting layer analyses individual team members and classifies reply sentiment automatically, so machine assessment of both employees and prospects is happening with no published account of how either is governed.

No base model, provider, family, version or hosting arrangement appears anywhere on the pages read. The response generation feature has its own dedicated page in the product navigation and the launch announcement from the parent company describes what it produces without naming what produces it. Reply sentiment classification is likewise undescribed. For a vendor whose parent is publicly listed and therefore subject to continuous disclosure obligations, the absence of any statement about model dependencies is notable rather than routine.

A large independent review corpus, thin first party evidence, and one contamination problem worth naming. The four testimonials on the homepage are transparently labelled as republished reviews from a major directory, which is honest sourcing, but each carries only a first name, an initial and a job title with no employer, so none can be verified. Six customer success story cards all link to the same single page.

The independent review base is genuinely substantial at several hundred reviews across four directories, and one named customer carries real figures, having scaled from three to eight touchpoints and halved implementation time. Against that, the trust badge image displayed is dated to 2024 and the social video asset is a review recording from 2022, both stale on a page updated in 2026.

The contamination finding is the sharpest part: since 2021 the parent has sold a fully managed outbound service under this brand, and multiple reviews in the same directory listings are plainly about that service rather than the software, complaining that thousands of sent messages produced no meetings. A buyer researching the platform reads campaign results as though they were product defects.

A platform sending cold mail and placing calls at published per user daily volumes, with no compliance position located on any page read. No statute is named, and no consent requirement, unsubscribe obligation, suppression list or complaint route appears. A terms of service document exists at its own route in the footer and was not read, and is flagged as the only place a position could sit.

One point that sits adjacent to this axis rather than on it: the mistake prevention safeguards include preventing two representatives from contacting the same prospect, which is a form of contact discipline even though it is framed as protecting the sender from embarrassment rather than the recipient from duplication.

A privacy policy exists at its own route, linked in the footer of every page, and was not read, so this row is held at the middle band and flagged. Two structural facts are verified and worth recording. The parent is a publicly listed company on two exchanges, named on every page with a link to its investor relations pages, so the controlling entity is identifiable and subject to continuous disclosure obligations, which is more corporate transparency than most vendors of this size offer.

It is also Canadian, which brings a different domestic privacy regime to bear than the United States default most of this index sits under. The processing in scope includes connected mailboxes, prospect records gathered from a professional network by browser extension, call activity and reply content classified for sentiment.

The prospecting mechanism is described plainly and its provenance is not. The browser extension searches a professional network and its paid recruiter product, finds the person's email address and adds them to a campaign, and nothing published states where the address comes from, what verification is performed, what coverage or accuracy to expect, or on what basis a record is created about someone who has not asked.

No contact database is sold, which is the clean part, so the vendor is a finder rather than a reseller. An independent directory notes the product works well alongside a named third party tool that scrapes professional network searches for verified addresses, which suggests the built in finder is not always sufficient, though that is third party commentary rather than a vendor statement.

The exposure is stated openly on the homepage and carries no position. The browser extension searches a professional network and its paid recruiter product, extracts the prospect's address and enrolls them in an outreach campaign, which is precisely the condition this band describes. No conformance statement, permitted use position or allocation of risk between vendor and customer was located anywhere, and the terms of service were not read.

Held at the middle band rather than lower because the language is neutral rather than promotional about the extraction, nothing rotates or rents an identity, no evasion is marketed, and the remaining architecture connects through ordinary supported synchronisation with systems of record.

Nothing on the pages read states whether customer content is used to train or improve any model, whether prompts are retained by any provider, or whether material crosses between accounts. The response generation feature reads inbound mail from a customer's prospects in order to draft a reply, so the content being processed is the prospect's own words rather than only the customer's, and reply sentiment classification runs across the same material.

No non use commitment, confidentiality undertaking or provider retention arrangement was located. The privacy policy and terms of service both exist and were not read, and are flagged as the documents that would settle this.

No position on article 50 of the European artificial intelligence regulation appears anywhere. Machine drafted replies go to real people who wrote to a human and receive an answer composed by a model, with nothing stating that a machine wrote it. Open tracking is described in the vendor's own republished review copy as a core benefit, and reply sentiment is classified automatically, so the recipient is both watched and machine assessed without notice.

Held at the middle band rather than lower for one reason that runs against the grain of this category: the product's marketed differentiator is helping a sender communicate like a human by preventing the mechanical errors that expose automation, and while that is framed as protecting the sender, its effect is a less deceptive and less obviously automated experience for the person receiving it.

Real depth on a narrow set, and the homepage cannot list it correctly. Native bi directional synchronisation is offered rather than one way push, covering contact records, activities and the ability to enroll a contact into a campaign from inside the system of record, which is genuine integration rather than a connector badge.

A browser extension, a dedicated page for one system of record and an integrations page complete it, and independent sources report further connections including one gated to the higher tier. The defect is on the homepage itself: the sentence naming the three native integrations lists the vendor's own product as one of the three, so the published integration list includes the product doing the integrating. No interface, webhooks, protocol server, developer documentation or marketplace was located anywhere.

Nothing on hosting provider, country, region, data centre or residency option appears on any page read. What is available is corporate rather than technical: the parent company is publicly listed and headquartered in Vancouver, British Columbia, and is named on every page with a link to its investor pages, so the controlling jurisdiction is Canadian and identifiable.

That answers who is accountable and where they are incorporated, and says nothing about where connected mailbox contents, prospect records or call activity physically sit. No security or trust page exists in the footer where a residency statement would normally live.

No certification, audit, attestation, penetration test, control set, security page, trust centre, status page or vulnerability disclosure route was located, and the footer is complete and short enough to establish that as an absence rather than a failure to find: it carries integrations, one system of record, investor relations, pricing, blog, referral, careers, a privacy policy and terms of service, and nothing else.

Set against a platform holding connected mailbox sending credentials, prospect databases and call activity, that would ordinarily sit at the floor. It is held one band above for a specific structural reason absent from the vendors graded lower in this index: the controlling entity is a publicly listed company on two exchanges, named on every page, subject to continuous disclosure and audited reporting obligations. That is corporate accountability rather than a security control, and it is the reason this is thin rather than unaccountable.

There is no pricing page. Both routes a buyer would take terminate at the same demand form: the primary navigation item labelled pricing points directly at the demonstration booking page, and the separate footer link to a pricing path redirects to that same booking page.

No price, no unit of pricing, no tier name and no allowance appears anywhere on the vendor's own site, at a self serve product that simultaneously offers a free signup button on the homepage, so a buyer is invited to create an account without being told what it will cost.

The consequence is visible in the independent record and is unusually stark: one directory concludes outright that pricing is not publicly listed, while four others publish entry figures of fifty, fifty nine, seventy five, ninety nine and one hundred and twenty dollars per user per month, and they disagree with each other on whether monthly billing exists at all. When the vendor publishes nothing, the market invents five different answers. Note also that the demonstration page serves the customer results page's title, description and social tags, so a search engine describes the booking form as a page about customer outcomes.

One structural strength and no published position. Bi directional synchronisation with the system of record means contact records and activity history are written into a database the customer already governs rather than living only in the vendor's store, which materially lowers the practical cost of leaving and is the same reasoning credited elsewhere in this index.

Against it, no export function, file format, bulk retrieval route, post termination window, deletion timeline or retention period was located on any page read. What would stay behind includes campaign libraries, templates, reply histories, call activity, reporting history and the prospect lists built through the extension. A terms of service document exists in the footer and was not read, and is flagged as the document that would settle this.

One real governor, reported rather than published, and nothing else. Independent pricing research describes per user daily sending ceilings that differ by tier, at three hundred messages a day on the lower plan and six hundred on the higher one, which is an explicit throughput limit and a meaningful one, but it appears only in third party reads because the vendor publishes no pricing page on which it could be stated.

Split testing exists and reply tracking distinguishes sentiment, both of which support message quality rather than sender reputation. No warmup, sender authentication position, address verification, complaint rate threshold, bounce handling policy or reputation monitoring was located anywhere on the pages read. Mail leaves through the customer's own connected mailbox, which at least leaves the reputation with whoever controls it.

The segment claim is clear and singular and nothing sits behind it. The homepage headline names the target twice, positioning the product as the leading sales engagement platform for small teams, which is a genuine statement of fit and an implicit statement of who should look elsewhere.

Four buyer roles are illustrated through the republished reviews, covering an account executive, a sales manager, a co founder and a sales development representative, which reflects a small team where those four people are the whole revenue function. Independent research corroborates the segment, placing the customer base in small and mid sized technology and business services companies.

Against that, no region, country, language, industry, seat band or customer count appears anywhere, the only quantitative claim is that thousands of growing companies are served, and there is no statement of the size at which a team outgrows the product despite the whole positioning resting on team size.