Outstack

The model layer is the middle of three steps and the other two are plumbing, yet a working pipeline survives its removal. Strip the models and the buyer loses per prospect research, the generated personalised copy and the automated lead scoring, and retains a professional network search scraper, an email verification lookup and an export endpoint, which is a real if thin category of product in its own right.

The vendor's own pricing meters the two halves separately, selling prospect profile enrichments and copy generation credits as distinct units on the entry plan, which is the clearest available evidence that the generation layer is an addition priced on top of a data pipeline rather than the substrate itself.

One genuine oversight mechanic is named and nothing surrounds it. The vendor lists a manual review and approval system as a product feature sitting between message generation and export, which puts a human gate in front of machine written copy before it reaches a sending tool, and that is more than most vendors at this size publish. Workspace partitioning by company is real too, running from one to five to unlimited across the tiers, which lets an agency separate client data properly.

Held at the middle band because the approval feature is a name without mechanics: nothing states who may approve, what they see, whether approval can be bypassed or skipped in bulk, or whether anything is recorded when it is. No role model, permission structure, audit trail or guardrail on what the generator may write appears anywhere.

No base model, provider, family, version or hosting arrangement appears anywhere on the page read. The product researches individual people, scores them, and writes copy in the customer's voice after learning their company and product, and none of that machinery is described beyond what it produces. Generation is metered commercially as credits, so the model layer is billed by the unit and specified nowhere, which is the same asymmetry recorded repeatedly this session.

There is no evidence of any kind. No customer is named, no logo appears, no testimonial is published, no case study exists, no review platform is cited and no usage or scale figure is given anywhere on the site. What pushes this to the floor rather than leaving it as a thin absence is that the one number on the page is an unsupported outcome promise attached to a price: the entry tier is positioned as suitable for a buyer who needs an extra fifteen to twenty sales calls a month, which tells a prospect what result to expect from a specific spend with nothing whatsoever behind it, no method, no sample, no customer and no qualification.

The vendor sits upstream of sending, which genuinely changes the shape of this axis without removing it. Nothing here dispatches mail or messages; the product hands enriched records and drafted copy to whatever tool the customer already sends from, so the statutory obligations attach at that downstream step rather than here. What the vendor does supply is the address, freshly verified, for a person who has not asked to be contacted, together with the message that will be sent to them.

No statute, consent position, suppression mechanism or list source restriction appears anywhere, and there is no terms of service or acceptable use document in which such a position could sit, so nothing passes any obligation to the customer either.

No privacy policy exists. The footer of this site contains a location line and a copyright notice and nothing else, with no privacy policy, cookie policy, terms of service or data protection contact linked anywhere on the page. That absence is established rather than merely unlocated, because the footer is complete and short.

Set against what the product does it is the floor: the platform scrapes profiles of named individuals from a professional network, attaches verified email addresses to them, generates written content about each person, stores the result and exports it onward, all without publishing a lawful basis, a retention period, a data subject rights route, a deletion mechanism or an address to send a request to. The operating entity is a United Kingdom registered company, so the domestic data protection regime applies to it directly.

Unusually candid about the source and silent on the right to use it. The vendor names its primary data source outright and describes the method plainly, listing a search scraper for a professional network's paid prospecting product as a bulleted feature, which is more provenance than most vendors in this category disclose and is creditable as honesty even where it is damaging elsewhere.

Beyond that the sourcing goes dark: the freshly verified email addresses attached during enrichment have no stated verification provider, coverage figure or accuracy rate, the auto enrichment layer names no partner, and no licence, permission or lawful basis is stated for collecting, storing or reselling any of it.

The vendor names a feature after the act of breaching a platform's terms. A search scraper for a professional network's paid prospecting product appears as the first bullet of the import step, so extracting search results from a subscription product that prohibits automated collection is not an inference from the architecture, it is a named capability on the homepage. Automated profile addition by webhook sits beside it.

No conformance position for that platform or any other appears anywhere on the site, and because no terms of service exist, nothing passes the obligation or the risk to the customer either, which leaves a buyer relying on the vendor with no written allocation of who answers if the source account is restricted. Nothing here is marketed as evasion, and that is the only thing separating this from the very worst of the band.

Nothing states whether customer content trains or improves any model, and there is no document anywhere in which such a statement could live. The material at stake is real on both sides: the platform is explicitly told to learn the customer's company and product in order to write in their voice, so it holds each account's positioning as working input, and it simultaneously holds scraped personal data on the third parties being written about.

No non use commitment, confidentiality undertaking, provider retention arrangement or separation statement between accounts was located, because no privacy policy, terms of service or processing addendum is published at any route on this site.

Machine written copy about a scraped individual is produced here and sent elsewhere, which splits the disclosure question across two vendors and leaves it unanswered by either. This product generates the personalised message and hands it to the customer's sending tool, so the moment of contact happens outside its walls, but the authorship question originates inside them.

No position on article 50 of the European artificial intelligence regulation appears anywhere, no disclosure line, identifier or sender transparency option is offered in the generated output, and no guidance is given to the customer about disclosure downstream. Held at the middle band rather than lower because the marketing sells conversion rather than undetectability, and because the manual approval step means a human sees the message before it leaves.

Genuinely open in architecture and entirely unnamed in partners. Data enters by webhook and leaves by an outbound endpoint that will post to any system the customer nominates, with export explicitly unlimited on every plan including the cheapest, which is a more open posture than most vendors twice this size take. Integration support is sold as an escalating service across the three tiers, rising to implementation help for the whole outreach system on the top plan.

Against that, not one integration partner, system of record, sending tool or platform is named anywhere on the page, there is no integrations page, no developer documentation, no marketplace and no published specification for the endpoint, so a buyer knows the product will export anywhere without being shown it exporting anywhere.

Nothing on hosting provider, country, region, data centre or residency option appears anywhere. Two geographic facts are published and both concern the company rather than the data: the operating entity is named as a United Kingdom registered limited company, and the product is described as made in Exmouth in the United Kingdom and Toronto in Canada.

That establishes jurisdiction and a development footprint across two countries, which is more than several vendors this session disclose, and tells a buyer nothing about where scraped prospect records, verified addresses or their own product positioning are actually stored.

No security disclosure of any kind exists, and the absence is established rather than unlocated because the footer of this site contains only a location line and a copyright notice, with no security page, trust centre, status page, certification badge or vulnerability contact anywhere on it. There is no audit, attestation, penetration test, control set or encryption statement.

Set against what the platform holds this is the floor: personal data scraped on named individuals at named employers, verified email addresses attached to them, the commercial positioning each customer supplies so the generator can write in their voice, and credentials or session access sufficient to run searches against a professional network's paid product.

A buyer can very nearly compute their bill from the page. Three tiers carry real monthly figures at sixty seven, one hundred and ninety seven and six hundred and ninety seven dollars, with an annual option and the saving stated as a percentage. Every gating quantity is published: prospect enrichments, copy generation credits, team member seats and the number of separate company workspaces, which is the allowance that actually matters to the agency buyer this product targets.

Fifty free credits open the account. The vendor states plainly that all features are available on all plans with only volume limits differing, which removes the feature gate that decides the real bill at most competitors, and it is the kind of commitment a buyer can hold them to.

Off the top band on one specific and load bearing gap: credit add ons are offered on two tiers with no rate published for them, so a customer who exceeds their monthly enrichment or generation allowance cannot find out what the next block costs, and no overage behaviour is stated either.

The product half is unusually good and the contractual half does not exist. Unlimited export of prospect data is listed as a feature on every plan including the entry tier, the vendor states separately that all prospect data and generated messages can be exported, and an outbound endpoint will post the whole set to any system the customer chooses, so bulk retrieval is a designed capability available at the lowest price rather than a concession negotiated at the end.

That is the pattern this index has repeatedly found separates vendors who genuinely have portability from those who claim it. Against it, there is no terms of service, no subscription agreement and no privacy policy anywhere on the site, so no post termination retrieval window, deletion timeline, retention period, data return obligation or cancellation procedure is published, and a departing customer has a working export button and no written right to anything.

Structurally a weak fit for this axis, graded on what is knowable, with the context carried here. This product operates no sending infrastructure: it connects no mailboxes, runs no campaigns and dispatches nothing, handing finished records to the customer's own sending tool where the reputation consequences land.

The one contribution it does make to deliverability is real and worth naming, in that email addresses are verified at the point of enrichment and described as freshly verified, which removes invalid records before they can generate bounces downstream. No verification provider, accuracy rate or bounce guarantee is published, and none of the warmup, authentication, pacing, complaint threshold or volume governance this axis normally measures applies to what the vendor actually does.

The buyer is named at the top of the page as agencies, individual sales professionals and go to market teams, and each tier carries a short statement of who it suits, running from a single operator wanting more calls a month through a small team or agency to an organisation.

The agency case is backed by a real product decision rather than a claim, since the number of separate company workspaces rises from one to five to unlimited across the tiers and a dedicated question addresses using the tool across client accounts, which is the feature an agency actually needs.

Beyond that the coverage half is empty: no region, country, language, industry or company size band is published, no customer count or usage figure appears, and there is no statement of who this is wrong for.