Salesgear

The research to draft loop is the product's central claim and it is model work rather than automation. The vendor states that it researches each prospect across a large number of sources and writes the outreach, and positions that against templated personalisation with the line about no longer sending a first name merge field. Deep research is priced as a first class metered unit at ten credits, the same as a phone number and ten times an email, which is the vendor's own statement that research is expensive to produce and valuable to the buyer.

Surrounding it are a writing assistant, a personalisation capability, an in product copilot and a development sold as an automated development representative, each with its own surface.

What keeps this off the top band is the substantial deterministic platform underneath. The sequencer, dialer, mailbox warming, domain health diagnostics, website visitor tracking, meeting scheduler and customer relationship synchronisation all function with no model involved, and the contact database is data retrieval rather than inference. A customer on a data only plan receives finders and research with no engagement layer at all, which is the vendor separating the two cleanly.

The model is genuinely constitutive of the differentiated half and additive to the rest.

Ask what a deep research credit actually returns and how the hundred plus sources are selected.

The vendor draws its autonomy boundary explicitly and publishes it, which in this category is worth more than any feature.

The stated position is that representatives still review every touch and simply skip the digging. That places the model on research and drafting and keeps a human on the send decision, and it is stated plainly on the home page rather than buried. In a category where competitors compete on how completely they remove the human, a vendor publishing the opposite boundary has taken a defensible position and made itself accountable to it.

The free research offer reinforces the same posture: the vendor invites a prospect to send fifty hard accounts for research at no cost with no signup required, which is research delivered as an artefact for a human to judge rather than an autonomous system let loose.

Sequences are configurable step by step and the dialer is representative operated rather than agentic.

What is undescribed is enforcement. Nothing states whether the review step can be bypassed, whether a representative can bulk approve a batch without reading it, or what audit record exists showing who approved what. A commitment that depends entirely on user diligence with no product level control behind it is a norm rather than a guardrail.

Ask whether the review step is enforceable and what the approval audit trail records.

Two passes located no model, no provider, no version, no architecture and no evaluation for any capability across the entire site.

The metering makes the opacity commercially awkward as well. Deep research costs ten credits, which is ten times an email lookup and equal to a phone number, so it is the most expensive single action a customer can take and the one whose output is least defined. Nothing states what a research returns, how long it takes, what the hundred plus sources are, how they are selected, or what happens when the research finds nothing useful. A buyer allocating a credit budget across finders and research has no basis for the research half of that decision.

The reply rate claim of three to five times is attributed to the research and writing quality, which is a claim about model output, and no evaluation supports it.

One element runs the other way and deserves credit. The model context protocol server exposes the product inside external assistants the customer already chooses and pays for, so for that surface the model identity is the customer's own decision rather than an undisclosed vendor choice. That is transparency by architecture rather than by disclosure.

Ask which provider powers research and drafting, and whether customer data reaches it.

Eleven customer logos are displayed and one substantial named testimonial is published from an operator who describes having built outbound motions at multiple companies, which reads as a credible practitioner rather than a manufactured quote.

The scale claim needs reading precisely. The vendor states more than a thousand sales leaders and representatives, which counts individuals rather than companies. For a per seat product that is a materially smaller base than a company count would imply, and the vendor has phrased it accurately rather than inflating it.

Two disclosures are unusually candid and are credited. The vendor publishes the assumption underpinning its own return calculator, roughly twelve minutes saved per researched account across forty eight working weeks, so a buyer can test the arithmetic rather than accept an output. And it publishes a self referential proof point, stating it runs the product on itself and books thirty five to forty meetings a month with it. Dogfooding stated with a number is rare.

What is missing is verification. The headline claim of three to five times reply rates carries no methodology, period, baseline or sample. No case study with attributed figures was located across two passes, and the independent review base is thin.

Ask what the reply rate multiple is measured against and over what sample.

Squarely applicable and substantially unaddressed. This product supplies contact data including direct phone numbers, then sends cold email at volume, runs professional network outreach and operates a dialer, which is three regulated channels in one account.

Two passes across the pricing page, product pages, solutions pages, use case pages and footer located no compliance surface whatsoever. Nothing on consent basis for contacting a supplied record, nothing on unsubscribe capture or suppression across campaigns, nothing on do not call screening or calling hour restrictions for the dialer, and nothing on messaging registration.

The absence is sharper because the vendor sells the data. A customer buying an email address or a phone number from this platform and dialling it is contacting a person the vendor identified, and the vendor publishes no basis on which that contact may lawfully be made in any jurisdiction.

The operating entity is Singapore registered and the customer base spans multiple continents, so the applicable regimes differ materially by target. A vendor selling globally with no jurisdictional guidance leaves that entire analysis to a buyer who is unlikely to know it is required.

The extensive how to library covers deliverability and reply rates and contains nothing on lawful sending.

Ask how suppression works across channels, and on what basis a supplied phone number may be dialled.

A privacy policy and terms of service are published, and a responsible disclosure page sits alongside them in the footer, which is a genuine artefact and is credited on the stewardship axis.

What two passes could not locate is everything a buyer's reviewer asks for after the policy itself: no data processing agreement, no subprocessor list, no retention period, no processing location, no regional framework named, and no route for a data subject other than general contact.

The holdings make those omissions consequential in two distinct directions. Outward, the vendor supplies personal data at scale, including direct telephone numbers, on individuals who have no relationship with it and were never asked. Inward, the platform connects to customer mailboxes over standard send and receive protocols, which means it holds credentials or tokens and processes correspondence content in order to sequence and track it.

Holding both sides simultaneously, a database of people who did not consent and live access to the mailboxes used to contact them, is the configuration that most needs a documented processing position, and none is published.

The Singapore entity selling into European and North American markets compounds it, because the transfer question has no stated answer.

Ask for the processing agreement, the retention period on supplied contact records, and where mailbox data is processed.

The vendor is a data supplier as a primary business rather than incidentally, and this is the axis where that matters most.

An email finder, a phone number finder and a contact database are core products with their own pages, metered by credit at published rates, with the entire data plan family existing to sell nothing else. Third party material reports a database in the hundreds of millions of records. Selling a phone number at ten times the price of an email is a statement that the vendor knows direct dials are the valuable and hard to source item.

Two passes located nothing about where any of it comes from. No source named, no aggregator, no licence basis, no consent basis, no coverage statement by geography, no accuracy or verification figure, no refresh cadence and no indemnity. For a company whose unit of sale is a record about a person, publishing nothing about the provenance of those records is the central omission.

One independent review, published by a competitor and therefore to be weighed rather than accepted, reports users citing unreliable contact data. The vendor publishes no accuracy figure to set against it either way.

The research capability compounds the question, since researching across many sources produces derived personal data whose basis is equally unstated.

Ask where records originate, what accuracy is achieved by geography, and what indemnity covers a provenance claim.

The professional network appears across the product in three distinct ways and none of them is explained.

An email finder specific to that network is a named product page. A browser extension is a named product page. Multichannel outreach includes automated activity on that network as a sequence channel alongside email and calls.

That platform does not license general contact extraction, email resolution from profiles, or automated outreach to third parties. Two passes located no partner programme, no sanctioned interface, no described mechanism and, notably, no trademark or non affiliation disclaimer of the kind other vendors in this cohort publish. Where a peer explicitly stated it holds no relationship with the platform, this vendor says nothing at all.

Independent review material notes users finding the network outreach options limited, which is consistent with automation operating inside constraints rather than through a sanctioned route.

The exposure lands on the customer, because it is their account and their credentials that connect, and enforcement restricts them rather than the vendor.

The mail side is cleaner, connecting through both major providers or any provider over standard protocols, which is the sanctioned path where one exists.

Ask how profile email resolution and network outreach are performed, under which programme, and who bears an account restriction.

One genuine artefact exists and it is worth naming before the gaps. A responsible disclosure page is published and linked from the site footer alongside the terms and privacy policy, which gives security researchers a route and signals the vendor expects to be tested. Many vendors of this size publish nothing of the kind.

Everything else is absent. Two passes located no security page, no encryption statement in transit or at rest, no access control description, no monitoring or logging position, no business continuity statement, no incident history, no status page and no subprocessor list.

The stewardship exposure is unusually broad because of what the platform touches. It holds live connections to customer mailboxes for both sending and receiving, so it can read correspondence. It holds a contact database of individuals. It runs research across many sources producing derived profiles of named people. And it exposes an interface into external assistants.

Against all of that, nothing states whether customer correspondence, research output or engagement data trains any model, whether tenants are isolated, how long anything is retained, or who at the vendor can access a customer's mailbox connection.

The disclosure page is a floor rather than a programme.

Ask whether mailbox content or research output trains any model, and what internal access controls apply to connected mailboxes.

The sending architecture is honest. Messages go from the customer's own connected mailbox, through both major providers or any provider over standard protocols, in the customer's own name and from their own domain. There is no shared sending infrastructure, no alias rotation, no synthetic voice and no automated reply agent answering a prospect on the customer's behalf.

The vendor's published human in the loop commitment reinforces it. If a representative reviews every touch before it sends, then the named sender has genuinely authorised the message, which is the substance of authenticity rather than its appearance.

The residual question concerns the research rather than the sending. A message that opens with a specific and accurate observation about the recipient's company reads as the product of human attention, and the vendor's own positioning contrasts it explicitly with obvious templating. The recipient cannot distinguish research a person did from research a model did in seconds, and while a human approved the send, the apparent effort behind the message is not what it appears.

That is a soft form of the problem rather than misrepresentation of identity, and it is why this sits mid band.

Website visitor tracking adds observation the recipient is not told about.

Ask whether any disclosure line can be configured, and what the visitor tracker records.

Broad for a company this size and genuinely open at the mail layer. Four customer relationship systems are integrated with a dedicated page each, a general automation platform covers the long tail, and a browser extension supports in context prospecting. Mail connects through both major providers or, importantly, any provider at all over standard send and receive protocols, which means a customer is not locked to a particular mailbox vendor.

The distinguishing element is a model context protocol server with its own page, exposing the product inside external assistants. This is the third vendor in the current sweep shipping one, and doing so from a small bootstrapped team is early. Alongside it sit an in product copilot and a published interface and partners programme, so the developer posture is deliberate rather than incidental.

A knowledge base is maintained on its own subdomain.

The gap is depth and documentation quality. Independent review material reports users finding documentation limited and integrations missing, and two passes located no interface reference with endpoints or rate limits, no marketplace listing, and no description of what synchronises back to a customer relationship system after an engagement.

Ask for the interface reference and what writes back on a reply or a booked meeting.

Two passes located no hosting provider, no region, no country, no residency option, no single tenant deployment, no customer managed encryption keys, no subprocessor list and no transfer mechanism.

The corporate position makes the silence consequential rather than routine. The operating entity is a Singapore private limited company, and the product is sold globally with comparison pages targeting North American incumbents and industry pages addressing sectors across several regions. A European or North American buyer therefore contracts with a Singapore entity and is told nothing about where anything rests.

The data crossing that unstated boundary is substantial and sensitive in two directions. Personal contact records including direct telephone numbers on individuals across many jurisdictions, and live access to customer mailboxes carrying commercial correspondence.

For a European buyer the transfer analysis cannot begin without a processing location, and none is stated anywhere. For a buyer in a regulated sector, mailbox access with an unknown processing location will not clear review.

The absence of a subprocessor list compounds it, since data enrichment, research across many sources and mail infrastructure are almost certainly performed with third parties whose identities are undisclosed.

Ask in which country contact data and mailbox content are processed and stored, and for the subprocessor list.

Two dedicated passes located no security page, no trust centre, no attestation, no certification, no auditor, no assessment period, no penetration test summary, no completed questionnaire, no subprocessor list and no status page. There is no credential to apply the test to because none is claimed anywhere.

The one security artefact is a responsible disclosure page, published and linked from the footer beside the terms and privacy policy. That is a real commitment, it gives researchers a route, and it is more than several records in this index offer. It is also a single page describing how to report a problem rather than any statement of what controls exist to prevent one.

The gap matters more here than for a lighter product because of the mailbox connection. A customer granting send and receive access to their business mail is handing over the most sensitive integration in the outbound stack, and the vendor publishes nothing about how that access is stored, scoped, monitored or revoked.

The vendor otherwise publishes extensively, with dozens of product, solution, comparison and guide pages and a knowledge base subdomain, so this is a choice about what to publish rather than an absence of capacity to publish.

Ask what attestation exists or is planned, and how mailbox credentials and tokens are stored and scoped.

Strong in mechanics and undermined by two contradictions on the pricing page itself.

What works is genuinely good. Four tiers are published with exact monthly prices and exact credit allowances, a permanently free tier requires no card, annual billing states a twenty percent saving, and the metered unit is defined precisely: one credit for an email, ten for a phone number, ten for a deep research. Publishing the consumption table is what makes a credit model forecastable and most competitors omit it. The questions section answers the awkward cases directly, stating that credits reset monthly and do not roll over, and describing exactly what happens when an allowance is exhausted.

The two plan families are explained clearly, including that an account may hold only one at a time and must contact support to switch, which is a real constraint disclosed rather than discovered.

The contradictions are on the same page. The free tier card states five credits while the questions section states fifty credits every month, a tenfold discrepancy on the entry offer. And the page headlines that all features are unlocked across plans with no restricted features, while the tier cards themselves differentiate basic filters from advanced filters.

A further practical gap: only the data plan prices rendered on retrieval, so the outreach family, which is the one most buyers here would want, is harder to price than the cheaper one.

Ask which free credit figure is correct and for the current outreach tier rates.

The commercial side is comparatively clean and the data side is undocumented.

No annual lock in is stated as a positioning claim, monthly plans are available across the range, and a permanently free tier means an account can be downgraded rather than closed. Removing term commitment is the most consequential portability feature a subscription vendor can offer and this one has done it deliberately.

A published interface and partners programme gives a technically capable customer a programmatic extraction route, and the model context protocol server offers a second path to the same records.

One small forfeiture is disclosed rather than hidden: credits reset monthly and do not roll over, so a customer cancelling mid cycle loses the remainder of that month's allowance. Stating it plainly is better than the alternative.

What two passes could not locate is any export route, file format, statement of what an export contains, or retention and deletion position after an account closes. For a platform holding sequence structures, engagement history, research outputs and a contact list assembled over time, none of that is addressed.

Mailbox derived data is partly mitigated, since sent mail and replies already exist in the customer's own mail system independently.

Ask what an export contains and what is retained after cancellation.

Deliverability is treated as a capability area in its own right rather than a warmup toggle, and the structure of the site says so. Three separate surfaces exist: a deliverability optimiser, a domain health centre and mailbox warming, each with its own page under solutions rather than buried in a feature list.

A domain health centre is the element that distinguishes this from a checkbox. Diagnosing the sending domain's own configuration and reputation, as a standing surface a customer returns to, is the correct place to catch a problem before it costs a campaign.

The educational layer reinforces it. Dedicated guides cover improving deliverability, avoiding the spam folder, increasing reply rates and finding verified addresses, which is a vendor teaching the discipline rather than only selling the tool.

The architecture helps too. Sending runs from the customer's own connected mailbox through both major providers or any provider over standard protocols, so reputation accrues to a domain the customer already owns.

What is absent is measurement. Two passes located no published delivery rate, placement figure, bounce rate or complaint rate, and no verification guarantee on the addresses the vendor itself sells, which is the obvious source of a bounce. The dialer side carries no caller reputation or spam labelling position.

Ask whether supplied addresses are verified before sending and what bounce rate is typical.

Segment is addressed with unusual precision through five named role pages: founders, engineers responsible for go to market systems, heads of growth, development representatives and revenue operations. Naming the go to market engineer as a distinct buyer is current positioning that most vendors have not caught up to, and addressing revenue operations separately from representatives shows the vendor understands who actually configures an outbound stack.

Four industry pages cover artificial intelligence companies, software, financial technology and agencies, which is a coherent technology and services focus rather than a scattered list.

Thirteen comparison pages target named competitors including the largest incumbents in both data and engagement, which is an explicit and confident market position.

The practical centre of gravity shows in the customer logos, which skew heavily toward Indian and wider Asian technology and services firms, consistent with a Singapore operating entity and a small bootstrapped team.

The base should be read carefully. More than a thousand sales leaders and representatives counts individuals, not companies, which for a per seat product is a modest installed base and the vendor has stated it accurately.

Gaps: no geographic coverage statement, no supported country list for phone data, no language statement.

Ask what contact coverage and accuracy look like in your specific territory.