Salesmate

Model driven capability is real, named across several surfaces, and separately priced, which places it as a substantial layer on a deterministic platform rather than the platform itself.

An in application assistant appears from the middle tier, auto pilot and co pilot surfaces are published, and a distinct agent product is sold with its own pricing page separate from the relationship management tiers. Call transcription runs on the telephony side. Agent consumption is metered through a dedicated credit type sold as an add on.

The tier structure settles the placement. Across four plans the differentiators are pipelines, sequences, ticketing, quotes, custom objects, permissions, dialing and reporting, with automation metered in workflow execution credits. Those workflow credits meter rule execution rather than inference, and the agent credits are a separate purchase entirely.

So a customer can run the full relationship platform, the telephony, the campaigns and the automation with no model involvement at all, and the vendor has structured its pricing to make that possible.

The separation is honest rather than evasive: the agent product is priced where it belongs instead of being bundled to justify the tier.

Ask which actions consume agent credits and at what rate, since the add on description does not say.

Governance is provisioned properly and, more usefully, it is tiered legibly so a buyer can see exactly what controls cost.

Field based permissions and team management arrive at the middle tier alongside single sign on, which is unusually early: restricting visibility at field level rather than record level is a control most platforms reserve for enterprise. Custom duplicate management rules and service level agreements arrive at the third tier. Audit logs and address restriction sit at enterprise, together with custom validation rules and a quarterly business review.

Audit logging is the control that matters most for a system of record, and its placement at the top tier is worth a buyer noting rather than assuming.

Autonomy is bounded and metered. Workflow automations execute on customer authored rules with consumption capped by credit allowance per user, so runaway automation has a natural ceiling. The dialer and voicemail drop execute deliberate operator actions. The agent product introduces higher autonomy and is sold separately.

What is undescribed is agent oversight specifically: nothing states whether an agent action can be held for approval, what it declines to do, or how its activity appears in the audit log.

Ask whether agent actions are captured in audit logs and whether any approval step exists.

Two passes located no model, no provider, no version, no architecture and no evaluation for any of the several model driven surfaces this vendor markets.

That covers an in application assistant, an auto pilot surface, a co pilot surface, an agent product with its own pricing page, and call transcription. Five distinct capabilities, none attributed.

The metering makes the omission commercially awkward. Agent consumption is sold as a credit based add on, described as giving clear visibility into usage and predictable costs, and the description then says only that certain actions performed by agents will consume credits. Which actions, and at what rate, is not stated anywhere on the pricing page, so the add on promising predictable costs does not publish the numbers that would make them predictable.

Call transcription carries no accuracy figure and no supported language list, which matters for a platform sold into eleven named industries across global markets.

Nothing states whether customer records, email content or call audio reach a third party model provider, or whether any of it trains anything. For a system of record holding an entire company's customer relationships, that is the question a buyer should press hardest.

Ask which provider powers the assistant and the agents, and whether customer data trains any model.

The corroboration is broad, current and mostly third party derived, which is the right shape even without a headline case study.

Scale is stated at more than eight thousand five hundred businesses globally. Named customer logos include a major music label, an agricultural research firm and an insurance brokerage, spanning very different sizes and sectors.

The review evidence is displayed rather than asserted: ratings from three separate review platforms appear together, and nine award badges are shown, all dated to the current season. Those awards are review derived rather than purchased placements, and several measure things a buyer actually cares about, including estimated return, ease of setup, ease of doing business and likelihood to recommend. Two momentum and high performer placements sit alongside an industry award for best overall platform.

Currency is demonstrable rather than claimed. A product updates page carries release notes for the current month, so development is visibly active at the time of review.

A customer stories library is published separately.

What is missing is a quantified outcome attached to a named customer. No percentage, no time saved, no revenue figure and no methodology appears on the pricing page, so the evidence establishes satisfaction and scale rather than result.

Ask for a customer story in your industry with figures attached.

Three regulated channels are sold natively and the commercial plumbing is published while the compliance guidance is not.

The platform sells email campaigns and marketing automation, bulk text messaging, and from the third tier a power dialer with voicemail drop. Numbers are purchased inside the product from just over a dollar a month, and dedicated rate cards for both voice and messaging are published, so a customer can compute the cost of an outbound programme precisely.

What is absent is the rules layer. Two passes located no consent basis, no suppression list handling, no unsubscribe description or propagation across channels, no do not call screening, no calling hour restriction and no carrier registration guidance for bulk messaging, which is a prerequisite rather than a best practice in the vendor's primary market.

The power dialer and voicemail drop combination is the specific concern. Both are capabilities that attract regulatory attention, both arrive at a mid priced tier available to any self serve customer, and neither carries any published constraint.

A data protection compliance overview is published, and it addresses processing obligations rather than outbound conduct, which is a different regime.

Ask how opt outs are captured and enforced across email, messaging and voice, and who completes messaging registration.

The documentary set is complete for a vendor at this price and the technical commitments are specific rather than adjectival.

A data protection compliance overview is published as a standing resource, alongside terms, a privacy policy and a separate cookie notice, all linked from the footer.

The vendor's own security statement, submitted to a third party marketplace where such claims are reviewed, provides concrete detail: data at rest is encrypted with a named strong algorithm, customer databases replicate across multiple availability zones in near real time, customer data and files are logically separated in storage specifically to prevent leakage and accidental access, and data resides only on dedicated virtual private clouds and is not shared with third parties without consent.

Deletion is addressed directly and unusually. The vendor states that when a subscription ends, the system automatically removes every piece of client data within an hour. As a privacy commitment that is strong and rare, since most vendors retain indefinitely by default. As an exit risk it is severe, and it is graded there.

What two passes could not locate is a data processing agreement, a subprocessor list, a retention schedule for active accounts, or a stated processing location.

Ask for a processing agreement and the subprocessor list covering telephony and transcription.

Not applicable in the supplier sense and rated accordingly rather than penalised. Two passes located no contact database, no enrichment product, no email finder, no purchased records and no third party data sourcing. This vendor sells software, not data.

Every record in a deployment originates with the customer's own activity: web form submissions from their site, imports from their existing systems, contacts synchronised from their connected mailboxes, and inbound calls and messages to numbers they purchased. That is a materially cleaner provenance position than any data vendor in this index, and a buyer inherits no upstream licence terms or consent questions from the vendor.

Two residual questions arise from capabilities rather than from data supply. Call transcription captures and stores the speech of the person on the other end of the line, who is a customer or prospect rather than an employee, and the resulting transcript is new personal data created about them. And email synchronisation ingests correspondence content into the platform, including messages from people who never dealt with the vendor.

Neither is a licensing question and both are provenance questions about the analysed corpus.

Ask what is retained from call transcripts and synchronised correspondence, and for how long.

The extraction exposure common in this index is absent. Two passes located no scraping utility, no professional network automation and no profile harvesting. A browser plugin exists and operates inside a mail client through that provider's own extension framework, which is the sanctioned pattern.

Integrations span both major productivity suites, two commerce platforms, two accounting packages, a website platform, a telephony provider and a general automation platform, each accessed through its published interface. A partner marketplace and public interface documentation support the rest.

Where exposure genuinely sits is telephony and messaging, because those are native rather than integrated. Numbers are purchased through the platform and calls and bulk messages originate from them, which places the arrangement under carrier acceptable use terms and, for application to person messaging in the vendor's primary market, under a registration requirement. Nothing published states who holds that registration or whose terms bind the customer.

The consequence is practical rather than theoretical: an unregistered or non compliant sender sees messages filtered or surcharged by carriers, and the customer's numbers carry the reputation.

Ask who registers messaging traffic and whose carrier terms govern purchased numbers.

The technical disclosure is more specific than most vendors at this price offer, and it surfaced on a third party marketplace rather than prominently on the vendor's own site.

What the vendor states: data at rest encrypted with a named strong algorithm, databases replicated across multiple availability zones in near real time, customer data and files logically separated in storage explicitly to prevent leakage and accidental access, data held only on dedicated virtual private clouds, no third party sharing without consent, and complete automatic deletion within an hour of subscription end.

Logical separation stated as a deliberate anti leakage measure, and dedicated virtual private clouds, are real architectural commitments rather than assurances.

A security page exists on the vendor's own site and was not opened during this review, so it is recorded as existing rather than assessed. A community forum, a knowledge base and monthly release notes are maintained.

What two passes could not locate is an incident history, a responsible disclosure route, a status page, or any statement on whether customer records, correspondence or call transcripts are used to train models.

That last omission matters for a platform holding an entire company's customer relationships and its recorded conversations.

Ask whether customer data or transcripts train any model, and for a security disclosure contact.

The fundamentals are authentic. Calls and messages originate from numbers the customer purchased and owns, emails send from mailboxes they connected, and a representative places the calls. There is no synthetic voice on the core platform, no alias sending and no identity substitution.

Two capabilities introduce questions the vendor does not address.

Call transcription captures the speech of the person on the other end of the line. A substantial number of jurisdictions require both parties to consent before a call is recorded, and two passes located no recording notification capability, no default announcement and no guidance on which rule applies where a customer operates. The obligation lands on the customer.

Voicemail drop delivers a pre recorded message into a recipient's mailbox without the representative speaking. The message is in the representative's real voice and was recorded by them, so it is authentic in substance, and the recipient nonetheless receives something presented as a call that no one placed to them personally.

The separately sold agent product introduces higher exposure and its disclosure posture was not established in this review.

Ask what recording notice is available and whether the agent product identifies itself.

Broad, documented, and carrying one commercial commitment that most competitors do not make.

Integrations are stated on the pricing page to be included at no extra cost, on every plan. Connector fees and integration tiers are routine in this category, and removing them entirely is a real term rather than a marketing line.

Coverage spans both major productivity suites, two commerce platforms, two accounting packages, a website builder, a telephony provider and a general automation platform, with a fuller catalogue behind an integrations page and a partner marketplace on its own subdomain.

The developer surface is properly separated and documented, with interface documentation on a dedicated subdomain and access to public interfaces listed as included with every plan rather than gated to higher tiers.

The surrounding infrastructure is complete: mobile applications on both stores, a browser plugin for mail, a community forum, a knowledge base, video tutorials, webinars, and release notes published monthly. Three partner programmes cover solutions delivery, affiliates and startups.

The gaps are technical detail rather than absence. Two passes located no published rate limits, and no model context protocol server, which several smaller vendors in this cohort now ship.

Ask for interface rate limits and what synchronises back on a two way integration.

More architectural disclosure than most at this price and no answer to the question that matters for a regulated buyer.

What the vendor states is genuinely informative: the platform runs on cloud infrastructure providers, customer databases replicate across multiple availability zones in near real time, and customer data sits on dedicated virtual private clouds with files logically separated in storage. That describes a tenant isolation model and a resilience posture, and it is more than an assurance.

What is missing is location. Two passes located no provider named, no region, no country, no residency option, no single tenant deployment, no customer managed encryption keys and no subprocessor list.

The enterprise tier includes address restriction, which controls where users may connect from rather than where data rests, and a buyer should not read the one as the other.

The holdings make residency a real question rather than a formality. This is a system of record holding an entire customer base, full correspondence history, recorded calls and their transcripts, quotes and support cases, sold into eleven named industries including finance, insurance and education, several of which face sectoral requirements on data location.

Ask in which country records and call recordings are stored, whether any regional option exists, and for the subprocessor list.

The credential test resolves cleanly here, and unusually the vendor's own wording is what settles it.

In its published security statement the vendor writes that the platform is hosted with cloud infrastructure providers holding a service organisation control report and an information security certification, among others. That sentence attributes both credentials to the hosting providers rather than to this company, and it does so accurately: the vendor does not claim to hold them. Stating the position honestly is better than the badge rows found elsewhere in this index, and it leaves the buyer in the same place, because the landlord's certificates evidence the landlord's controls.

Two passes located no attestation held by this vendor, no auditor, no assessment period, no report request route, no trust portal, no completed questionnaire, no subprocessor list, no status page and no responsible disclosure route.

A security and reliability page exists on the vendor's own site and was not opened during this review, so it is recorded as existing rather than assessed and a buyer should read it directly.

The gap matters given the segment. Finance, insurance and education are named industry targets, and the enterprise tier offers audit logs and a quarterly business review, so the vendor is selling into buyers whose procurement will ask for a report.

Ask what this company itself holds or has in progress, and request the report.

Comprehensive, and it earns the band partly by publishing terms that work against the vendor.

Three tiers carry exact per user figures with a monthly and annual toggle and the saving stated, each with a full feature list and a separate comparison page. Workflow credit allowances are published per tier. Four add ons are named and described. A fifteen day trial allows unlimited users and full feature access.

Three disclosures stand out. Enterprise onboarding is published as starting from a stated figure, which almost no vendor does for a custom tier and which tells a buyer the implementation floor before any conversation. Telephony costs are published separately with number pricing from just over a dollar a month and dedicated rate cards for calls and messages. And integrations are stated to be included at no extra cost on every plan.

The refund policy is the most telling. It is published in full on the pricing page and it is unfavourable: all fees are non refundable, no credit is given for periods where an account went unused or partially used, and none is given where a customer deactivates or terminates mid interval, with credit toward an upgrade the single exception.

Publishing an adverse term in plain language, on the pricing page rather than buried in a contract, is exactly what this axis should reward.

Payment methods, invoicing and the bank transfer threshold are also stated.

The most consequential finding on this record, and it comes from the vendor's own words.

In its published security statement the vendor writes that there is no explicit option for a client to delete their data, and that when a subscription ends the system automatically removes every piece of data related to that client immediately, within an hour.

For a system of record that is severe. What disappears is the customer's entire relationship history: contacts, companies, deal pipelines, full email synchronisation history, call recordings and transcripts, support tickets, quotes and reporting. A grace period, an export window, a retention buffer or a recovery path would each mitigate it, and two passes located none.

It compounds with the refund policy on the same site, which states that fees are non refundable and that no credit is given where a customer terminates during an interval. So a customer cancelling forfeits the remaining paid term and, on the vendor's own account, their data within the hour.

What mitigates it is that public interfaces are included with every plan and documented on a dedicated subdomain, so a prepared customer can extract everything before cancelling. The risk falls entirely on the customer who cancels first and asks afterwards.

This statement appeared on a third party marketplace rather than the vendor's own pricing page and should be confirmed.

Ask in writing what grace period exists after cancellation and what export is provided.

One control is properly productised and the surrounding discipline is absent.

A dedicated sending address is sold as a named add on and described precisely, as protecting sender reputation and maintaining consistent deliverability for high volume campaigns, with one included in the enterprise plan. That is the correct control for a customer sending at volume from shared infrastructure, and productising it rather than burying it is right.

Email campaigns, marketing automation, full mailbox synchronisation and template tracking complete the sending workflow.

Everything else is missing. Two passes located no mailbox warm up, no domain health monitoring, no sender authentication record guidance, no inbox placement testing, no spam or content checking, no bounce or complaint handling surface, no address verification and no published delivery figure.

The messaging side is equally unaddressed. Bulk text messaging is sold with published per message rates and nothing on carrier registration, throttling, opt out keyword handling or delivery reporting.

A dedicated address without warm up guidance is a partial answer, since a new dedicated address has no reputation at all and must be built up carefully or it performs worse than shared infrastructure.

Ask what warm up guidance accompanies a dedicated address, and what delivery rates customers see.

The most specific segment definition in this cohort, and the specificity is evidenced by dedicated pages rather than asserted in a list.

Eleven industries carry their own pages: jewelry, retail, manufacturing, hospitality, consulting, real estate, insurance, finance, education, travel, and technology consulting and software. Jewelry and travel are the telling entries, because neither is a category a vendor writes about speculatively. Building for a jewelry retailer means understanding consignment, repairs and appraisal workflows, and a vendor only does that after winning customers there.

Five buyer roles are addressed separately, spanning representatives, sales leadership, customer success, marketing and business owners, which recognises that a platform combining sales, marketing and support is evaluated by different people for different reasons.

Nine use case pages describe outcomes rather than features.

The tier ladder matches the claim, running from a plan explicitly described for startups and small businesses through mid market to a custom enterprise tier with audit logs and a quarterly review, and scale is stated at more than eight thousand five hundred businesses globally. Mobile applications on both stores support field use in the trade and property verticals named.

Three partner programmes cover delivery partners, affiliates and startups.

Geography is the gap: no supported country list, no language statement and no regional data options.

Ask which countries telephony numbers are available in.