Showpad

A fifteen year old content platform carrying an extensive and recently expanded model layer, and the removal test decides it. Strip Showpad Genie entirely and the customer keeps the product they bought: a governed content library with collections, pages and curated experiences, automation rules enforcing standards and permissions, divisions segmenting content across large user groups, training libraries with courses, paths and certifications, shared spaces for buyers, three dimensional and augmented reality asset hosting, offline mobile access for field sellers, analytics joining activity back to the record system, and seventy five or more integrations.

What disappears is authoring assistance, practice roleplay, conversational search across the library, generated analytics commentary and the agent tier. Those are substantial and the vendor markets them hard, and they are additions to a working platform rather than the platform itself. The vintage supports the reading: the company was founded in 2011 as a sales content and presentation tool, and the model layer arrived as a branded programme on top.

This sits with the other established enablement platforms in this index rather than with the model native entrants. Ask which capabilities cease to function without the model layer as opposed to becoming manual again.

A real governance vocabulary sitting in front of an agent tier whose limits are undescribed. The controls named are genuine and specific to this product's risk: automation rules enforce content standards and user permissions without manual oversight, divisions segment content and data across large user groups, permissions run to the file level, read only licences let content be shared without editing risk, and the vendor states that administrator controls allow guardrails to be customised so model output aligns to company standards.

The stated design principle is that output is grounded in governed data and that users receive only what their role and permissions allow, which is the right architecture for a platform whose failure mode is a seller showing a customer something they should not have. The gap is the agent tier.

The top package adds an agent studio, custom agents, agent templates, agent access to external sources and a field seller agent, and nothing published describes what any of them may do without a person approving it, what the guardrails actually constrain, or how a field seller agent behaves when it engages a buyer directly. Agent tooling also sits only in the highest package, so the sophistication of the controls a customer gets may vary with what they paid. Ask what a custom agent can do without human approval, what the administrator guardrails constrain, and whether agent capabilities are available below the top package.

Fifteen model driven capabilities named individually and nothing published about any of them beyond the name. The inventory is unusually granular, covering an assistant with conversation starters, text chat, voice chat, vision intelligence, deep research and podcast creation, alongside roleplay, authoring, library search, content governance, analytics commentary, asset question answering, page building, record system recommendations, an agent studio and a field seller agent.

Naming capabilities at that resolution is a form of disclosure and it is credited here. What none of it includes is a model provider, family, version, or any measure of quality. No accuracy, grounding or hallucination figure appears for the search and question answering features that sit on top of a governed content library, which is the specific risk this product creates, since a confidently wrong answer drawn from approved material is precisely what a regulated field seller must not repeat to a customer.

The vendor asserts that output is explainable, auditable and reliable, and describes no audit artefact, evaluation or measurement behind those three words. A page addressed to machine readers is published separately. Ask which models power the assistant and the search features, and what grounding accuracy has been measured against the content library.

Analyst evidence cited with unusual precision, customer evidence without figures. The vendor names its recognitions specifically rather than gesturing at them, giving the report, the category and the quarter in each case, covering a leading position in one research firm's revenue enablement evaluation in the third quarter of 2024 and its sales readiness evaluation in the fourth quarter of 2023, a leader placement in another firm's sales enablement assessment, a customers' choice designation from a peer review programme dated April 2024, and a leader position in a 2025 vendor quadrant.

Citing the quarter and the report title is better practice than most vendors manage and lets a reader locate the source. Twelve enterprise customers are named by logo and a healthcare customer story is published. What is missing is outcome. No study, population, period or method appears anywhere for any performance claim, no percentage lift or return figure is offered, and the customer story reached carried no numbers.

The analyst citations also predate the merger that reshaped the product, with the most recent evaluation running to 2025 and the platform now combining two previously separate products, so a buyer is reading recognition for something structurally different from what is sold today. Ask for outcome data with a stated customer population and period, and whether any analyst evaluation has covered the merged platform.

The axis applies narrowly to a platform that distributes content rather than running campaigns, and the scope is stated so a reader does not mistake narrowness for a gap. There is no sequencer, dialer or bulk sending capability here. What reaches outward are shared spaces delivered to buyers as links, content distributed to mobile devices, and integrations into the customer's own mail and sales engagement tools, so the sending obligations sit with those tools rather than this one.

The vendor's framing of the relevant control is access rather than consent, describing secure distribution across devices and shareable links to keep information out of the wrong hands, which is a confidentiality control and not a communications compliance one.

What is unaddressed is the buyer side of a shared space: nothing published describes what a recipient of a link is told about tracking, since engagement with shared content is captured and reported back to the seller, nor whether a buyer can decline that. For a product sold heavily into healthcare and life sciences, where what a seller may show and record is itself regulated, that is worth establishing. Ask what a buyer is told about engagement tracking in a shared space, and what controls exist for regulated industries.

Certification does real work here rather than sitting as decoration. The vendor names the international privacy information management standard alongside its information security certifications, and names an assurance standard used mainly by European auditors for service organisations, which is a combination that indicates the privacy programme has been examined rather than merely written.

Compliance with the European and Californian regimes is stated explicitly and a dedicated page exists for the former. Two published documents are separated in a way most vendors do not bother with, a privacy policy and a distinct set of privacy principles, and a responsible disclosure route is published as its own page.

The technical measures that bear on privacy are specific: encryption at rest and in transit to named standards, single sign on with automated provisioning and deprovisioning, permissions to file level, divisions segmenting data across user groups, and read only licence types that limit what a holder can do with content.

Held below the top band because the instruments a buyer would need to verify any of it were not reached: no processing addendum, sub processor list, retention schedule or transfer mechanism was located on this pass, and the trust centre the security page points to is hosted on the merged partner's domain rather than this brand's own, which a buyer signing with this entity should resolve. Ask for the processing addendum and sub processor list, and which legal entity the trust centre documents cover.

The material is the customer's own, which makes this axis narrow, and one recent capability widens it again. Content is uploaded and governed by the customer, training material is authored by the customer, and buyer engagement data is generated by the customer's own sellers and prospects, so there is no acquired dataset here and no licensing chain of the kind this axis usually examines.

The vendor's stated architecture reinforces that, grounding model output in governed customer data rather than in an outside corpus. The widening comes from the top package, which grants agents access to external sources, and nothing published describes what those sources are, who licenses them, what accuracy is warranted, or what happens when an agent draws a customer facing answer from one.

For a platform whose entire value proposition is that a seller can trust what the library tells them, the provenance of anything entering from outside that library is the question that matters. Authoring assistance raises a smaller version of the same point, since generated content becomes governed content once approved and nothing states what it was generated from. Ask what external sources agents may access and under what licence, and how content produced by authoring assistance is distinguished in the library from human authored material.

Official connections throughout with no evasion anywhere in the design, on a platform that runs its own infrastructure rather than reselling somebody else's application layer. Named integrations run through the documented routes of the platforms they connect to, covering a customer record system, a marketing automation platform, mail integrations and sales engagement tools, with more than seventy five pre built connections and a published open interface and software development kit.

Nothing scrapes, no browser extension harvests a third party network, and content distribution runs through the vendor's own mobile applications and hosted links rather than through another company's surface. Connections to model providers are handled through the vendor's own layer rather than by pushing the customer into direct provider relationships.

Held below the top band for the same reason most records in this index are: no conformance position is stated toward any of the platforms the product depends on, nothing describes what happens if a record system changes its interface terms or its data access policy, and the agent connectors added in the top package create a new dependency on assistant platform terms that is not addressed anywhere. Ask what the vendor's position is when a connected platform changes its interface terms, and whose terms govern content once an agent moves it into an external assistant.

An explicit training commitment, which is rare enough on this axis to be the finding. The vendor states plainly that customer data remains the customer's own and is never shared or used to train third party models. That is a specific and falsifiable commitment rather than a reassurance, and it names the exact concern an enablement buyer has, since the corpus at stake is a company's own proprietary sales content, pricing material and competitive positioning.

Two mechanisms are published alongside it. Administrator controls allow guardrails on model output to be customised to company standards, and output is stated to be grounded in governed data with users receiving only what their role and permissions allow, so the model operates inside the same access boundary as the person using it rather than across the whole library. The vendor also asserts that its models are explainable, auditable and reliable.

Held below the top band on evidence rank rather than on content. The commitment appears on a marketing page and no contractual anchor for it was located on this pass, so a buyer cannot yet tell whether the guarantee lives in the master agreement or only in the copy, and this record does not assert that the contract is silent because the contract was not read.

No retention period, model provider or evaluation record is published, and the audit implied by the word auditable has no described artefact. Ask where the training commitment lives contractually, and what retention applies to prompts and outputs.

The buyer facing surface is substantial and nothing describes what the buyer is told. Shared spaces put curated content in front of a prospect, engagement with that content is captured and reported back to the seller, a field seller agent is sold as engaging buyers directly, and authoring assistance produces material that reaches those buyers, so a person on the receiving end may be looking at generated content, inside a tracked environment, presented by or with the help of an agent.

None of that is disclosed to them in anything published. No statement describes whether a shared space discloses that engagement is tracked, whether content produced by authoring assistance is marked as generated, or what a buyer interacting with the field seller agent is told about what they are speaking to.

The controls the vendor does publish all point inward, protecting the customer from their own sellers rather than informing the buyer: read only licences, file level permissions, automation rules and divisions. Two things keep this in the middle band rather than lower.

The seller is a real named person throughout, with no persona or fabricated identity anywhere in the design, and the vendor sells heavily into regulated industries where content approval workflows exist precisely so that what a buyer sees has been checked. Ask what a recipient of a shared space is told about engagement tracking, and whether generated content and agent interactions are disclosed as such.

Broad, documented and genuinely extensible, with the extensibility sold rather than given. The vendor states more than seventy five pre built integrations alongside an open interface and software development kit, and the supporting apparatus is real: a developer portal on its own subdomain, a partner ecosystem site, a partner programme, a help centre, an academy and a customer community.

Named connections cover a customer record system, a marketing automation platform, mail integrations and sales engagement tools, with content syncing to keep assets current across systems. The agent tier adds connectors into an external assistant protocol and a hosted server for the same purpose, which is a forward looking surface few vendors in this category ship. The reason this sits below the top band is the packaging rather than the depth.

The developer bundle carrying the interface, the software development kit, webhooks and the reporting interface is gated to the middle package, and the assistant connectors and hosted server are gated to the highest, so a customer's ability to integrate at all is a purchase decision rather than a property of the platform, and neither package publishes a price. Ask which package is required for interface access, and what the assistant connectors expose.

Strong architectural claims about protection and nothing about location. The vendor describes securely distributing content across mobile devices and shareable links, offline access as a first class capability for field sellers, encryption at rest and in transit to named standards, automated provisioning and deprovisioning, and divisions that segment content and data across large user groups, which is a tenancy adjacent control described more concretely than most.

Compliance is stated to hold wherever a customer's teams operate and the vendor speaks of safeguarding interactions across borders. What none of that answers is where anything is. No hosting provider is named, no region or residency option is described, no isolation model is stated, and no recovery time or recovery point objective appears on any surface reached. The question carries weight for this particular buyer profile.

The company is European in origin with headquarters in two jurisdictions, sells into healthcare and life sciences where data location is frequently a contractual requirement, and enables offline copies of governed content on devices that leave the building, which is a residency question in its own right. The trust centre that might answer it sits on the merged partner's domain and was not reached. Ask which regions host content and analytics, whether regional storage is available, and what governs content cached on offline devices.

A named certification set, separated by platform in a live trust centre, with the documents behind it held back. The product page names five standards together: the international information security standard, the privacy information management standard, a service organisation control report at the continuous type, an assurance standard used mainly by European auditors for service organisations, and compliance with the European and Californian privacy regimes.

Controls are described specifically rather than in adjectives, covering encryption at rest and in transit to named algorithm and protocol versions, single sign on with automated provisioning, permissions to file level, divisions for segmenting large user populations, and read only licence types. A responsible disclosure route is published as its own page.

The trust centre the security page directs a buyer to is hosted on the domain of the platform this company merged with, and reading it resolves rather than deepens the ambiguity that routing creates. It states the certification scopes separately, giving this platform the information security standard, the privacy information management standard and the continuous type report, and the merged platform the information security standard and the continuous type report, and it explains that the modular nature of both platforms requires multiple audits through the year.

It also publishes named policies covering change management, business continuity and disaster recovery, and acceptable use, and states encryption standards and least privilege across both platforms. So the two brands share one trust programme that distinguishes what each holds rather than blurring them. Three things hold this below the top band. The substantive documents require a confidentiality agreement.

The trust centre answers the question of which platform's documentation to request by directing the buyer to their sales representative, so a buyer must first establish which platform they bought before they can ask for the right report. And the assurance standard named on the product page does not appear in the trust centre's own certification answer, so two of the vendor's own surfaces describe the certification set differently. Ask for the current report with its period and auditor, and for confirmation of which standards cover the platform being purchased.

Three packages itemised in detail, built cumulatively, and priced nowhere. The structure is published properly: a foundation package, a middle package described as everything in the first plus a named list, and a top package described the same way again, with each tier's additions enumerated across six capability groups covering assistant and agent features, content management, readiness, buyer engagement, analytics and extensions.

A separate table sets out what every package includes regardless of tier. The licensing unit is stated, being per user, and the vendor adds that further licence types exist for adjacent teams. Every package carries the same call to action, which is to request a quote, and no figure, band, range or currency appears anywhere on the page. Three specific gaps compound the absence of a number.

Three add on families are named and their contents were not readable on the page, so a buyer knows extra modules exist without knowing what they contain or cost. The interface and developer bundle is gated to the middle package and the assistant connectors to the top, so the cost of integrating at all depends on a tier whose price is withheld.

And the second licence class for adjacent teams, which is exactly the mechanism that decides the real bill for a large deployment, is mentioned in a single sentence with no rate. Independent aggregations of buyer contracts report annual totals in the tens of thousands with per user rates spread across a wide band, and they do not converge closely enough to record. Ask for the per user rate at each package, the adjacent team licence rate, and the contents and price of each add on family.

Better extraction machinery than most records in this category, and no statement about leaving. The routes are real and documented: an open interface and software development kit, webhooks, a reporting interface, and content syncing designed to keep assets current across a customer's other systems, which means the library and much of the analytic output can be pulled programmatically while a subscription is live.

That is materially more than a vendor offering a support ticket and a spreadsheet. Two things hold it in this band. Those routes are gated to the middle package and above, so a customer on the foundation package has no documented programmatic export at all, and the price of moving up is unpublished.

And nothing describes termination itself: no export scope or format is stated for the assets that matter most, which are the governed content library with its structure and permissions, the training courses and certification records, the shared space history and the engagement analytics that justify the spend, and no post termination access window, retention period or deletion timeline was located.

Recorded as a retrieval limit as well as a gap, since the online terms were not retrieved on this pass and a contractual provision may exist there. Ask what exports at termination and in what format, whether certification records and engagement history are included, and what is deleted and when.

The axis applies weakly to a platform that distributes content rather than sending campaigns, and the scope is stated rather than left for a reader to infer. No sequencer, dialer or bulk mail capability forms part of this product, so warmup, authentication, bounce handling and complaint thresholds have nothing to attach to, and where mail is involved it travels through the customer's own connected mail and sales engagement tools whose sending discipline is a separate question about a separate vendor.

The delivery mechanisms that do belong to this product are content distribution to mobile devices, shareable links to buyer facing spaces, and offline synchronisation to devices in the field, and the vendor addresses those as access and governance problems rather than as delivery ones.

The nearest thing to a discipline question here is content currency: material cached offline on a seller's device can diverge from the governed library, and the vendor markets automation rules and content syncing as the answer without describing how stale content on a disconnected device is expired or blocked. Ask how content already cached on an offline device is expired or withdrawn once the governed version changes.

One of the more clearly drawn coverage maps in this category, defined by selling motion rather than by company size. Six buyer roles each have a dedicated surface, covering marketing, field sellers, enablement, revenue leadership, revenue operations and channel selling, with channel selling in particular being a segment few competitors name at all.

Five industries each have their own page, covering healthcare and medical devices, manufacturing, consumer packaged goods, chemicals and technology, and the customer roster matches that claim rather than contradicting it, naming industrial, pharmaceutical and consumer goods enterprises rather than a generic technology logo wall.

The product design corroborates the positioning, since offline mobile access, three dimensional and augmented reality asset hosting and device level content distribution are capabilities that only matter to sellers who are physically in front of customers away from a desk. Scale is stated at more than two thousand organisations across more than fifty countries. Two limits keep it below the top band.

No employee count, revenue band or seat minimum is published, so a smaller organisation cannot tell whether it is served or priced out, and the packaging gives no clue because no package carries a price. And geographic coverage is a country count rather than a statement of where support, data or language coverage actually reaches. Ask what organisation size each package assumes, and which regions have local support and language coverage.