1 of the 6 capabilities tracked on both differs.
| Measure | EvaluAgent | Surfboard |
|---|---|---|
| Starts at | $35/seat/mo | No longer sold |
| Charged per | Each agent seat | Not published |
| Included, where the other says no | 1 capability | 0 capabilities |
1 of the 6 capabilities tracked on both products differs.
| Capability | EvaluAgent | Surfboard |
|---|---|---|
| Voice | IncludedAll plans, as an imported channel rather than a live one. Speech-to-text on imported voice conversations, speaker diarisation separating agent from customer, custom vocabulary for brand and industry terms, and audio playback synchronised to the transcript. | Not published |
| Capability | EvaluAgent | Surfboard |
|---|---|---|
| Shared inbox | NoNo inbox. Conversations are imported from the desk or contact centre that handled them. | NoNo inbox. Surfboard read volume out of the ticketing and telephony systems and scheduled against it. |
| Ticket queues | NoNo ticket object of its own. | NoNo ticket object of its own. |
| SLA policies | Not published | IncludedPro and above, as SLA driven capacity planning: the forecast was built to hit a service level rather than to fill a rota. |
| Business hours | Not published | IncludedAll plans. Timezone management was listed in Scalers, which matters for the distributed support teams this was built for rather than for a single-site contact centre. |
| Capability | EvaluAgent | Surfboard |
|---|---|---|
| Autonomous resolution agent | NoNothing here answers customers. It evaluates AI agents rather than being one, and fabrication detection exists to flag "bot responses that hallucinate or invent information" produced by somebody else's agent. | NoNothing customer-facing. |
| Multilingual AI | IncludedAll plans. Transcription across major European, Asian and Latin American languages. | Not published |
| Capability | EvaluAgent | Surfboard |
|---|---|---|
| Summarization | IncludedConversation Intelligence tier only. AI-generated summaries of every interaction, alongside the feature the vendor calls Spotlight. | Not published |
| Auto-tagging | IncludedConversation Intelligence tier only. Automatic categorisation of why customers got in touch, plus insight topics and theme detection across conversation volume, with custom topics buildable by the customer. | Not published |
| Sentiment analysis | IncludedConversation Intelligence tier only, from $65 per user. Conversation-level sentiment scoring with trend tracking is one of seven capabilities marked absent from the AutoQM tier. | Not published |
| Capability | EvaluAgent | Surfboard |
|---|---|---|
| Public help center | NoNo knowledge base product. | NoNo knowledge base product. Surfboard ran its own product help centre, now offline with the domain. |
| Capability | EvaluAgent | Surfboard |
|---|---|---|
| Conversation QA scoring | IncludedAll plans. Custom scorecards for any team or use case, with manual review and calibration sessions alongside the automated scoring. | NoNo quality management at all, which separated Surfboard from Playvox and from the contact centre suites. It scheduled and forecast; it did not score conversations. |
| Automated QA coverage | IncludedAll plans, and the product's central claim: "Automated scoring on 100% of conversations". SmartScore AI scores qualitative line items with stated reasoning rather than only the objective ones, and auto-publish and auto-fail rules route the results. | Not published |
| Calibration | IncludedAll plans. Human-in-the-loop review and calibration sessions are listed against both published tiers. | Not published |
| Agent scorecards | IncludedAll plans. Individual and team-level performance dashboards with coaching workflows. | Not published |
| Forecasting | Not published | IncludedPro and above for anything driven by real demand. The forecasting comparison lists unlimited ticketing platform integrations, forecast by custom channel or inbox, SLA driven capacity planning, forecast versus actuals comparison, manual forecast adjustments, custom shrinkage levels, concurrency calculations and scenario planning tools. |
| Scheduling | Not published | IncludedAll plans, and the deepest part of the product. Reusable shift plans, rotating patterns, automated round robins, time off in lieu, split shifts, compliance with contracted hours, draft and publish control, overtime scheduling, availability collection, shift swap and change requests, and break scheduling. |
| Adherence tracking | Not published | IncludedSplit by plan. Tracking agent adherence as time-on-task appears in the reporting comparison at the lower tiers; "Real-time adherence and productivity" and real-time management dashboards were Enterprise only. |
| Capability | EvaluAgent | Surfboard |
|---|---|---|
| Custom reports | Not published | IncludedPro and above for performance and schedule analytics. The reporting set covered shrinkage, coverage accuracy analysis, agent utilisation and productivity alongside adherence. |
| Raw data export | Not published | IncludedTimesheet exports on every plan including Scalers. Custom data exports on Enterprise only. |
| Capability | EvaluAgent | Surfboard |
|---|---|---|
| SAML SSO | IncludedBoth published tiers, including the $35 entry. SSO, MFA and role-based access control are bundled at the floor price rather than held back for an enterprise tier, which is the opposite of the pattern across this category. | IncludedEnterprise only, described as "Enterprise grade security including SAML SSO" with no provider list. |
| Role-based access control | IncludedBoth published tiers, alongside SSO and MFA. | Not published |
| SOC 2 | IncludedSOC 2 Type 2, announced by the vendor and stated on the about page as "SOC 2. Plus ISO certification and a published security and trust posture." The trust centre runs on SafeBase. | Not publishedNo certification claim appears on the archived pricing page, and the site is gone, so nothing can be confirmed. "Enterprise grade security" at the top tier is marketing language rather than an attestation. |
| ISO 27001 | IncludedClaimed alongside SOC 2 Type 2. The about page names "ISO certification" without the standard number, so the specific scheme is asserted rather than itemised in that sentence. | Not published |
| GDPR | IncludedA UK company with a full public subprocessor register naming each vendor, its purpose and its processing region, including the AI model providers. That register is the strongest GDPR artefact in this category: most competitors name no model provider at all. | Not published |
| Data residency | IncludedCustomer-designated and unusually granular. AWS hosts "customer databases, uploaded objects, and application servers" in "EU, AUS or US at customer's designation", Deepgram transcribes in the same designated region, and Azure AI runs in "UK, US or AUS at customer's designation". Two subprocessors are pinned regardless of choice: Google Cloud's LLM and redaction services run in Europe, and Anthropic stores in the US. | Not publishedNever established. A UK company, but no hosting region statement is recoverable. |
| Capability | EvaluAgent | Surfboard |
|---|---|---|
| REST API | IncludedAdvertised as "an open API for everything else" beside the native connectors. No public reference, endpoint list or rate limit is published, so its scope cannot be assessed before a contract. | Not publishedNo API was advertised anywhere on the pricing page, and the archived captures reference no developer documentation. Unlike Playvox, which promoted a public API in writing, nothing here establishes that one existed. |
Both entries last checked Sep 9, 2026.
2 of 4 not published by EvaluAgent, checked
2 of 4 not published by Surfboard, checked