Customer Service & Helpdesk
Crisp
Provenance unknown Report an error0–5 in half steps. 5 means the rubric's top anchor is met on the evidence.
by Crisp IM SAS · crisp.chat
Report an error on this page Is this your product? →
Read this page as one judge. Each weighs the same scores by what they care about.
The panel's verdict
Crisp, a customer-service helpdesk from Crisp IM SAS, is strongest on omnichannel, where scores cluster at 5-6: chat widget, email, WhatsApp and Messenger plus eight further channels land in one shared inbox, with a built-in CRM showing each person's data and past interactions. Knowledge and self-service scores 3-4: a knowledge base feeds an AI trained on articles, PDFs and the website, though judges found no public information on editorial workflow, versioning or deflection reporting. The lowest scores sit in customer data protection, scored 0-1 with no public information on retention, deletion paths, agent roles, a signable DPA or audit logs; ticketing core scores 2-3, with the shared inbox, no-code internal workflows and day-by-day metrics evidenced but assignment, SLA timers, routing rules, escalation and full-text search not. Integrations shows no spread at 2, its API evidenced only by a testimonial. Sovereignty scores 3-4 on "Made in Europe" and Crisp IM SAS, with no data-centre provider or subprocessor list named; pricing transparency scores 1 — a Crisp Plus plan and a 14-day, all-features trial with no card required, no figures.
Speaks for it
- Chat widget, email, WhatsApp and Messenger plus eight further channels land in one shared inbox, with omnichannel scored 5-6.
- A built-in CRM shows a person's data and past interactions whenever you communicate with them.
- A knowledge base makes customers more autonomous and trains the AI on articles, PDFs and website content.
- No-code internal workflows let teams speed up replies and reduce manual work, alongside day-by-day performance metrics.
- A 14-day free trial covers all Crisp features with no card required.
Held against it
- Customer data protection scored 0-1, with no public information on retention periods, deletion paths, agent roles, a signable DPA or audit logs.
- Ticketing core scored 2-3, with no public information on assignment, SLA timers, routing rules, escalation or full-text search across ticket bodies.
- The API is evidenced only by a customer testimonial, with no public information on documentation, webhooks, SSO or named CRM and shop integrations.
- Sovereignty rests on "Made in Europe" and Crisp IM SAS, with no public information on a data-centre provider or a subprocessor list.
- The "automate 50% of your inquiries" claim is not broken down into queue mechanics any judge could verify.
Best for
- You want chat widget, email, WhatsApp and Messenger plus more channels in one shared inbox with shared customer history.
- You want an AI copilot trained on your own website, PDFs and knowledge-base articles to answer routine inquiries.
- You prefer a built-in CRM with past interactions beside every conversation over connecting a CRM you already run.
- You want to try everything first — a 14-day trial of all Crisp features with no card required.
Avoid if
- You carry years of customer conversations and must evidence retention, deletion paths, agent roles or audit logging before signing.
- Your queues run on SLA timers, routing rules, escalation chains and per-queue reporting.
- You need documented API endpoints, webhooks or SSO to connect existing CRM, shop or identity systems.
The scores
Ticketing, queues & SLA
Show reasoningHide reasoning
How this is scored
The engine: how work is routed, prioritised, escalated and measured — and whether an agent can find the ticket they need among ten thousand.
0 — A shared mailbox with labels; no ownership, no status model, no history beyond the thread.
3 — Tickets with assignment and open/closed status, but routing is manual, there are no SLA timers and search covers subject lines only.
5 — Queues with rules-based routing, priorities, a working status model, SLA timers with breach warnings, and full-text search across ticket bodies.
8 — Business-hours-aware SLA policies per queue or customer, escalation chains, macros and triggers, merge and split, and reporting on first-response and resolution time by agent and queue.
10 — The workload is managed rather than merely tracked: capacity-aware assignment, SLA per contract with reporting an account manager could show a customer, audit of every status change, and search that finds the ticket from a half-remembered phrase.
The Support Lead
A shared inbox, no-code internal workflows, a built-in CRM with past interactions, and day-by-day performance metrics sketch a working engine. But we found no public information on SLA timers and breach warnings, rules-based routing, priorities, escalation chains, merge and split, or first-response and resolution reporting by agent and queue — and the "automate 50% of your inquiries" claim is not broken down into any queue mechanics I could verify. 1
The Agent Advocate
A shared inbox is named and no-code internal workflows exist, but I found no public information on assignment, a status model, routing rules, SLA timers, macros, or search across ticket bodies — the things that decide whether the hundredth reply is survivable. The only measurement shown is day-by-day team performance metrics, which tells me nothing about finding the old thread. 1
The Data Protection Officer
A shared inbox centralising inbound communications and no-code internal workflows are described, alongside day-by-day performance metrics. We found no public information on assignment and status models, SLA timers, priorities, escalation, or full-text search across ticket bodies. 1
The Shop Operator
A shared inbox that centralizes inbound communications is evidenced, with no-code internal workflows and day-by-day team metrics around it. We found no public information on ticket assignment, a status model, routing rules, SLA timers or search across ticket bodies, so from these pages I cannot tell who owns a where-is-my-parcel thread or when it breaches. 1
The Integrator
A shared inbox that centralises inbound communications, no-code internal workflows and day-by-day performance metrics are evidenced, which clears the labelled-mailbox floor. We found no public information on assignment and status, queues, rules-based routing, SLA timers, escalation or search across ticket bodies, so I cannot credit the engine's core. 1
The Skeptic
The page names a shared inbox, no-code internal workflows and day-by-day performance metrics, which is more than a labelled mailbox. We found no public information on assignment and status models, routing rules, priorities, SLA timers with breach warnings, escalation chains, or full-text search — none of the engine a queue manager actually needs is evidenced. 1
Channels in one queue
Show reasoningHide reasoning
How this is scored
Email, chat, phone, portal, messengers and social — judged on what actually lands in the same queue with the same history, not on how many channel logos the marketing page carries.
0 — Email only.
3 — Email plus one more channel, but the second lives in its own inbox: no shared history, and a customer who switches channel starts again.
5 — Email, a web form or portal and live chat all landing as tickets in one queue, with the customer's history visible whichever channel they used.
8 — The above plus telephony integration with call logging, at least one messenger (WhatsApp, Signal or similar) with its consent handling stated, and a customer portal where a requester can see their own tickets.
10 — Channel is an implementation detail: every channel including voice and messengers writes to one conversation with one history, agents answer from one screen, and the customer can move between channels mid-issue without repeating themselves.
The Support Lead
Twelve channels — chat widget across website, iOS, Android and React, email, WhatsApp, Messenger plus eight more — land in one shared inbox, and the CRM shows a customer's data and past interactions whichever channel they used. We found no public information on telephony with call logging, a requester-facing ticket portal, or WhatsApp consent handling, so the picture stops short of the fuller anchor. 1
The Agent Advocate
Chat widget, email, WhatsApp and Messenger are named as landing in one shared inbox, and the CRM shows past interactions whatever channel the customer used — that's the one-screen setup I want to see. I found no public information on telephony, a customer portal where requesters see their own tickets, or messenger consent handling, so it stops short of the full spread. 1
The Data Protection Officer
Chat widget, email, WhatsApp and Messenger plus eight further channels are named as landing in one shared inbox, and the built-in CRM shows past interactions whoever you communicate with. We found no public information on telephony with call logging, messenger consent handling, or a customer portal where requesters see their own tickets. 1
The Shop Operator
Chat widget, email, WhatsApp and Messenger are named as feeding one shared inbox, and the CRM shows each person's data and past interactions whenever you communicate with them — that is the WhatsApp-and-everything-else-in-one-queue test I run. We found no public information on telephony with call logging, messenger consent handling, a requester-facing portal, or any named shop or order channel. 1
The Integrator
Chat widget, email, WhatsApp and Messenger are named as landing in one shared inbox, and the CRM copy promises past interactions visible whenever you communicate with someone — one queue with shared history. We found no public information on telephony with call logging, a customer portal where requesters see their own tickets, or messenger consent handling, and the further eight channels are counted but not named. 1
The Skeptic
Twelve channels are advertised — chat widget, email, WhatsApp, Messenger plus eight others — and the shared-inbox and CRM copy claims every conversation and past interaction in one place. The eight extra channels go unnamed, and we found no public information on telephony with call logging, messenger consent handling, or a portal where requesters see their own tickets. 1
Knowledge base & deflection
Show reasoningHide reasoning
How this is scored
Whether the product reduces the number of tickets as well as organising them: public help centre, article workflow, suggestions to agents and to customers.
0 — No knowledge base; answers live in agents' heads and old tickets.
3 — A basic article list, public or internal, with no editorial workflow, no versioning and no link between articles and tickets.
5 — A searchable public help centre with categories, draft and publish states, and agents able to insert an article into a reply.
8 — Article suggestions to the customer before they submit and to the agent while they answer, multilingual articles, review dates that flag stale content, and reporting on which articles deflect.
10 — Knowledge is a managed asset: gaps identified from unanswered tickets, article performance measured against ticket volume by topic, versioned content with approval, and deflection reported as a number the team can act on.
The Support Lead
A knowledge base is named as a module to make customers more autonomous, and it doubles as training material for the AI copilot, which suggests content linked to answering. We found no public information on editorial workflow, versioning, review dates, suggestions before a customer submits, or reporting on which articles deflect. 1
The Agent Advocate
A knowledge base exists to make customers more autonomous, and the AI is trained on articles, PDFs and the website, so content feeds both customers and agents. I found no public information on editorial workflow, versioning, multilingual articles, or any deflection number, so I can't tell whether the base shrinks ticket volume or just holds text. 1
The Data Protection Officer
A knowledge base intended to make customers more autonomous exists and feeds an AI trained on website content, PDFs and articles, with a claim of automating half of inquiries. We found no public information on editorial workflow, versioning, article suggestions at submission, or deflection reporting. 1
The Shop Operator
A knowledge base meant to make customers more autonomous is evidenced, and the AI is trained on website content, imported PDFs and those articles. We found no public information on draft-and-publish workflow, article suggestions to customers or agents, stale-content review dates, or reporting tying articles to deflected tickets; the automate-fifty-percent figure is an AI claim, not a deflection number I could act on. 1
The Integrator
A knowledge base that customers use to become more autonomous is evidenced, the AI is trained on its articles, and the vendor claims to automate 50% of inquiries — that is a deflection claim. We found no public information on editorial workflow, versioning, article suggestions to agents at reply time, or deflection reporting, so this reads as an article store with an answering bot rather than a managed knowledge asset. 1
The Skeptic
A knowledge base module is marketed as making customers autonomous, and the AI trains on articles, PDFs and the website, so some link between content and answers exists. We found no public information on categories, draft and publish workflow, versioning, stale-content review dates, multilingual articles, or reporting on which articles deflect. 1
Customer data protection
Show reasoningHide reasoning
How this is scored
A ticket archive is personal data written by the data subject. Retention, deletion, access control, subject rights, and what the vendor does with attachments — judged on what executes rather than what is promised.
0 — No retention policy stated, no deletion path, agents all see everything, and no DPA is published.
3 — A DPA on request and coarse roles; deletion is described as something the customer arranges, and there is no stated retention period.
5 — A signable DPA published, configurable agent roles and queue-level visibility, a stated retention period, and deletion of a requester's data that can actually be executed.
8 — Automatic retention rules per queue or data category, attachment handling stated, an audit log of who opened which ticket, documented support for access and erasure requests, and pseudonymisation or redaction of ticket content.
10 — Built for a data-protection audit: retention executed and evidenced per category, field-level redaction, full audit trail, subprocessor list published, and the vendor's own support access to customer instances documented and consent-gated.
The Support Lead
We found no public information on retention periods, deletion paths, agent roles and queue-level visibility, audit logs, or a signable data protection agreement. "Made in Europe" speaks to origin, not to anything a data-protection audit would test, so nothing here is evidenced as executing. 1
The Agent Advocate
A ticket archive is personal data, and I found no public information on retention periods, deletion paths, agent roles, a signable DPA, or audit logs — nothing here is shown to execute. For a buyer carrying six years of customer conversations, that's the biggest hole on the page. 1
The Data Protection Officer
We found no public information on retention periods, an executable deletion path, agent role scoping or queue-level visibility, a signable data processing agreement, an audit log of who opened which ticket, or redaction. The AI features trained on the customer's own content make that silence the deciding fact, since the captured pages state no data-use policy either way. 1
The Shop Operator
We found no public information on retention periods, an executable deletion path for a requester's data, agent roles or queue-level visibility, a signable DPA, attachment handling, or an audit log of who opened which ticket. A ticket archive is customers writing about themselves, and none of those protections is evidenced on the captured pages. 1
The Integrator
We found no public information on retention periods, a deletion path, agent roles or queue-level visibility, a data-processing agreement, attachment handling or audit logging. A ticket archive is the data subject's own words, and on this capture there is nothing to assess against what actually executes. 1
The Skeptic
The captured page says nothing about retention periods, deletion paths, agent roles or queue-level visibility, a signable DPA, audit logs, attachment handling, or a published subprocessor list — we found no public information on any of it. Silence on every one of those controls leaves this at the bottom. 1
Integrations & API
Show reasoningHide reasoning
How this is scored
The systems a helpdesk has to reach — CRM, shop, order management, identity — and whether the API is documented for building or gated behind a partner conversation.
0 — No API and no named integrations; context is copied in by hand.
3 — A handful of native integrations and a read-mostly API, with no webhooks and no documented rate limits.
5 — Named integrations for common CRM and shop systems, a documented REST API with keys, webhooks for the core ticket events, and SSO.
8 — Maintained bidirectional integrations, customer context from other systems shown inside the ticket, SCIM provisioning, documented rate limits and a sandbox.
10 — A component rather than a destination: versioned API with a deprecation policy, event streaming both directions, an app framework for in-ticket extensions, and integrations the vendor maintains rather than lists.
The Support Lead
An API exists and one customer testimonial calls it a "mini-backoffice", but that is praise, not documentation. We found no public information on named CRM or shop integrations, webhooks, SSO, documented rate limits, or a sandbox — the built-in CRM softens the need but does not replace the reach. 1
The Agent Advocate
An API exists — a customer testimonial calls it a mini-backoffice for them — and there are desktop and mobile apps. But I found no public information on named CRM or shop integrations, documented endpoints, webhooks, SSO, or rate limits, so a buyer can't tell what they can actually wire up. 1
The Data Protection Officer
API availability is confirmed only through a customer testimonial describing it as a mini-backoffice, and the CRM is built in rather than reached through integration. We found no public information on API documentation, webhooks, rate limits, SSO, or named CRM, shop or identity integrations. 1
The Shop Operator
An API is available and a built-in CRM synchronizes customer data alongside conversations, but we found no public information on named CRM or shop-system integrations, documented endpoints, webhooks, rate limits, SSO or SCIM. My baseline — the order sitting inside the ticket without a second login — is not evidenced here. 1
The Integrator
The only API evidence is a customer testimonial about building a mini-backoffice — we found no public information on API documentation, rate limits, webhooks, SCIM, a sandbox or SSO. The CRM is built in rather than reached, which helps no buyer who already runs one, and no named integrations for CRM, shop or identity systems appear anywhere on the page. 1
The Skeptic
An API exists, evidenced only by a customer testimonial calling it "a mini-backoffice for us", and the messaging channels are named. We found no public information on API documentation, webhooks, rate limits, SSO or SCIM, or maintained CRM and shop integrations — the built-in CRM is native product, not a connector. 1
European sovereignty
panel opinion
Show reasoningHide reasoning
How this is scored
Where the ticket archive lives, who the contracting entity is, which subprocessors touch it, and whether vendor support can read customer data. Independently sourced by the sovereignty pipeline; scored here as this buyer weighs it.
0 — Non-EU vendor and contracting entity, hosting unstated or non-EU, subprocessors unnamed.
3 — EU hosting offered as an option while the contracting entity is non-EU, or the subprocessor list is absent.
5 — EU hosting as standard and an EU contracting entity, but parts of the chain — support tooling, analytics, AI features — are non-EU without an explained safeguard.
8 — EU or DACH hosting with a named data-centre provider, EU contracting entity, full subprocessor list published, and any non-EU processing named with its legal basis.
10 — Sovereign end to end and evidenced: vendor, entity, hosting and every subprocessor in the EU, certification published, and a self-hosted or private-cloud option for buyers who need the archive on their own infrastructure.
The Support Lead
The contracting entity is Crisp IM SAS and the vendor presents European origin, which is the right starting point for this buyer. But no data-centre provider is named, we found no public information on subprocessors or certification, and nothing on a self-hosted option — the chain is asserted rather than evidenced. 1
The Agent Advocate
"Made in Europe" and a European legal entity, Crisp IM SAS, point the right way, and that is all I have. I found no public information on the data-centre provider, a subprocessor list, certifications, or how vendor support access to customer data is handled — exactly the questions a European buyer will ask first. 1
The Data Protection Officer
European origin is stated and the contracting entity is Crisp IM SAS, which is more than nothing. Beyond that we found no public information on the data-centre provider, a subprocessor list, where the AI processing occurs, or whether vendor support can read customer data. 1
The Shop Operator
The contracting entity is Crisp IM SAS, hosting origin is given as Europe, and the product is presented as made in Europe. We found no public information on the data-centre provider, a published subprocessor list, or the vendor's own support access to customer instances, so the chain stops at the vendor's front door. 1
The Integrator
Crisp IM SAS is the contracting entity and "Made in Europe" states European origin, with no non-EU processing in evidence. The hosting claim is a slogan though: we found no public information on a named data-centre provider, a published subprocessor list, certification, or how the vendor's own support access to customer instances is governed. 1
The Skeptic
The contracting entity is Crisp IM SAS and the page says "Made in Europe", but that speaks to product origin, and we found no public information on hosting location, a named data-centre provider, or a published subprocessor list. An apparently European entity with unstated hosting and an absent subprocessor list cannot carry this higher. 1
Pricing transparency
not rated — the vendor publishes no price
Show reasoningHide reasoning
How this is scored
Whether a support lead can compute the real annual invoice for their agent count — including the channels and features they actually need — from public pages alone.
0 — No public prices at all; every tier is a sales conversation.
3 — A per-agent headline exists, but the tier where the needed channel or SLA feature begins is unstated, or light-agent and contact limits are not mentioned.
5 — Per-agent prices public for the main tiers with billing period stated, but at least one commonly needed capability (telephony, messengers, SLA policies) sits in an unpriced bundle.
8 — Every tier and add-on priced publicly with per-agent maths, billing period, minimum term and VAT treatment stated; only genuinely custom enterprise work lacks a number.
10 — Complete price computability: a calculator or table producing the annual invoice for a given agent count and channel selection, including usage-metered channels and overage.
The Support Lead
The captured page names a Crisp Plus plan that includes the AI helpdesk features and offers a 14-day, no-card trial of all features, but shows no figures: we found no public information on per-agent prices, billing period, VAT treatment, or add-on pricing. I could not begin to compute an annual invoice for my twelve agents from this. 1
The Agent Advocate
A 14-day free trial of all features with no card required is public and honest, and a "Crisp Plus plan" is named. But I found no public information on per-agent prices, billing period, minimum term, or VAT treatment — a support lead cannot compute even a rough annual invoice from this page. 1
The Data Protection Officer
A Crisp Plus plan is named and a fourteen-day all-features trial without a card is stated, but no price figures, billing period, per-agent maths or add-on prices appear anywhere on the captured pages. A support lead cannot compute an annual invoice from this. 1
The Shop Operator
The only pricing-adjacent facts are a fourteen-day free trial of all Crisp features with no card required, and the statement that AI Helpdesk features sit in the Crisp Plus plan. No per-agent price, billing period, minimum term or VAT treatment appears, so an annual invoice for a given agent count cannot be computed from these pages. 1
The Integrator
The only pricing information is the trial — "14 days free trial — All Crisp features — No card required" — plus the name of a Crisp Plus plan whose features are described but which carries no price. We found no public information on per-agent prices, tier boundaries, billing period, minimum term or VAT treatment, so no annual invoice could be computed from this page. 1
The Skeptic
Some structure is visible: "14 days free trial — All Crisp features — No card required", and AI helpdesk features "Included in the Crisp Plus plan", so the AI summariser at least is bundled rather than visibly metered. Beyond that we found no public information on per-agent or tier prices, billing period, minimum term or VAT treatment, and no annual invoice could be computed from this page. 1
European sovereignty — proven facts
0 of 4 dimensions provenBuilt only from facts shown on the vendor's own pages. A dimension we could not prove is left open, not scored as zero.
| Legal entity | Not determined ⚠ unverified | — | uncited Report an error |
|---|---|---|---|
| Ownership | Not determined | — | uncited Report an error |
| Data residency | Not determined ⚠ unverified | — | uncited Report an error |
| Subprocessors | Not determined ⚠ unverified | — | uncited Report an error |
Where this could be wrong
- Evidence ages. The oldest capture behind this page is from 29 Sep 2026. Vendors change pricing and policies without notice; every fact reflects its source as of the capture date shown in the registry.
- Weak sourcing — Data residency, Subprocessors, Legal entity. Not confirmed on the vendor’s own pages as captured.
- AI can misread a source. Extraction and judgement are automated; a citation guarantees traceability, not infallibility. If something here is wrong, say so — no account needed, every report is decided within 5 business days, and accepted corrections are published.
What we left out
A claim that does not survive our checks costs us the claim, not the page. This is what was taken off this one.
- 35 compliance facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material. Know more? Tell us
- 9 legal facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material. Know more? Tell us
- 5 product facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material. Know more? Tell us
- 2 pricing facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material. Know more? Tell us
- 1 data fact could not be confirmed on the vendor’s page as captured and was left out of this page and of the panel’s material. Know more? Tell us
- 1 subprocessors fact could not be confirmed on the vendor’s page as captured and was left out of this page and of the panel’s material. Know more? Tell us
- 3 sovereignty dimensions could not be confirmed on the vendor’s own pages and are shown as unknown. Know more? Tell us
- 6 of the readings below were written against an earlier fact sheet — a fact has been corrected, added or pulled since. Until the panel next runs on this product you are reading the older judgement. Know more? Tell us
Sources (14)
The pages every claim on this page was read from — each one checked, dated, and kept verifiable.
- 1 Vendor page crisp.chat Checked 29 Sep 2026 Details →
- 2 Terms of service — found from the homepage crisp.chat Checked 30 Sep 2026 Details →
- 3 Privacy policy — found from the homepage crisp.chat Checked 30 Sep 2026 Details →
- 4 Security / trust page — found from the homepage docs.crisp.chat Checked 30 Sep 2026 Details →
- 5 Ticketing, queues & SLA — found from sitemap crisp.chat Checked 1 Oct 2026 Details →
- 6 Ticketing, queues & SLA — found from sitemap help.crisp.chat Checked 1 Oct 2026 Details →
- 7 Channels in one queue — found from sitemap crisp.chat Checked 1 Oct 2026 Details →
- 8 Channels in one queue — found from sitemap crisp.chat Checked 1 Oct 2026 Details →
- 9 Knowledge base & deflection — found from sitemap crisp.chat Checked 1 Oct 2026 Details →
- 10 Knowledge base & deflection — found from sitemap crisp.chat Checked 1 Oct 2026 Details →
- 11 Customer data protection — found from sitemap help.crisp.chat Checked 1 Oct 2026 Details →
- 12 Customer data protection — found from sitemap help.crisp.chat Checked 1 Oct 2026 Details →
- 13 Integrations & API — found from sitemap crisp.chat Checked 1 Oct 2026 Details →
- 14 Integrations & API — found from sitemap docs.crisp.chat Checked 1 Oct 2026 Details →