Front

Email, Front Chat (website widget), SMS, and social messaging are documented as native channels from the Professional tier up — but Starter is explicitly capped to a single channel type, so the full channel breadth isn't available at the entry price.

Front's core architecture is built around one shared, collaborative view of a conversation across channels, with internal comments and assignment happening on the same thread — genuine unification, consistent with the product's founding purpose.

Autopilot is a genuine, documented AI agent for complex workflows, and Copilot provides real-time agent assistance — both real capabilities, though structurally unbundled as paid add-ons below the Enterprise tier.

Front's own cited 'Topic accuracy' figure of around 85% measures categorization, not a true resolution rate, and independent analysis specifically notes that Autopilot's restricted knowledge sources (no Confluence, Google Docs, Notion, or Slack) lead to more escalations and a lower real-world resolution rate than that headline figure implies.

Autopilot learns automatically from Front's own knowledge base and past conversations within Front — genuine automatic learning, but explicitly restricted to Front-native sources, a real limitation for teams whose documentation lives elsewhere.

No detailed public documentation was found describing specific hallucination-prevention mechanisms (source citation, confidence thresholds) beyond the general observation that Autopilot's answers are constrained by, and limited to, its narrow knowledge sources.

Automation rules scale by tier (10 on Starter, 20 on Professional, unlimited plus 'Smart rules' on Enterprise) alongside Autopilot's documented handling of complex, multi-step workflows — genuine depth, with the most advanced tooling gated to the top tier.

Smart CSAT measures post-interaction satisfaction rather than real-time sentiment, and AI Topics auto-categorizes conversation subject matter rather than detecting frustration or urgency — no documentation of proactive, in-conversation sentiment detection was found.

This is the founding purpose of the product and its most mature capability — a single shared view of team communication with no separate ticketing system required, consistently the top-praised feature in independent reviews.

Autopilot is documented as escalating to a human when it can't resolve a conversation from its available knowledge, which is functional handoff — but we found no detailed documentation of an explicit confidence-threshold trigger comparable to what some competitors publish.

Internal comments and assignment directly on the customer-facing thread are the product's original, defining feature from 2013 onward — mature, well-reviewed, and consistently named by users as the reason they chose Front over a traditional inbox.

Macros are a documented, tier-gated feature (available from Professional up alongside automation rules) for consistent, fast agent responses.

A documented Shopify integration surfaces order information directly in the conversation ('where's my order? is a one-glance answer'), but this is described in comparatively general terms without the explicit refund, cancellation, or write-action detail more ecommerce-focused competitors document for their own Shopify integrations.

No dedicated WooCommerce integration was found in the materials reviewed.

Front Chat can be embedded on a WordPress site via the universal JavaScript snippet, but no dedicated first-party WordPress plugin was found — it's the same generic embed used on any website.

No dedicated Wix app was found; installation would rely on the same generic embed snippet used elsewhere.

No dedicated Squarespace integration was found.

No dedicated Webflow integration was found.

Both Zapier and Make are documented as supported integration paths for less common, long-tail connections — solid no-code coverage, though positioned as the fallback for tools without a dedicated Front app rather than a deeply built-out native trigger/action set.

Both Salesforce and HubSpot integrations are documented as genuinely bidirectional — changes in either Front or the CRM stay synced automatically, with contact and deal context visible directly in the conversation. This is deeper than the one-directional CRM sync several competitors in this category offer.

The Shopify integration surfaces order status directly in the conversation for a quick answer, but documentation is comparatively shallow versus competitors that explicitly detail refund or cancellation actions from inside the tool.

A multi-language knowledge base is documented as an Enterprise-tier feature, but no automatic language detection or live-translation capability for conversations themselves was found documented at any tier.

Basic analytics ship on Starter, advanced analytics arrive at Professional, and Enterprise adds AI-powered Smart QA (quality review) and Smart CSAT (satisfaction scoring) — genuine depth concentrated at the top tier.

Every tier and every AI add-on is individually, clearly published with exact per-seat and per-conversation pricing — genuinely one of the more transparent pricing pages in this category, even though the total cost at scale requires real modeling.

160+ documented integrations plus a public API with purchasable rate-limit increases ($200 per additional 100 requests/minute/month) — a mature, genuinely extensible developer platform.