Frequently asked questions

Can a vendor pay to rank higher on Crail?

No. Vendor subscriptions pay for data completeness, freshness, and analytics dashboards — never for rank, review exposure, or review volume. Where sponsored placements exist, they are shown in a visually and structurally separate shelf, never blended into ranked or organic results.

Does Crail have an API or MCP server for agents?

Yes. JSON feeds are published at /api/vendors.json, /api/categories.json, and /api/comparisons.json, and a read-only MCP server (mcp/) exposes search_software, get_vendor, compare_vendors, get_pricing, get_reviews, list_categories, and recommend_software tools over the same data.

How are reviews on Crail verified?

Human reviews are tiered: Gold (LinkedIn OAuth or corporate-domain cross-check), Silver (corporate email OTP only), or flagged/unverified (personal email + screenshot, always shown with a visible badge). Crail also publishes agent_test_report reviews — a versioned, pre-published AI agent test protocol with a full public transcript. See /glossary/human-verified-review and /agent-test-protocol.

What is Crail?

Crail is an independent directory and comparison engine for enterprise software, built for both human buyers and AI procurement agents. Every vendor listing includes pricing, an agent-readiness score, and MCP/compliance data, all dated and sourced.

How is Crail different from G2 or Capterra?

G2 and Capterra were built for human clickstream SEO and vendor-paid lead-gen. Crail is built agent-native from the ground up: every page ships structured data and an MCP server so AI agents can query the catalog directly, and vendor payment (Crail Verified/Reach) is structurally separated from the agent-readiness score and ranking — a vendor can't pay to rank higher.

How often is vendor data updated?

Every field group on a vendor record carries its own lastVerifiedAt date and verificationMethod, so freshness is auditable per-field rather than asserted for the whole page. Vendors on a Crail Verified plan get a committed update SLA; unpaid listings are re-verified on a rolling basis.

What is the agent-readiness score?

A 0-100 score across six sub-dimensions — discoverability, machine-readable content, programmatic access, MCP/agent support, self-serve transactability, and transparency. See /methodology and /glossary/agent-readiness-score for the full breakdown.