Klue

What it is

Klue is a competitive intelligence and win-loss platform, founded in Vancouver in 2015. A product marketing or CI team uses it to collect competitor news, pricing changes, and field intel, curate that into battlecards, and push the cards to sellers inside Salesforce, Slack, Microsoft Teams, and the browser. Win-loss is the second half of the product. Klue bought DoubleCheck Research, a win-loss interview firm, in January 2023, then Goldpan.ai (March 2025) and Ignition (September 2025) to automate that work with AI. In April 2026 Gartner named Klue a Leader in its first Magic Quadrant for Competitive and Market Intelligence Platforms, alongside Crayon. It holds 4.7/5 on G2 across roughly 440 reviews.

Two months later it became the source of a supply-chain breach. Any buying decision about Klue in late 2026 starts with that incident, so this page covers it in full under Watch-outs.

Why it shows up in RevOps stacks

• Battlecards where reps already work. Cards surface in a Salesforce opportunity, in Slack and Teams, and through a browser extension, so the rep does not open a separate intel hub. Ask Klue answers “how do we beat X on pricing?” in Slack or Salesforce from the curated card set rather than from the open web.
• Call-driven deal coaching. With Gong or Chorus connected, Deal Tips emails the rep a competitor-specific briefing shortly after a rival is mentioned on a call. Auto Insights (October 2025) runs the same analysis continuously across open deals, so RevOps can see which competitor shows up in which segment and at which stage.
• Win-loss without a research agency. The Win-Loss AI Interviewer runs short voice interviews with buyers and sellers on closed deals, and the results feed the same platform as the battlecards. This is the piece Klue assembled from DoubleCheck, Goldpan, and Ignition, and it is the clearest gap between Klue and a DIY setup.
• An MCP server with write access. The Klue MCP server (December 2025) lets internal agents search intel, pull a specific battlecard, and retrieve pricing and objection-handling content. Klue says it tested it with Claude Desktop, OpenAI Agent Builder, and Copilot Studio. Admin workflows can also create and update Klue cards through it.

Pricing

Klue publishes no prices and has no self-serve plan or free tier. Every contract is quoted.

• What teams pay. Vendr puts the median Klue contract at $30,000 a year across 106 purchases, in a range of roughly $16,000 to $60,000, with buyers saving 18% against the first quote on average.
• Crayon costs the same. Vendr’s median for Crayon is also $30,000 a year (93 purchases, $12,700 to $46,000, 20% average savings). Price will not separate the two finalists; who has a credible alternative at the table will.
• Budget beyond the license. Klue’s value depends on someone curating the cards. If you have no PMM or CI owner with at least a quarter of their week for it, add that headcount to the business case, or buy something smaller.

Best for

• Product marketing and competitive intelligence leads at B2B companies with 30+ quota-carrying reps and 5 or more competitors that come up in real deals, who need cards inside Salesforce and Slack and a win-loss program on the same contract
• RevOps leaders who already run Gong or Chorus and want competitor mentions on calls turned into per-deal briefings without building the pipeline themselves
• Not for teams under about 20 reps without a dedicated CI owner (the $30,000 median buys a library nobody maintains), and not for analysts whose job is market sizing, earnings, and expert-call research rather than sales enablement (look at AlphaSense)

Alternatives — and when to pick them instead

• Crayon — the other Leader in Gartner’s 2026 Magic Quadrant and the rival in nearly every Klue deal, at the same $30,000 Vendr median. Crayon’s centre of gravity is automated monitoring of competitor websites and public sources, and it shipped an official competitive intelligence MCP server in September 2025, three months before Klue. Pick Crayon when breadth of automated tracking across many competitors matters more than win-loss, or when your security review will not approve Klue’s re-connection to Salesforce yet.
• AlphaSense — the largest vendor in the category by revenue (past $500 million ARR in October 2025) and also in the Gartner MQ. It searches filings, earnings calls, broker research, and expert transcripts. Pick AlphaSense when the buyer is strategy, corporate development, or product leadership asking market questions, not reps asking how to win a deal.
• DIY on your own call data — the fastest-growing option in practice, because most teams already pay for Gong and Claude or ChatGPT. Our competitive-battlecard-skill builds cards from your own call transcripts, and competitive-intel-tracker-n8n watches competitor pages and sends a weekly digest for a few dollars a month in model costs. Pick DIY when you have a technical RevOps owner, fewer than 20 reps, and no budget for a $30,000 platform. You give up the win-loss interviewer, the Salesforce-native card UI, and coverage of review sites that block scraping.

Watch-outs

• The 2026 breach ran through Klue’s Salesforce integration. On June 11-12, 2026 an attacker used a compromised legacy credential tied to a Klue integration service to obtain customers’ OAuth tokens and export Salesforce data: contact names, emails, titles, phone numbers, and account records. At least nine companies disclosed impact, including HackerOne, Huntress, Jamf, OneTrust, Recorded Future, Snyk, and Tanium. Klue disabled its Salesforce, HubSpot, Gong, Chorus, Clari, Slack, Zoom, SharePoint, and Google Drive integrations the same day, published a CrowdStrike investigation summary on July 1, and reported Salesforce and Gong reinstating it in their marketplaces on July 27. Guard: before reconnecting, get the CrowdStrike summary and Klue’s remediation list in writing (static egress IPs, PKCE platform-wide, shorter OAuth token lifetimes, no long-lived GitHub tokens), reconnect through a dedicated Salesforce integration user with read access limited to the objects Klue actually needs, allow-list Klue’s published egress IPs, and treat any phishing that impersonates your own staff as a likely use of the stolen contact data.
• The library decays without an owner. Users on G2 and in independent comparisons repeat the same complaint: Klue needs a dedicated CI or PMM owner, and a densely populated instance gets hard to navigate. Guard: name one owner before signing, cap the card set at the competitors that appear in at least 5% of opportunities, and retire any card not viewed in two quarters.
• AI agents can write to your card library. Compete Agent drafts intel and the MCP server can create and update cards, so an unreviewed agent can put a wrong price or claim in front of every rep. Guard: route agent-created cards through a review state before they publish, and connect MCP clients through a read-only role until that review step exists.
• Opaque pricing and renewal uplift. With no list price, the first quote is the anchor. Guard: bring the Vendr median and a parallel Crayon quote to the first call, ask for a multi-year cap on renewal increases, and cite the 2026 incident when you ask for a credit or a price hold at renewal.