LinkedIn Comments

A prospect's comments on someone else's LinkedIn post.

Overview

LinkedIn Comments capture when a prospect engages with someone else's post — often revealing more about their thinking than their own posts do. On average, LinkedIn users produce comments 1.2 - 2x more frequently than they post.

We track comment activity across millions of contacts, prioritizing substantive comments where prospects ask questions, share opinions, or reveal pain points. Each comment is analyzed for intent, pain points, initiatives, and technologies mentioned, with full context from the parent post.

The result: high-intent engagement signals that show what topics a prospect cares about, what they're evaluating, and how they think—perfect for starting a conversation that feels relevant, not random.

Schema

{
  "signal_id": "b99a32af-43fb-554f-8ec3-43ce5a9228fa",
  "signal_type": "linkedin-comments-contact",
  "signal_subtype": "linkedinPostComment",
  "detected_at": "2026-09-20T23:48:09.116Z",
  "association": "contact",
  "contact": { // The contact who commented 
    "email": "[email protected]",
    "first_name": "David",
    "last_name": "Barajas",
    "full_name": "David Barajas",
    "job_title": "Senior Software Engineer, Billing and Monetization",
    "linkedin_url": "https://www.linkedin.com/in/david-barajas-swe" // Recommended match key 
  },
  "company": { // We also enrich with the contact's employer
    "name": "Calendly",
    "domain": "calendly.com",
    "linkedin_url": "linkedin.com/company/calendly",
    "industries": ["Software Development"],
    "employee_count_low": 1001,
    "employee_count_high": 5000,
    "revenue": "250 Million to 500 Million"
  },
  "data": { // The signal data itself 
    "comment_text": "I think the challenge is framing around what customer success looks like. Most of a SaaS DAU or MAU counts come from app usage. Logins to the site, the mobile app, interactions with the dashboard. If you let users consume you as an MCP/API and let someone else own the UX, you give up a lot of upsell opportunities. Being invisible has a cost that people start thinking \"Do I really need that service?\". I think it's a tricky balance to play but I agree. As a consumer I want everything to plug and play into a central hub. Business growth is at odds with that in the traditional sense.", // The prospect's own words - quote them in outreach
    "comment_summary": "Engineer agrees agent-first consumption is coming but warns invisible products lose upsell leverage.",
    "comment_url": "https://www.linkedin.com/feed/update/urn:li:activity:7506838202792615936?commentUrn=urn%3Ali%3Acomment%3A%28activity%3A7506838202792615936%2C7507527351925108737%29&dashCommentUrn=urn%3Ali%3Afsd_comment%3A%287507527351925108737%2Curn%3Ali%3Aactivity%3A7506838202792615936%29",
    "comment_posted_at": "2026-09-20T19:53:23.813Z",
    "comment_num_likes": 2,
    "comment_num_comments": 3,
    "comment_intent": "insight", // question, disagreement, celebration, praise, tagging, insight, recommendation
    "signal_quality": 0.85, // 0-1 scale, higher = more actionable
    "relationship_context": { // Inferred relationship between the commenter and the poster
      "inferred_relationship": "industry_peer",
      "confidence": 0.7
    },
    "pain_points": [
      {
        "topic": "losing upsell opportunities when others own the UX",
        "intensity": 0.7
      },
      {
        "topic": "invisible products get questioned at renewal",
        "intensity": 0.65
      }
    ],
    "initiatives": [
      {
        "topic": "balancing MCP and API exposure with product engagement",
        "urgency": 0.6 // 0-1 scale
      }
    ],
    "technologies_mentioned": [
      {
        "name": "MCP",
        "status": "considering"
      }
    ],
    "competitors_mentioned": [],
    "parent_post": { // The post they commented on - same AI enrichment as LinkedIn Post signals
      "signal_id": "2f3e2c35-8b9a-5578-a09c-cda70baf2d46", // Cross-reference to the parent post signal
      "post_url": "https://www.linkedin.com/feed/update/urn:li:activity:7506838202792615936/",
      "post_text": "If you're still focused on building AI into your product, you're missing the point.\n\nI just opened an app I use at work almost everyday.\n\nThey had a pop-up announcing new AI functionality in their product.\n\nI literally do not care.\n\nAll I care about is how I can get your API's/MCP's exposed so I can plug them into my universal agents that run everything for me.\n\nCan't stop thinking about how many eng hours were wasted on getting that feature live...\n\nThe big players have already won the race for the AI homepage.\n\nIf you're not Anthropic, SpaceX, Meta, etc., you shouldn't be spending time developing AI tools in your platform.\n\nInstead, you should be focusing on how to make your platform play well with agents. \nie. Instinct, GrokBot, Muse, Claude Agents, etc.\n\nAt this point, if I can't use your product/solution via my own internal tooling, I'm not buying it. \n\nSo think less about how to make AI work INSIDE your product, and instead think about how to make AI work OUTSIDE of your product.",
      "post_summary": "RevOps leader urges vendors to expose APIs and MCPs for agents over in-product AI.",
      "posted_date": "2026-09-18T22:14:57.855Z",
      "num_likes": 163,
      "num_comments": 71,
      "poster_name": "Mike Christensen", // The person whose post drew the comment
      "poster_job_title": "Head of Revenue Operations",
      "poster_company_name": "Redo",
      "poster_company_description": "Redo helps online retailers optimize their post-purchase operations by consolidating tech stacks and enhancing returns management, order tracking, and customer service with AI.",
      "pain_points": [
        {
          "topic": "engineering hours wasted on unused in-product AI features",
          "intensity": 0.7
        },
        {
          "topic": "products unusable through agent-based internal tooling",
          "intensity": 0.65
        }
      ],
      "initiatives": [
        {
          "topic": "plugging vendor APIs and MCPs into universal agents",
          "urgency": 0.85
        },
        {
          "topic": "evaluating products on agent interoperability before buying",
          "urgency": 0.7
        }
      ],
      "technologies_mentioned": [
        {
          "name": "MCP",
          "status": "using"
        },
        {
          "name": "Claude Agents",
          "status": "using"
        },
        {
          "name": "Instinct",
          "status": "considering"
        },
        {
          "name": "GrokBot",
          "status": "considering"
        },
        {
          "name": "Muse",
          "status": "considering"
        }
      ],
      "competitors_mentioned": []
    }
  }
}

Field Reference

Core Fields

FieldTypeDescription
signal_idstring (UUID v4)Unique identifier for this signal
signal_typestringAlways "linkedin-comments-contact"
signal_subtypestringAlways "linkedinPostComment"
detected_atstring (ISO 8601)Timestamp when signal was detected
associationstringAlways "contact"

Contact Object

FieldTypeDescription
contact.first_namestringContact's first name
contact.last_namestringContact's last name
contact.full_namestringContact's full name
contact.emailstringContact's email address
contact.job_titlestringContact's job title
contact.senioritystringSeniority level, when resolved (e.g., "Cxo", "Director")
contact.departmentstringDepartment, when resolved (e.g., "Marketing", "Operations")
contact.linkedin_urlstringContact's LinkedIn URL

Company Object

FieldTypeDescription
company.namestringCompany name
company.domainstringCompany website domain
company.linkedin_urlstringLinkedIn company URL
company.industriesarray[string]Industry classifications
company.employee_count_lowintegerLower bound of employee count
company.employee_count_highintegerUpper bound of employee count
company.descriptionstringCompany description, when resolved
company.revenuestringAnnual revenue bucket, when resolved (e.g., "250 Million to 500 Million")

Comment Data

FieldTypeDescription
data.comment_summarystringAI-generated summary of the comment
data.comment_textstringFull text of the prospect's comment
data.comment_urlstringDirect URL to the comment
data.comment_num_likesintegerNumber of likes the comment received
data.comment_num_commentsintegerNumber of replies the comment received
data.comment_posted_atstring (ISO 8601)When the comment was posted on LinkedIn
data.comment_intentstringIntent classification (question, disagreement, celebration, praise, tagging, insight, recommendation)
data.signal_qualityfloatQuality score (0.0-1.0) - higher = more actionable
data.relationship_contextobjectInferred relationship to poster
data.relationship_context.inferred_relationshipstringRelationship type (competitor, colleague, unknown, prospect, vendor)
data.relationship_context.confidencefloatConfidence score (0.0-1.0)
data.pain_pointsarray[object]Identified pain points
data.pain_points[].topicstringPain point topic
data.pain_points[].intensityfloatIntensity score (0.0-1.0)
data.initiativesarray[object]Identified initiatives
data.initiatives[].topicstringInitiative topic
data.initiatives[].urgencyfloatUrgency score (0.0-1.0)
data.technologies_mentionedarray[object]Technologies mentioned
data.technologies_mentioned[].namestringTechnology name
data.technologies_mentioned[].statusstringStatus (evaluating, using, implemented, migrating_from, migrating_to, churned, considering, integrated, building_on, hiring_for)
data.competitors_mentionedarray[object]Competitors mentioned in comment
data.competitors_mentioned[].namestringName of the company mentioned

Parent Post Object

FieldTypeDescription
data.parent_post.signal_idstringReference to parent post signal (if exists)
data.parent_post.post_summarystringAI-generated summary of the post
data.parent_post.post_textstringFull text of the parent post
data.parent_post.post_urlstring (URL)Link to the parent post
data.parent_post.posted_datestring (ISO 8601)When the parent post was published
data.parent_post.num_likesintegerLikes on the post
data.parent_post.num_commentsintegerTotal comments on the post
data.parent_post.pain_pointsarray[object]Pain points from parent post
data.parent_post.initiativesarray[object]Initiatives from parent post
data.parent_post.technologies_mentionedarray[object]Technologies from parent post
data.parent_post.competitors_mentionedarray[object]Competitors from parent post (name per entry)
data.parent_post.poster_namestringName of post author
data.parent_post.poster_job_titlestringJob title of post author
data.parent_post.poster_company_namestringCompany of post author
data.parent_post.poster_company_descriptionstringDescription of poster's company

Example Output

"Your recent comment on the Adobe–Marketo acquisition post really stood out. Calling out the challenge of scaling enablement after a merger shows deep awareness of the real pain points sales teams face..."

Identity Resolution

Every LinkedIn comment signal is pre-resolved to a business contact record with a work email. Here's how:

  1. LinkedIn profile URL captured from the comment activity — this is a deterministic, unique identifier
  2. Profile URL matched against our contact database (250M+ contacts, 75M+ companies), ingested monthly with continuous updates from our pipeline exhaust
  3. Business email, job title, and company resolved from the matched record
  4. Company firmographic data (domain, LinkedIn URL, industries, headcount) attached

Key Points

  • Business emails only. The contact.email field is a professional/work email. Our domain validation excludes generic providers (gmail.com, yahoo.com, etc.). We do not deliver personal emails.
  • Match accuracy: 99.8%. We prefer no match over a false match — if we can't resolve a profile to a business contact with high confidence, the signal is not delivered.
  • Coverage: contact.linkedin_url is populated on 100% of signals.
  • No false positives from common names. LinkedIn profile URLs are deterministic unique identifiers — there is no ambiguity in the match.
  • Quality scoring included. Every comment carries a signal_quality score (0.0-1.0) and a comment_intent classification, so you can prioritize the most actionable engagement.

Full matching guide with SQL examples: Resolution


Did this page help you?