Tool · Score Intent

Score any mention for buying intent. Know where to spend your time.

The intent scoring tool takes the text of any social post, forum thread, or online mention and returns a 0-to-100 intent score, a tier label, a suggested outreach angle, and the reasoning behind the score. It is what MentionFox uses internally to prioritize the Engagement HQ queue and filter scan results.

Open Engagement HQ

What the score measures

Intent scoring is post-level, not person-level. It asks: does this specific piece of text signal that the author is currently looking to buy, switch, or evaluate a solution? The scoring model looks for:

Explicit complaint language about a current tool or situation
Evaluation language — asking for recommendations, comparisons, or pricing
Switching language — tired of, looking for an alternative, switched from
Urgency signals — timelines, deadlines, just started a new role
Frustration signals — stuck with, fed up, cannot get this to work

Posts that score high are worth your time to engage with directly. Posts that score low are either low-signal (general chatter) or off-topic (spam, promotional, or unrelated to your category).

Intent tiers

High intent (70+)

Author is actively evaluating or expressing strong pain around a specific problem. Engage now. These posts age quickly.

Medium intent (30–69)

Author shows some signal — a complaint, a question, a comparison ask — but the context is less specific or urgent. Worth engaging with a thoughtful reply.

Low intent (0–29)

General chatter, announcement posts, or off-topic content. Hidden from the default Engagement HQ view but available via "Show hidden items."

The system applies additional rules for specific post types. Product Hunt launch threads — where the comment section is typically "wow, congrats!" rather than genuine evaluation — are capped at a lower score unless the post contains real complaint or switching signal text. This prevents high-activity threads from dominating the queue with posts that have no reachable intent behind them.

Inputs and outputs

Input is the post text and optionally a platform label and post URL. The platform label improves the scoring model's calibration (Twitter average intent scores differently from a Reddit software recommendations thread).

Output:

score — integer from 0 to 100
tier — "high", "medium", or "low"
reasons — list of signals detected in the post
suggested_angle — the recommended outreach frame given this specific post
intent_score_reason — present when a score cap or floor was applied, explaining why

Via API and MCP

Intent scoring is available as a standalone tool via the MentionFox API and MCP. The MCP tool name is score_intent.

POST /score-intent
{
  "post_text": "Anyone else fed up with Clay? Looking for something that doesn't require a PhD to set up.",
  "platform": "twitter",
  "post_context": "B2B SaaS audience, sales enrichment thread"
}

Response:

{
  "score": 76,
  "tier": "high",
  "reasons": ["competitor frustration", "switching signal", "recommendation request"],
  "suggested_angle": "Reach out acknowledging the setup complexity pain. Lead with how quickly users get to first value.",
  "intent_score_reason": null
}

See it in action

Intent scores appear on every scan card and Engagement HQ item inside MentionFox.

Open Engagement HQ