Lead Scorer (ICP Fit)

What it takes in

The inputs you provide on each run, validated before anything executes.

3 tasks, run in order.

Each task passes its result to the next. Here's what each does and what it produces.

1. 01
   Normalize and match the lead against the ICP. Parse {{leadDetails}} into structured attributes (company name, industry, company size, geography, contact title/role, and any other stated facts). Parse {{icpDefinition}} into a list of discrete criteria including must-haves and disqualifiers. Map each ICP criterion against the lead. For any criterion that cannot be evaluated from the inputs, mark it Unknown. Do not infer facts that are not stated. Present the result in clear, readable formatting, never JSON.

   Expected output

   A clear, readable breakdown using short headings and bullet points, with no JSON and no code blocks, containing three sections: Lead Summary, a one-paragraph plain-language summary of the lead; ICP Criteria Match, a bulleted list where each line names the criterion, the lead's actual value, and the match level (Full, Partial, None, or Unknown); and Disqualifiers Triggered, a short list of any disqualifiers the lead hits, or the word None.
2. 02
   Score the lead against a fixed weighted rubric using the Task 1 match results, {{engagementSignals}}, and {{offerContext}}. Dimensions and weights: firmographic fit 35 percent (industry, company size, geography), persona fit 25 percent (decision-making authority and relevance of the contact's title), engagement and intent 25 percent (strength and recency of signals in {{engagementSignals}}; if none provided, score this dimension conservatively and note low confidence), offer-need fit 15 percent (how well {{offerContext}} maps to the lead's likely need; if no offer context provided, score neutral and note it). Each dimension is scored 0 to 100. Then apply a disqualifier penalty (0 to 100) based on any disqualifiers triggered in Task 1. Compute the composite score as the weighted average of the four dimensions minus the disqualifier penalty, clamped to the range 1 to 100. Apply the rubric identically every run. Present the result in clear, readable formatting, never JSON.

   Expected output

   A clear, readable scoring breakdown using short headings and bullet points, with no JSON and no code blocks, containing: Dimension Scores, one line per dimension (Firmographic Fit, Persona Fit, Engagement and Intent, Offer-Need Fit) showing the score out of 100 followed by a one-line rationale; Disqualifier Penalty, the penalty applied out of 100; and Composite Score, the final weighted score out of 100 with a one-line explanation of how it was reached.
3. 03
   Convert the Task 2 composite score into a tier and a concrete recommendation. Tier bands: Hot 80 to 100, Warm 60 to 79, Cool 40 to 59, Low below 40. If any hard disqualifier was triggered in Task 1, set the tier to Disqualified regardless of score. Provide 3 to 5 key reasons referencing specific attributes, list any risks or unknowns (especially criteria marked Unknown), and recommend a specific next action, the best channel to use, and a priority level. Present the result as a polished, readable summary, never JSON.

   Expected output

   A polished, human-readable qualification summary in plain language using clear headings and bullet points, with no JSON and no code blocks, containing: a headline line stating the Score out of 100 and the Tier (Hot, Warm, Cool, Low, or Disqualified); a Key Reasons section with 3 to 5 bullets referencing specific lead attributes; a Risks and Unknowns section with bullets covering any criteria marked Unknown or other concerns; and a Recommended Next Step section stating the specific next action, the best channel to use, and a priority level of High, Medium, or Low.

What you can do with it