Lead Scoring & Triage Queue

The problem this kills

A pile of inbound leads lands every day, and the way they get worked is basically vibes. The rep who's fastest grabs the easy ones. The lead who emailed three times gets called back before the quiet VP at a perfect-fit account. Nobody can say why this lead got called and that one sat for a week. And when a great opportunity goes cold, the post-mortem is a shrug.

Most teams "prioritize" leads with a gut feel, a color-coded spreadsheet someone updates by hand, or a black-box score from a CRM that nobody can explain to a rep. So reps don't trust it, managers can't tune it, and the same lead sometimes gets scored and queued twice under two slightly different spellings of the company name.

This tool replaces the guesswork with a scoring rubric you define - the firmographics, sources, behaviors, and keywords that actually predict a good deal for your business - and turns it into a clear, explainable, ranked queue. Every score shows its work. The manager stays in control. And because rubrics are versioned, last month's scores still make sense next month.

What you'll build

A small internal web app, just for your team, that:

• Imports a batch of leads from a CSV (company, contact, source, and whatever attributes you score on).
• Reads your scoring rubric - the fields, weights, and thresholds you set for fit and intent - and computes a score for every lead.
• Shows a plain-language explanation for each score: which rules fired, how many points each added, and why this lead landed where it did.
• Buckets leads into hot / warm / cold using your thresholds.
• Lets the manager review and override any individual lead's tier, with a required reason that's logged.
• Captures disqualify reasons so a bad-fit lead is recorded as out, not just ignored.
• Dedupes on lead email so the same person can't be scored and queued twice.
• Stamps every score with the rubric version that produced it, so scores are reproducible and comparable over time.
• Publishes a ranked queue per rep and notifies each rep via Resend.
• Exports a scored CSV with score and reason columns in the shape your CRM expects.

What's inside the Implementation Plan

The plan is a single markdown file you paste into Claude Code (a free AI coding agent). It walks the agent through building the whole tool, step by step, each step ending with a ready-to-paste prompt.

The most important part: the plan opens by interviewing you about your business. Before it writes a single line, the agent asks how leads reach you today, what your rubric actually rewards, the real field names in your lead data, your typical and peak lead volumes, your exact hot/warm/cold thresholds and routing rules, and your messiest edge cases (no-name companies, free-email signups, duplicate submissions, disqualifiers). It reads a short tailored spec back to you, you confirm it, and only then does it build - so you get a scoring tool shaped to your definition of a good lead, not a generic template you have to bend to fit.

Inside you'll find:

• The discovery interview and how the agent turns your answers into the scoring model.
• The full build: database, login, CSV import with duplicate guards, the rubric engine with versioning, the explainable scoring, the review-and-override screens, the per-rep queue, and the rep notifications.
• The hard human gate: the manager approves a rubric version before it can score anything, and reviews/overrides tiers before the queue goes live.
• Verification steps so you can prove it works, and the CSV-export fallback so it's fully usable even before you connect it to your CRM.

The governance it includes (this is the point)

This isn't a toy. The plan builds in the controls a real sales operation needs:

• Login so only your team can see or touch anything.
• Row-level security so people only see the leads and queues that belong to your organization.
• A complete audit trail - every import, score, rubric change, tier override, disqualification, and queue publish is logged with who and when.
• A hard human-in-the-loop gate - the AI scores and ranks, but a manager must approve the rubric version before it's applied and must review/override tiers before the queue is published to reps. Nothing auto-publishes.
• Versioned rubrics with effective dates so any score can be traced back to the exact rules that produced it - reproducible months later.
• Duplicate guards (on lead email) so the same lead can't be scored and queued twice.

Who it's for

SDR managers, demand-gen and revenue-ops folks, and small sales teams who are tired of "whoever shouts loudest" lead prioritization - and who want a transparent, tunable, auditable triage tool without hiring a developer or buying a heavyweight scoring platform. You don't need to write code. You need your leads CSV, your scoring rubric, and an afternoon-to-a-weekend.

You've got this - paste the first prompt and let the agent interview you.