Create a Lead Scoring System with AI

1. Open https://claude.ai and start a new chat. 2. Paste three bullets: (a) your offer in one line, (b) what your best customers have in common, (c) what bad leads look like. 3. Ask: “Based on this, list (a) disqualifiers—who we should NOT pursue, (b) 3–5 positive signals that historically meant real progress (reply, meeting booked, budget or timeline mentioned), (c) what we should NOT score yet because we don’t have reliable data.” 4. If the list feels generic, add one concrete example of a lead that advanced and one that died—then ask Claude to sharpen signals.

1. In the same Claude chat, paste the column headers from your lead export (or list the fields you usually have: source, title, company size, last activity, etc.). 2. Ask: “Propose 3–6 scoring dimensions. For each: why it matters for likelihood to buy, and a weight percentage. All weights must sum to 100%. Use names that could become spreadsheet column headers.” 3. If you lack engagement data, say so and ask Claude to shift weight toward Fit + Intent. 4. Rename anything jargon-heavy until you could explain each dimension to a teammate in one sentence.

1. Ask Claude: “Using my dimensions and weights, build a rubric where each dimension has a maximum point value equal to its weight. For each dimension, define rules for 0 points, partial points, and full points in plain English (e.g. full Fit points only if title and company size match our ICP).” 2. Ask for the output as a markdown table: Dimension | Max points | 0 pts | Partial | Full. 3. Confirm the sum of all max points is 100 (or your chosen total—100 is easiest for tiers later). 4. Copy the markdown—you’ll paste it into a second tab or notes in the next step.

1. Go to sheets.google.com and create a blank spreadsheet named something like “Lead scores.” 2. On the first tab (rename it “Leads”), row 1: Lead name, Email, then one column per raw field you track, then one column per dimension’s points (use short headers matching Claude’s names), then Total, Tier, Next action. 3. Select row 1 and use View → Freeze → 1 row (or Format → Freeze). 4. Add a second tab named “Rubric” and paste the markdown table from the previous step so you can score without scrolling.

1. Pick the column letters for each dimension’s points (e.g. Fit in D, Intent in E, Engagement in F). 2. In the Total column, row 2, enter =SUM(D2:F2) and adjust the range to cover all your dimension point columns—fill down later. 3. In Tier (assuming Total is column G), use something like =IFS(G2>=70,"Hot",G2>=40,"Warm",G2<40,"Cold")—replace 70/40 with cutoffs you’ll refine next step; adjust column letters to match your sheet. 4. Paste 10 real leads from your export into rows 2–11. 5. For each lead, enter dimension points using the rubric (whole numbers that respect each dimension’s max). 6. Eyeball: known great leads should land high; known time-wasters low. If not, fix points before touching formulas.

1. In Claude, paste your tier cutoffs (e.g. Hot ≥70, Warm 40–69, Cold <40) and describe two example leads per tier from your sheet. 2. Ask: “For each tier, give one recommended next action for a solo seller (e.g. Hot: call within 24h; Warm: add to nurture; Cold: disqualify or long-cycle drip). Keep each to one sentence.” 3. In Google Sheets, add those defaults in the Next action column (or as notes on the Tier header) so empty cells have a clear playbook. 4. In cell A1 or a notes row at the bottom, add: Review scores weekly—if one tier has almost everyone, adjust one weight or one cutoff and re-check 5 leads. 5. Optional: share view-only link with a collaborator and ask them to score two leads using only the Rubric tab—if they disagree wildly, tighten the rubric wording.