Last updated:
Quick answer
Define what a good lead looks like, assign points for each trait, and let the AI score every inquiry. Hot leads get your call today. Cold ones wait.
At a glance
All outcome figures are estimates from cited sources, not independent verification by us.
- What it is allowed to do
- Drafts for you
Writes drafts (emails, posts, plans, replies). You review and send. Nothing goes out without you.
- Time saved (estimate)
- About 15 to 30 minutes per day on active inquiry volume
- Difficulty
- Advanced
- Tools needed
- ChatGPT, Google Sheets
- Typical cost
- Free tier of ChatGPT or Claude plus a free spreadsheet. Nothing to buy.
Reviewed Oct 19, 2026
Basis (estimate): Estimate only: triaging and prioritizing a busy inbox of inquiries by hand takes most owners half an hour daily; no timed study found.
What does this setup do?
Build a simple lead scoring rubric in a spreadsheet, then use AI to score each inbound lead and route the hot ones to immediate follow-up.
Who is it for?
Small businesses getting enough inquiries that the good ones sometimes wait behind the bad ones.
How do you set it up?
Tools
- A chat AI (free tier works)
- A spreadsheet with columns for each scoring trait
- Your last 20 to 30 leads, marked with which ones actually bought
Steps
- Look at your last 20 to 30 leads. Mark which ones became paying jobs.
- List the traits the buyers shared: service area, job type, budget signals, how they contacted you, timeline.
- Turn the traits into a rubric: each trait gets points, up to 100 total. Set a hot threshold (for example, 70+).
- Put the rubric in a spreadsheet with one row per new lead.
- For each new inquiry, paste the details plus the prompt below into the AI. It scores the lead against your rubric.
- Hot leads get a call or quote the same day. Warm leads get the standard process. Cold leads get a polite, slow lane.
- Review monthly: if hot leads are not buying, your rubric is scoring the wrong traits. Adjust it.
Example output
Illustrative. Your results will differ.
- 3 new leads logged from this week's inbox
- 1 follow-up draft ready for the lead waiting longest
- You review and send. Nothing goes out on its own.
Where to paste this prompt
ChatGPT: ChatGPT → Project instructions, or ChatGPT Tasks for scheduled reminders (once a day on Free)
Copy-paste prompt
Adapt bracketed fields to your situation.
Score this inbound lead against my rubric. Show your work for each trait, then give the total and the routing decision.
My scoring rubric (total 100 points):
- In my service area: [POINTS, e.g. 20]
- Asking for work I actually do: [POINTS, e.g. 20]
- Budget signal present (mentions budget, asks about pricing tiers, commercial client): [POINTS, e.g. 15]
- Decision-maker contacting me directly: [POINTS, e.g. 15]
- Timeline under [X WEEKS]: [POINTS, e.g. 15]
- Repeat customer or referral: [POINTS, e.g. 15]
Hot threshold: [SCORE, e.g. 70]
New lead:
[PASTE THE INQUIRY: their message, source, anything you know]
Return:
1. Points per trait with one-line justification each. Score 0 for anything unknown: do not guess.
2. Total score and HOT / WARM / COLD.
3. Routing: HOT → call today and quote within 24h. WARM → standard quote process. COLD → polite reply, no rush.
4. If WARM or COLD, draft the reply.Newsletter
Get the free starter pack
Free starter pack with five beginner copy-paste prompts after you confirm, plus three sourced setups in your inbox each week. Unsubscribe anytime.
What usually goes wrong?
Where agents tend to stall on this job. More patterns in Where AI gets stuck.
- Scoring on traits that do not predict buying. Review monthly against actual closes and cut dead traits.
- Letting hot leads sit because the system said hot but you got busy. The score is a promise to act, not a label.
- Overcomplicating it. Six traits and one threshold beats a 20-factor model you abandon in a week.
Which chat tools work with this prompt?
Paste the prompt into any of these assistants (free or paid tiers).
- ChatGPT
- Claude
Common questions
- How many leads do I need before scoring is worth it?
- If you sometimes lose track of who is hot, you have enough. Even ten inquiries a week benefits from a rubric. Below that, your gut is probably fine.
- Should cold leads get any reply at all?
- Yes, a polite one. "Thanks for reaching out: based on what you described, we are not the best fit, but here is who I would call." That email costs nothing and builds goodwill.
- Can this run automatically?
- The scoring can be semi-automated with a form plus AI, but keep a human in the loop on routing. An auto-scored hot lead that nobody calls is worse than no system.
What tips make it smoother?
- Build the rubric from your real buyers, not your imagined ideal customer. Your last 30 leads know more than your gut.
- Score 0 for unknowns. Guessing inflates scores and the whole system stops meaning anything.
- The threshold is a starting guess. If everything scores hot, raise it. If nothing does, lower it.
- Keep the rubric visible. A scoring system nobody looks at is decoration.
Where do the claims come from?
Summaries use public sources linked below, not independent verification. How setups are picked.
- Read MassivePeak’s write-up (www.massivepeak.com)
MassivePeak: Lead scoring setup: define criteria, assign points, set qualification thresholds, refine.
- Read Marion’s write-up (www.marion.com)
Marion: Fit vs interest scoring, thresholds, and reevaluating the strategy regularly.
Related setups
- Capture estimate leads and draft quote follow-ups
Log every estimate or quote request into a tracking sheet, then draft follow-ups when a lead has waited too long for a quote, without auto-sending mail.
- Summarize customer calls and flag local leads
Post per-call summaries and to-dos to Slack or email, and optionally list local forum posts from people asking for your trade.
- Build a lead-gen site with an AI bot
Use an AI coding or copy assistant to draft a simple landing page and contact form for a local service business, then you approve hosting spend, publish, and wire leads to email or a sheet.