realtor-lead-nurture
Find the leads you're about to lose — feed a leads CSV (name, contact, created/last-contact dates, hot/warm/cold temperature) and get a stale-lead revive list sorted by days silent with hot leads weighted first, each tagged with a day-1, day-7, or day-30 touch. Use when doing a weekly CRM review or re-engaging a quiet pipeline.
$ skillbank install realtor-lead-nurture@1.0.0 # pins exactly 1.0.0 · README + changelog included
Realtor Lead Nurture
Overview
A stale-lead detector for your CRM. You keep one CSV of leads with a temperature (hot/warm/cold) and the last time you contacted them. The script scores each lead as days silent × weight (hot ×3, warm ×2, cold ×1), sorts the revive list by that score, and tags each lead with the right next touch — a day-1 check-in, a day-7 value-add, or a day-30 revive message. All data stays local — nothing is uploaded.
Installation
1. Copy the realtor-lead-nurture/ directory anywhere convenient. 2. Make the script executable (Unix):
chmod +x scripts/stale.py
3. Export or write your leads CSV (see Inputs). No pip install needed — standard library only.
Configuration
No config file. Runtime options:
| Flag | Default | Description | |-----------------|---------|----------------------------------------------------------| | --today | real date | Override "today" (YYYY-MM-DD); useful for rehearsals. | | --stale-after | 7 | Only list leads silent at least this many days. |
The CSV must have exactly these columns (order does not matter):
| Column | Format | |---------------------|---------------------------------| | name | Lead's name. | | contact | Phone or email — wherever you reach them. | | created_date | YYYY-MM-DD. | | last_contact_date | YYYY-MM-DD. | | temp | hot, warm, or cold. |
Message templates for each touch live in references/nurture-sequences.md — personalize the bracketed parts before sending.
Inputs
stale.py <leads.csv> [--today YYYY-MM-DD] [--stale-after DAYS]
Returns, on stdout: each stale lead with temperature, days silent, last contact date, and the suggested next touch, plus a count summary. Missing or invalid input produces a clear Error: message on stderr and exit code 2 — never a raw traceback.
Worked example
Ten sample leads with varying staleness (tested with --today 2026-09-27):
cat > leads.csv <<'EOF'
name,contact,created_date,last_contact_date,temp
Maria Santos,555-0101,2026-08-01,2026-09-20,hot
James Park,555-0102,2026-07-15,2026-08-25,warm
Lena Brooks,555-0103,2026-06-10,2026-07-01,cold
Tom Reyes,555-0104,2026-09-01,2026-09-24,warm
Priya Nair,555-0105,2026-05-20,2026-09-15,hot
Sam Whitfield,555-0106,2026-08-20,2026-09-26,cold
Dana Cole,555-0107,2026-09-05,2026-09-10,hot
Greg Hall,555-0108,2026-04-01,2026-05-01,warm
Nina Petrova,555-0109,2026-09-12,2026-09-22,cold
Owen Blake,555-0110,2026-09-15,2026-09-27,hot
EOF
./scripts/stale.py leads.csv --today 2026-09-27
Output (abridged):
Stale-lead revive list (as of 2026-09-27, silent >= 7 days; hot weighted x3, warm x2, cold x1)
1. Greg Hall (warm, 149 days silent, last contact 2026-05-01)
contact: 555-0108
next touch: Day 30 — revive message (see references/nurture-sequences.md)
2. Lena Brooks (cold, 88 days silent, last contact 2026-07-01)
...
3. Priya Nair (hot, 12 days silent, last contact 2026-09-15)
contact: 555-0105
next touch: Day 7 — value-add touch (see references/nurture-sequences.md)
...
5 of 10 leads need attention.
Note how weighting moves hot leads up without ignoring staleness: Dana (hot, 17 days, score 51) outranks Priya (hot, 12 days, score 36) and both outrank Tom (warm, 3 days, score 6 — below the threshold anyway), while James (warm, 33 days, score 66) still earns his higher slot. Sam (1 day) and Owen (0 days) fall below --stale-after 7 and are not listed.
Version history
Changelog
1.0.0 — 2026-09-27
- Initial public release.
scripts/stale.py: leads CSV → stale-lead revive list sorted by days silent with hot/warm/cold weighting, tagged with a day-1/7/30 next touch.references/nurture-sequences.md: day-1 check-in, day-7 value-add, and day-30 revive message templates plus house rules.- Zero dependencies; tested with 10 sample leads of varying staleness plus bad-input cases (bad temp, bad date, future last contact, missing file).