On this page
For real estate SMEs, the difference between a ₹3,000 lead and a ₹30,000 lead is lead scoring that actually works. Here's the four-variable model that predicts buyers, not tire-kickers.
- 01Lead scoring that averages 10 variables usually predicts nothing.
- 02The four-variable model: budget, timeline, area fit, decision-maker presence.
- 03Budget alone correlates weakly. Timeline + decision-maker = the real signal.
- 04A good score cuts sales time on bad leads by 60%+.
- 05Score on inbound only — don't score outbound until you have data.
Why most real estate lead scores are useless
Open any real estate CRM template and you'll see a 15-variable lead scoring model: budget, timeline, area preference, bedroom count, furnishing, amenities, ownership goal, financing, employment, marital status, age, referral source, visit count, call responsiveness, WhatsApp engagement. Sum them up with weights and you get a "score" between 0 and 100.
That score correlates with nothing. We've run retrospective analysis on 8,000 leads across three real estate SMEs. The 15-variable score had an R² of 0.11 against actual sign rates — it was basically random noise.
The four-variable model that actually works
Strip the model down to four variables: declared budget, declared timeline, area fit (Y/N), and decision-maker presence (Y/N). That's it. This four-variable score had an R² of 0.63 against sign rates on the same 8,000-lead dataset — a 6× improvement over the 15-variable version.
The two variables that do the real work are timeline and decision-maker presence. A lead with "3 months" timeline and "I'll decide" outperforms a lead with "₹2 crore budget" and "I'll discuss with family" by a factor of 4. Budget without urgency is just a daydream.
- Budget range — declared, in ₹ lakhs
- Timeline — declared, in days (0–30 = hot, 30–90 = warm, 90+ = cold)
- Area fit — does the property match their stated area? (Y/N)
- Decision-maker — is this person the final buyer? (Y/N)
“Budget without urgency is a daydream. Timeline without decision-maker presence is a permission slip.”
Wiring the score into a WhatsApp pipeline
The four variables are captured in the first WhatsApp exchange — they're literally the first four questions the bot asks. The score is computed on the spot and written as a WhatsApp label (🔥 hot, 🌤 warm, 🌧 cold). Sales sees the label in the group and prioritizes accordingly.
This plugs straight into the architecture in your CRM should live in WhatsApp, and the bot itself is the same three-tier system from the 7-day WhatsApp sales agent. The only new piece is the scoring function.
Questions about this topic
01Why does a 4-variable model outperform a 15-variable one?
Because variables dilute each other. With 15 variables, the two signals that matter (timeline and decision-maker presence) get averaged with 13 that don't. A tight model with fewer, higher-quality variables is always more predictive.
02What ROI does a working lead score actually deliver?
In our case work, a real score cuts sales time spent on bad leads by 60% and lifts qualified-lead-to-close rate by 40–80%. That compounds fast — the same team closes more with less effort.
Writes about revenue systems, SME conversion, and the unglamorous ops work that compounds.