Lead Management

6 AI Lead Scoring Examples for Real Estate Agents

RW
Ryan Wanner

AI Systems Instructor • Real Estate Technologist

Quick Answer: AI lead scoring ranks your leads by conversion probability so you focus time where it matters most. These 6 examples show spreadsheet-ready models, CRM workflows, buyer intent scoring, seller motivation scoring, cold lead revival, and marketing budget optimization — all using AI prompts.

Every agent has too many leads and not enough time. AI lead scoring solves this by ranking your leads based on how likely they are to convert — so you focus your energy where it matters. 68% of agents use AI, but only 17% see real results. These 6 examples show you how to build lead scoring models using AI prompts, from simple spreadsheet systems to CRM-integrated workflows using the HOME Framework.

Generic AI vs. Context-Powered AI

The difference between a prompt with no context and a prompt built with a Context Card.

Before Generic AI Output

I follow up with all my leads the same way — first come, first served.

After Context Card Output

I score every lead 0-100 based on behavior, timeline, and engagement. The top 20% get personal calls. The middle 40% get targeted drip sequences. The bottom 40% get automated nurture until they score higher.

Same number of leads, dramatically different conversion rate. Lead scoring doesn't generate more leads — it helps you convert the ones you already have by focusing effort where the probability is highest.

Full Examples with Prompts

1

Basic Lead Scoring Model: Spreadsheet-Ready

An agent with 200 leads in a spreadsheet wants a simple scoring model they can apply manually. No CRM integration needed — just a weighted scoring system that helps them prioritize follow-up calls each morning.

H - Hero

You are a real estate lead management specialist who builds simple, spreadsheet-ready lead scoring models.

O - Outcome

Build a lead scoring model I can apply in a spreadsheet to prioritize my 200 leads.

M - Materials

Data available for each lead: source (Zillow, open house, referral, website, social media), first contact date, last response date, stated timeline, pre-approval status, property searches (if any), email engagement.

E - Execute

Create a 0-100 weighted model with clear scoring criteria. Include spreadsheet column formulas if possible. Show how to categorize leads into action tiers.

Prompt
You are a real estate lead management specialist who builds spreadsheet-ready lead scoring models.

Build a 0-100 scoring model I can apply manually to my 200 leads.

Data I track for each lead:
- Lead source (Zillow, open house, referral, website, social)
- First contact date
- Last response date
- Stated timeline (now, 1-3 months, 3-6 months, 6-12 months, just looking)
- Pre-approval status (yes, no, in process)
- Number of property links clicked in my emails
- Number of replies to my messages

Create weighted scoring criteria and action tier categories.
AI Output
LEAD SCORING MODEL — Spreadsheet Implementation FACTOR 1: LEAD SOURCE (Max 20 points) Referral: 20 pts (highest conversion rate in real estate, ~30%) Open house visitor: 15 pts (showed up in person = real interest) Website form: 12 pts (proactive inquiry) Social media DM: 10 pts (engaged but often earlier stage) Zillow/portal lead: 8 pts (high volume, lower intent) Cold/purchased list: 3 pts FACTOR 2: STATED TIMELINE (Max 25 points) Ready now / within 30 days: 25 pts 1-3 months: 20 pts 3-6 months: 12 pts 6-12 months: 6 pts Just looking / no timeline: 3 pts FACTOR 3: PRE-APPROVAL STATUS (Max 20 points) Pre-approved (letter in hand): 20 pts In process with lender: 12 pts No pre-approval, says they'll get one: 5 pts No pre-approval, hasn't mentioned it: 0 pts FACTOR 4: ENGAGEMENT RECENCY (Max 20 points) Last responded within 48 hours: 20 pts Last responded within 1 week: 15 pts Last responded within 2 weeks: 10 pts Last responded within 30 days: 5 pts No response in 30+ days: 0 pts FACTOR 5: BEHAVIORAL SIGNALS (Max 15 points) Clicked 5+ property links: 10 pts Clicked 2-4 property links: 6 pts Clicked 1 property link: 3 pts Replied to 3+ messages: 5 pts (add to click score, cap at 15) Replied to 1-2 messages: 3 pts SPREADSHEET SETUP: Column A: Lead name Column B-F: Raw scores for each factor Column G: =SUM(B:F) [Total score] Column H: =IF(G>=70,"HOT",IF(G>=40,"WARM","COLD")) [Tier] ACTION TIERS: HOT (70-100): Call today. These leads have high intent, recent engagement, and financial readiness. Personal phone call within 4 hours. If no answer, text + voicemail + email within 24 hours. WARM (40-69): Structured follow-up. These leads are interested but not urgent. Weekly personalized email or text. Monthly phone call attempt. Targeted property alerts. COLD (0-39): Automated nurture. Monthly market update email. Quarterly check-in text. Re-score monthly — cold leads warm up when timelines shorten or engagement increases. RE-SCORING CADENCE: Update scores every Monday morning. Change engagement recency based on last week's activity. Adjust timeline scores as stated timelines progress. A lead who said "6 months" three months ago should now score as "3 months."

Why It Works

The model uses five factors weighted by their actual predictive power in real estate lead conversion. Referrals score highest because they convert at 30%. Pre-approval scores high because it's a commitment signal. The spreadsheet implementation means any agent can use this without buying software.

1

CRM-Integrated Scoring: Follow Up Boss Workflow

An agent uses Follow Up Boss and wants to implement automated lead scoring based on CRM activity. AI helps design the scoring rules that the CRM can track and apply automatically.

H - Hero

You are a real estate CRM specialist who designs lead scoring automations for Follow Up Boss.

O - Outcome

Design a lead scoring automation workflow for Follow Up Boss.

M - Materials

Follow Up Boss capabilities: tracks call/text/email activity, website visits, property views, stage pipeline, tags, smart lists. Agent has ~400 active leads.

E - Execute

Design the scoring triggers, point values, and smart list criteria. Include the tag system and pipeline stages that correspond to score ranges.

Prompt
You are a real estate CRM specialist who designs Follow Up Boss lead scoring workflows.

Design a lead scoring system for an agent with 400 active leads in Follow Up Boss.

Available tracking in FUB:
- Calls (made/received), texts, emails sent/opened/replied
- Website visits and property page views
- Pipeline stages (New, Contacted, Engaged, Showing, Under Contract, Closed)
- Tags (custom)
- Smart lists (filter by any combination of fields)
- Last activity date

Design scoring triggers, point values, and smart list criteria. Show how to set up the daily workflow.
AI Output
FOLLOW UP BOSS LEAD SCORING WORKFLOW SCORING TAG SYSTEM: Create these tags to track scores (update weekly): - Score: Hot (70+) - Score: Warm (40-69) - Score: Cold (0-39) - Trigger: Engaged This Week - Trigger: Gone Silent (14+ days) SCORING CRITERIA: ACTIVITY-BASED SCORING (Track weekly): - Replied to email or text: +10 per reply (cap 30) - Answered phone call: +15 - Initiated contact (inbound call/text): +20 - Clicked property link in email: +5 per click (cap 20) - Visited website: +8 per visit (cap 24) - Viewed 3+ property pages in one session: +15 - Requested showing: +25 - Provided pre-approval letter: +25 DECAY SCORING (Deduct weekly for inactivity): - No activity in 7 days: -5 - No activity in 14 days: -10 (add "Gone Silent" tag) - No activity in 30 days: -15 - No activity in 60 days: downgrade to Cold automatically PIPELINE ALIGNMENT: - New (no score yet) → Initial outreach sequence - Contacted (score 10-39) → Cold tag, automated nurture - Engaged (score 40-69) → Warm tag, weekly personal touchpoint - Showing (score 70+) → Hot tag, daily attention - Under Contract → Remove from scoring, move to transaction management SMART LISTS TO CREATE: 1. "Priority Calls Today" — Filter: Score Hot tag + last activity within 7 days + NOT in pipeline stage Under Contract. This is your daily call list. 2. "Warming Up" — Filter: Score Cold tag + Trigger Engaged This Week tag. These leads just did something — reach out before they cool off. 3. "Going Cold" — Filter: Score Warm tag + Trigger Gone Silent tag. These were engaged and stopped. Time for a pattern-interrupt outreach. 4. "Pre-Approved & Active" — Filter: any score + pre-approved tag + activity in last 14 days. Highest conversion probability. Always in your top call priority. DAILY WORKFLOW (15 minutes each morning): 8:00 AM — Open "Priority Calls Today" smart list. Call every lead on this list before doing anything else. 8:30 AM — Open "Warming Up" list. Send personalized text to each lead that re-engaged. Reference their specific activity: "I noticed you were looking at homes in Brentwood — want me to set up a search?" 8:45 AM — Open "Going Cold" list. Send a pattern-interrupt message: a new listing, a market stat, or a simple "Hey, are you still thinking about buying? No pressure either way." WEEKLY MAINTENANCE (30 minutes, Friday): - Review all leads with Score Warm tag. Update scores based on this week's activity. - Move improved leads to Hot. Downgrade inactive leads to Cold. - Review pipeline stages for accuracy. Move leads forward or backward based on reality. EXPECTED IMPACT: Most agents with 400 leads spend equal time on all of them — or worse, focus on whoever emailed most recently. This system identifies the 40-60 leads (10-15%) that represent 80% of your near-term closings. You'll make fewer calls but close more deals.

Why It Works

The AI designed a scoring system using data Follow Up Boss already tracks — no additional software needed. The smart lists automate the daily prioritization that most agents do manually (or skip entirely). The decay scoring is critical: it automatically demotes leads who stop engaging, keeping your Hot list focused on truly active prospects.

1

Buyer Intent Scoring: From Inquiry to Ready-to-Buy

An agent receives 30-40 buyer inquiries per month from various sources. They need a model specifically designed to measure buyer intent — not just general engagement, but readiness to purchase.

H - Hero

You are a buyer intent analyst who measures purchase readiness signals in real estate leads.

O - Outcome

Build a buyer-specific intent scoring model that measures readiness to purchase.

M - Materials

Buyer signals: property search specificity, budget clarity, timeline statements, pre-approval status, showing requests, mortgage conversations, neighborhood narrowing, emotional language in communications.

E - Execute

Score buyer intent on a 0-100 scale. Weight signals by their correlation with actual purchase timing. Include the 5 strongest purchase-readiness indicators.

Prompt
You are a buyer intent analyst who measures purchase readiness in real estate leads.

Build a buyer intent scoring model focused specifically on readiness to purchase.

Buyer signals I can observe:
- Search specificity (broad area vs. specific neighborhood)
- Budget communication (vague vs. specific price range)
- Timeline language ("someday" vs. "by March")
- Pre-approval status
- Showing requests
- Questions about process (closing costs, inspections, etc.)
- Neighborhood comparison questions
- Emotional language ("love," "perfect," "this is the one")
- Partner/spouse involvement in communications
- School district questions

Score on 0-100. Identify the 5 strongest indicators. Include real examples of language that signals high intent.
AI Output
BUYER INTENT SCORING MODEL THE 5 STRONGEST PURCHASE-READINESS INDICATORS (Research-backed, ranked by correlation with purchase within 60 days) 1. SEARCH SPECIFICITY (Max 25 points) Broad: "We're looking at Nashville" = 5 pts Narrowing: "We like Brentwood and Franklin" = 12 pts Specific: "We want to be in the Brentwood High district, south of Old Hickory" = 20 pts Hyper-specific: "We want Owl Creek or Taramore subdivision" = 25 pts Why it matters: Buyers who've narrowed to specific subdivisions are 6x more likely to purchase within 60 days than those still exploring cities. 2. TIMELINE LANGUAGE (Max 25 points) "Someday" / "when the market changes": 3 pts "Within the next year": 8 pts "In the next 3-6 months": 15 pts "By [specific month]" or "Before school starts": 22 pts "We need to be moved by [date]" or "Our lease ends [date]": 25 pts High-intent language examples: "We need to close before June 1." "The job starts March 15, so we need to be there by March 1." 3. FINANCIAL READINESS (Max 20 points) No mention of budget or financing: 0 pts "We can probably afford around $X": 5 pts "We're pre-qualified": 10 pts "We're pre-approved up to $X": 15 pts "We're pre-approved, cash for the gap, and our current home is under contract": 20 pts High-intent language: "We have 20% down and our lender said we can close in 21 days." 4. PROCESS QUESTIONS (Max 15 points) No process questions: 0 pts General: "What's the buying process like?": 5 pts Specific: "What are typical closing costs in Nashville?": 10 pts Advanced: "Can we do a pre-inspection before writing the offer?" or "How does an escalation clause work?": 15 pts Why it matters: Buyers asking about escalation clauses and closing costs are mentally rehearsing the purchase. They've moved from researching to planning. 5. PARTNER/DECISION-MAKER INVOLVEMENT (Max 15 points) Single decision-maker, engaged: 10 pts Partner mentioned but not involved: 5 pts Partner actively participating in communications: 12 pts Both partners on showing or phone call: 15 pts Both partners + asked about "our" needs: 15 pts Why it matters: When both decision-makers are involved, the timeline compresses. A solo researcher is gathering information. A couple on a showing is making a decision. INTENT TIERS: READY NOW (80-100): Narrowed to 1-2 neighborhoods, specific timeline with a deadline, pre-approved, asking process questions, both decision-makers engaged. Action: Daily contact. Show properties this week. Be ready to write an offer. SERIOUSLY LOOKING (55-79): Narrowed to a few areas, 3-6 month timeline, working on pre-approval, asking general process questions. Action: Weekly showing schedule. Send targeted listings. Educate on process. RESEARCHING (30-54): Broad search area, vague timeline, no financial preparation, general questions. Action: Bi-weekly check-in. Send neighborhood guides. Recommend lender introduction. BROWSING (0-29): No specifics on anything. "Just seeing what's out there." Action: Monthly email drip. Don't invest personal time until signals change. LANGUAGE RED FLAGS (Subtract 10 points each): - "We'll buy when the market crashes": -10 (unrealistic expectations) - "We're in no rush at all": -10 (no timeline urgency) - "We want to see 50 homes before deciding": -10 (analysis paralysis) - No response to direct timeline questions: -10 (avoiding commitment)

Why It Works

This model scores INTENT, not just activity. A lead who's clicked 50 property links but says 'someday' scores lower than a lead who's clicked 5 links but says 'we need to close by March.' The language examples give the agent concrete signals to listen for in every conversation.

1

Seller Lead Scoring: Motivation and Readiness

An agent receives seller leads from home valuation tools, farming, and referrals. They need a scoring model that differentiates between curious homeowners and motivated sellers.

H - Hero

You are a real estate seller lead analyst who measures seller motivation and listing readiness.

O - Outcome

Build a seller-specific lead scoring model that separates curious homeowners from motivated sellers.

M - Materials

Seller signals: home valuation request, equity position, stated reason for selling, timeline, home condition, agent interview willingness, market awareness, already talked to other agents.

E - Execute

Score on 0-100. Weight motivation signals higher than information-seeking signals. Include language that distinguishes curiosity from intent.

Prompt
You are a real estate seller lead analyst.

Build a scoring model for seller leads that separates curious homeowners from motivated sellers.

Seller signals I can observe:
- Source of lead (home valuation tool, direct inquiry, referral, farming response)
- Stated reason for selling
- Timeline language
- Home condition (prepared or not)
- Willingness to schedule listing appointment
- Questions about pricing vs. process
- Whether they've talked to other agents
- Equity awareness
- Life event indicators

Score 0-100. Weight motivation higher than curiosity. Include distinguishing language examples.
AI Output
SELLER LEAD SCORING MODEL FACTOR 1: SOURCE & INITIAL SIGNAL (Max 15 points) Referral from past client: 15 pts (someone recommended you to a seller = high intent) Direct inquiry ("I want to sell my house"): 12 pts Farming response to market update: 10 pts Home valuation tool request: 5 pts (often curiosity, not intent) General website inquiry: 3 pts FACTOR 2: STATED REASON FOR SELLING (Max 25 points) "We're relocating for work" / "Job transfer": 25 pts (external deadline = must sell) "Going through a divorce": 25 pts (legal process requires resolution) "Inherited property": 22 pts (don't want to maintain it) "Downsizing — kids moved out": 18 pts (motivated but flexible timeline) "Want to upgrade to a bigger home": 15 pts (dependent on finding replacement) "Just curious what we could get": 5 pts (exploring, not committed) "Want to know our home value": 3 pts (information gathering) Curiosity language: "What do you think it's worth?" "We've been thinking about maybe selling." Motivation language: "We need to sell by June." "How fast can we get this on the market?" FACTOR 3: TIMELINE (Max 25 points) "How quickly can we list?" / "This week": 25 pts "Within the next month": 20 pts "This spring" / "By summer": 15 pts "Sometime this year": 8 pts "No rush" / "Just exploring": 3 pts FACTOR 4: PREPARATION & ACTION SIGNALS (Max 20 points) Already started decluttering/packing: 20 pts Asking about staging or repairs: 15 pts Asking about net proceeds / closing costs: 15 pts Asking about the listing process: 10 pts Already interviewed another agent: 15 pts (actively shopping = imminent decision) Asking only about home value: 5 pts FACTOR 5: ENGAGEMENT QUALITY (Max 15 points) Scheduled or completed listing appointment: 15 pts Willing to schedule but "need to check with spouse": 10 pts Responded to follow-up within 24 hours: 8 pts "Send me information" (no appointment): 5 pts No response to follow-up: 0 pts SELLER INTENT TIERS: MOTIVATED (75-100): External deadline, specific timeline, asking process/proceeds questions, willing to meet. Action: Listing appointment within 48 hours. Bring CMA and marketing plan. This seller is making a decision, possibly between you and another agent. CONSIDERING (45-74): Valid reason to sell, general timeline, asking pricing questions, some engagement. Action: Schedule a no-pressure consultation. Provide a detailed CMA. Follow up weekly. These sellers often need 4-8 weeks to commit. CURIOUS (20-44): Requested home value, vague reason, no timeline, asking "what's it worth" not "how do we sell." Action: Provide the valuation. Add to quarterly touchpoint list. Send market updates showing neighborhood sales. Don't invest heavy time — yet. COLD (0-19): Home value request with no follow-up. No reason, no timeline, no engagement. Action: Automated email drip with market updates. Re-score if they re-engage. RE-SCORING TRIGGERS: Any of these should trigger an immediate re-score and personal outreach: - Second home valuation request (they're checking again = timeline moved up) - Responded to a market update email with a question - Life event detected (divorce filing, probate, job listing on LinkedIn) - Listing appeared from same address on another platform (they called a competitor)

Why It Works

The model separates motivation (external pressure to sell) from curiosity (wondering about value). The language examples help agents distinguish between 'what's my home worth?' (curiosity) and 'how fast can we list?' (motivation). Most agents treat all seller leads equally — this model concentrates effort on the sellers who are actually going to sell.

1

Re-Engagement Scoring: Revive Dead Leads

An agent has 500 leads that went cold over the past year. Instead of mass-blasting them all, they want AI to help identify which dead leads are most likely to re-engage based on their original behavior patterns.

H - Hero

You are a lead revival specialist who identifies which cold leads have the highest re-engagement potential.

O - Outcome

Build a re-engagement scoring model for my database of 500 cold leads.

M - Materials

Historical data: original lead source, initial engagement level (# of interactions before going cold), stated timeline at initial contact, time since last interaction, seasonal patterns, original property criteria.

E - Execute

Score cold leads by re-engagement potential. Identify the top 50 worth personal outreach. Include the outreach strategy for each tier.

Prompt
You are a lead revival specialist who prioritizes cold lead re-engagement.

I have 500 leads that went cold over the past 6-18 months. Score them for re-engagement potential.

Historical data available:
- Original lead source
- Number of interactions before going cold (calls, texts, emails)
- Stated timeline at first contact
- Months since last interaction
- Original budget range
- Original preferred neighborhoods
- Whether they were pre-approved at any point
- Season when they first inquired

Score for re-engagement potential. Identify my top 50. Include outreach strategy.
AI Output
COLD LEAD RE-ENGAGEMENT SCORING MODEL FACTOR 1: ORIGINAL ENGAGEMENT DEPTH (Max 30 points) Had a phone conversation + showing: 30 pts Had a phone conversation only: 22 pts Multiple email/text exchanges (5+): 18 pts Few exchanges (2-4): 10 pts Single response then silent: 5 pts Never responded: 0 pts Why this matters most: Leads who invested time in conversations had real intent. Something interrupted them — life change, rate shock, competing agent, personal issue. The relationship existed. It can be rekindled. FACTOR 2: TIMELINE MATH (Max 25 points) Calculate: stated timeline at first contact + months elapsed. If they said "6-12 months" and 8 months have passed: 25 pts (they're IN their window now) If they said "3-6 months" and 12 months have passed: 20 pts (overdue, may have bought elsewhere — or still waiting) If they said "just looking" and 12+ months have passed: 10 pts (could have matured into real intent) If they said "now" and went cold: 15 pts (something blocked them — the block may have cleared) FACTOR 3: FINANCIAL READINESS SIGNAL (Max 20 points) Were pre-approved at some point: 20 pts (financial commitment was made) Discussed specific budget numbers: 12 pts General budget range: 6 pts Never discussed finances: 0 pts FACTOR 4: RECENCY (Max 15 points) Last interaction 3-6 months ago: 15 pts Last interaction 6-9 months ago: 12 pts Last interaction 9-12 months ago: 8 pts Last interaction 12-18 months ago: 4 pts FACTOR 5: SEASONAL ALIGNMENT (Max 10 points) Original inquiry was in same season we're in now: 10 pts (seasonal buyers often circle back) Original inquiry in adjacent season: 5 pts Opposite season: 2 pts TIER STRATEGY: TOP 50 (Score 65+): Personal Outreach Campaign Week 1: Personalized text referencing your last conversation. "Hey [Name] — we talked about homes in [area] back in [month]. A lot has changed in that market since then. Still thinking about making a move?" Week 2: If no response, send a specific property link. "This one just hit the market in [their preferred area] — thought of you." Week 3: Phone call. If voicemail, leave a brief, personal message. MIDDLE 150 (Score 35-64): Targeted Email Revival Send a market update specific to their original search area. Include one property match. Subject line: "[Area] market update + a home that matches what you described." Follow up once. If they engage, move to personal outreach. BOTTOM 300 (Score 0-34): Automated Re-Nurture Add to monthly email drip. If they open 3+ emails in a 30-day period, auto-move to Middle tier and trigger personal outreach. RE-ENGAGEMENT SUCCESS METRICS: Expect 15-20% response rate from Top 50 (7-10 responses). Of those, 30-40% will re-activate as warm leads. Expected outcome: 2-4 active clients recovered from a "dead" database. At an average commission of $12,000, that's $24-48K in recovered revenue from leads you were ignoring.

Why It Works

The model's key insight: leads who had real conversations (not just form fills) are the most recoverable. The timeline math factor catches leads who are now inside their originally stated buying window. Most agents treat all cold leads as equally dead — this model identifies the ones worth resurrecting.

1

Multi-Source Lead Comparison: Which Sources Convert Best

An agent generates leads from Zillow, Realtor.com, open houses, social media ads, and referrals. They want AI to help analyze which sources produce the highest-scoring leads so they can allocate marketing budget more effectively.

H - Hero

You are a real estate marketing analyst who evaluates lead source quality using conversion data.

O - Outcome

Analyze my lead sources and recommend budget allocation based on lead quality scores.

M - Materials

12-month data: leads by source, average score by source, cost per lead, conversion rate, average days to close.

E - Execute

Calculate cost per converted lead for each source. Rank by ROI. Recommend budget shifts with projected impact.

Prompt
You are a real estate marketing analyst.

Analyze my lead sources and recommend budget allocation:

12-Month Data:
Zillow Premier Agent: 180 leads, avg score 32, cost $850/mo, 4 closings, avg 142 days to close
Realtor.com: 120 leads, avg score 28, cost $600/mo, 2 closings, avg 156 days to close
Facebook/Instagram ads: 95 leads, avg score 38, cost $500/mo, 3 closings, avg 118 days to close
Open houses: 60 leads, avg score 52, cost $200/mo (materials), 5 closings, avg 78 days to close
Referrals: 25 leads, avg score 74, cost $0 (time only), 8 closings, avg 45 days to close
Google Ads: 40 leads, avg score 45, cost $400/mo, 2 closings, avg 95 days to close

Total annual marketing spend: $30,600
Total closings from marketing: 24

Analyze by cost per closing, lead quality, and time to close. Recommend budget shifts.
AI Output
LEAD SOURCE ANALYSIS — 12-Month Performance | Source | Leads | Closings | Conversion | Cost/Year | Cost/Closing | Avg Score | Days to Close | |--------|-------|----------|-----------|-----------|-------------|-----------|---------------| | Referrals | 25 | 8 | 32.0% | $0 | $0 | 74 | 45 | | Open houses | 60 | 5 | 8.3% | $2,400 | $480 | 52 | 78 | | Google Ads | 40 | 2 | 5.0% | $4,800 | $2,400 | 45 | 95 | | FB/IG Ads | 95 | 3 | 3.2% | $6,000 | $2,000 | 38 | 118 | | Zillow | 180 | 4 | 2.2% | $10,200 | $2,550 | 32 | 142 | | Realtor.com | 120 | 2 | 1.7% | $7,200 | $3,600 | 28 | 156 | KEY FINDINGS: 1. Referrals are your best source by every metric: highest conversion (32%), zero cost, fastest close (45 days), highest lead score (74). Every dollar spent should try to generate more referrals. 2. Open houses are your most efficient paid channel: $480 per closing, 8.3% conversion, and an average score of 52. These leads are in-person connections — they convert faster and score higher because you've already built rapport. 3. Zillow and Realtor.com are your most expensive channels: combined $17,400/year for 6 closings = $2,900 per closing average. The leads are high volume but low quality (scores 28-32) and slow to convert (142-156 days). RECOMMENDED BUDGET REALLOCATION: CUT: - Realtor.com: Eliminate entirely. $7,200/year for 2 closings at the lowest conversion rate. Save: $7,200. - Zillow: Reduce from $850/mo to $500/mo. Keep for volume but don't overspend. Save: $4,200. INVEST: - Open houses: Increase to $500/mo ($300 increase). Host 2x per week instead of 1x. Add professional signage and refreshments. Projected: 8-10 closings instead of 5. - Referral program: Allocate $400/mo to a client appreciation program (closing gifts, annual events, pop-by gifts). Projected: 10-12 referrals instead of 25 leads → potentially 10-12 closings. - Google Ads: Increase to $600/mo ($200 increase). Optimize for neighborhood-specific terms. Projected: 3-4 closings instead of 2. - Facebook/IG: Maintain at $500/mo but shift to retargeting warm leads instead of cold audiences. PROJECTED IMPACT: Current: $30,600/year → 24 closings → $1,275/closing average Proposed: $25,200/year → 28-32 closings → $787-900/closing average You'd spend $5,400 less and close 4-8 more deals. The math isn't complicated — invest in what works, cut what doesn't.

Why It Works

The AI turned raw performance data into a clear budget recommendation with projected impact. The key insight most agents miss: lead volume is not lead quality. Zillow generates the most leads but the lowest conversion rate. Open houses generate fewer leads but convert at 4x the rate. Lead scoring data makes marketing spend decisions obvious.

Pro Tips

1

Score leads weekly, not monthly. Lead intent changes fast — a cold lead who requests a showing this week shouldn't wait until month-end to be re-categorized.

1

Your best leads already have the highest scores. If you find yourself spending equal time on all score ranges, you're not using the model. The whole point is unequal attention allocation.

1

Decay scoring is as important as positive scoring. A lead who was Hot 60 days ago but hasn't responded since is no longer Hot. Automated decay keeps your priority lists honest.

1

Use the OODA Loop with lead scoring: Observe the scores, Orient them against your capacity (how many Hot leads can you actually serve?), Decide your daily priorities, Act on the top-scored leads first.

1

Share lead scoring insights with your team. If open house leads consistently score higher than Zillow leads, the whole team should know — it changes how you allocate floor time.

Frequently Asked Questions

Do I need a CRM to do lead scoring?
No. The spreadsheet model in Example 1 works without any CRM. You need a lead list, the scoring criteria, and 15 minutes each week to update scores. A CRM automates the tracking and scoring, which saves time at scale, but it's not required to start. Start with the spreadsheet model, prove the value, then invest in CRM automation when the ROI is clear.
How often should I re-score my leads?
Weekly is the sweet spot for most agents. Score every Monday morning before you make your first call. If you have a large database (500+), use the CRM to auto-score based on activity and manually review the Hot tier weekly. The key is consistency — irregular scoring creates blind spots where hot leads get missed.
What if my scoring model isn't accurate?
Track your predictions against reality. After 90 days, check: did the leads you scored Hot actually convert? If your Hot leads aren't converting at a higher rate than your Warm leads, your weights need adjusting. Common fix: increase the weight on financial readiness and decrease the weight on engagement volume. Actions speak louder than clicks.
Can AI automatically score my leads?
Some CRMs offer AI-powered auto-scoring (KVCore, Real Geeks, Sierra). These tools analyze behavior patterns and assign scores automatically. If your CRM doesn't offer this, you can build a semi-automated system by exporting weekly data, running it through an AI prompt with your scoring model, and importing the updated scores. The AI Acceleration 5 Essentials framework teaches both approaches.

Related Articles

More AI Examples