CRITICAL: Pricing drift detected in all 4 markets — LSR at 0.95 across portfolio, conversion deteriorating. Agent auto-generated Decision Packets.
March 30, 2026 · Auto-scan · No prompt required
Pricing & Conversion Decision Engine · Opendoor Analytics

From signal detection
to action in seconds.

The agent monitors pricing accuracy and funnel health continuously. When it finds a problem, it doesn't surface a chart — it generates a Decision Packet: what it costs, what to do, who owns it, and what outcome to expect.

$14.7M
Annualized Margin at Risk
60
Homes at Risk / Month
0.95
Portfolio-wide LSR
4 sec
vs 4 Hours Manual
Severity Framework
Every issue has a clear operational tier.
No ambiguity about who needs to act and how fast.
Critical
Act within 24 hours
Escalate to VP Acquisitions + Pricing Lead. Block new offers until resolved.
High
Act within 24–72 hours
Owner produces action plan by next business day.
Medium
Adjust within one week
Flag for weekly review. Adjust if trend continues.
Low
Informational
Log and monitor. Include in next weekly report.
Healthy
No action required
Continue monitoring. Use as comparison baseline.
Decision Packets · Auto-generated
Not insights. Actions.
Every packet contains evidence, ranked actions, financial impact, expected outcome, and owner. Ready to send to a VP or paste into Slack.
Pricing Drift + Velocity Breakdown — Charlotte
Critical SLA: Immediate — 24 hours Owner: market-ops
Evidence
List-to-Sale Ratio0.95 (warn: 0.98)
Conversion Rate WoW−5.9 pts
Homes Sold MoM−17.1%
DOM vs Avg+16 days above avg
Inventory Build+7.4% MoM
Pricing SignalDrift outside acceptance behavior
Segment Priority
1
mid_tier
Highest volume, most elastic to pricing changes
2
entry_level
Affordability sensitive — pricing deviations hit conversion hard
3
premium
Lower volume, longer DOM tolerance — monitor
Recommended Actions
1
Narrow offer band −1% to −2.5% in Charlotte starting with highest-DOM ZIP clusters.
2
Re-run pricing model inputs for Charlotte — LSR at 0.95 indicates pricing drifting outside observed acceptance behavior.
3
Prioritize disposition review for Charlotte homes >60 days DOM — velocity signal indicates demand softening independent of pricing.
Financial Impact
$2.31M / yr
10 homes/month at risk
Expected Outcome
+29.9% acceptance
6 homes/month recovered
Timeframe
2–3 weeks
Track LSR weekly
Confidence
Medium-High
Weight: 0.82
Conversion Collapse — Dallas (Highest Volume Market)
High SLA: 24–72 hours Owner: pricing
Evidence
List-to-Sale Ratio0.95 (warn: 0.98)
Conversion Rate WoW−6.6 pts (worst in portfolio)
Homes Sold MoM−16.0%
Inventory Build+7.3% MoM
DOM vs Avg+15 days above avg
BottleneckMulti-factor conversion pressure
Recommended Actions
1
Narrow offer band −1% to −2.5% in Dallas. Dallas has highest volume — pricing correction has largest absolute impact on portfolio.
2
Investigate multi-factor conversion pressure — both pricing and demand signals active. Isolate by ZIP to determine if pricing-specific or demand-structural.
3
Pause acquisitions in highest-DOM ZIPs until absorption rate stabilizes above 25%.
Financial Impact
$5.12M / yr
19 homes/month at risk
Expected Outcome
+29.9% acceptance
$426k margin/month recovered
Timeframe
2–3 weeks
Higher volume = faster signal
Confidence
Medium-High
Weight: 0.82
Funnel Diagnostics · analyze_funnel_drop()
Where exactly is conversion breaking?
Stage-by-stage breakdown across all markets. The bottleneck signal tells the pricing team where to focus first.
Charlotte
VELOCITY — DOM expanding. Demand softening or pricing mismatch.
Inventory
2,900 (+7.4%)building
Conversion Rate
20.0% (−5.9pts)deteriorating
Acceptance Proxy (LSR)
0.95below benchmark
Days on Market
49 (+6 days)slowing
Homes Sold MoM
580 (−17.1%)declining
Dallas
CONVERSION — Fewer listings converting to sales. Multi-factor pressure.
Inventory
4,400 (+7.3%)building
Conversion Rate
23.9% (−6.6pts)deteriorating
Acceptance Proxy (LSR)
0.95below benchmark
Days on Market
46 (+5 days)normal
Homes Sold MoM
1,050 (−16.0%)declining
Atlanta
CONVERSION — Fewer listings converting to sales. Multi-factor pressure.
Inventory
3,400 (+9.7%)building
Conversion Rate
22.1% (−6.3pts)deteriorating
Acceptance Proxy (LSR)
0.95below benchmark
Days on Market
48 (+5 days)normal
Homes Sold MoM
750 (−14.8%)declining
Phoenix
CONVERSION — Fewer listings converting to sales. Multi-factor pressure.
Inventory
4,000 (+8.1%)building
Conversion Rate
23.8% (−5.9pts)deteriorating
Acceptance Proxy (LSR)
0.95below benchmark
Days on Market
50 (+4 days)normal
Homes Sold MoM
950 (−13.6%)declining
Financial Impact · estimate_business_impact()
What is this actually costing us?
Every issue translated into homes per month, margin at risk, and holding cost exposure. Executive-ready numbers.
Market Homes at Risk / Mo Margin at Risk / Mo Annualized Margin Risk Hold Cost / Home Total Hold Cost / Mo Priority
Dallas
19
$426,360
$5,116,320
$475
$2,090,000
 #1
Phoenix
17
$362,780
$4,353,360
$380
$1,520,000
 #2
Atlanta
14
$244,860
$2,938,320
$475
$1,615,000
 #3
Charlotte
10
$192,500
$2,310,000
$570
$1,653,000
 #4
PORTFOLIO TOTAL
60
$1,226,500
$14,718,000
—
$6,878,000
 ACT NOW
Feedback Loop · The System Gets Smarter
Predictions tracked. Weights recalibrated.
Every Decision Packet is logged. When outcomes are recorded, confidence weights update automatically. The agent improves with every cycle.
1
Decision Packet Generated
Agent detects issue, generates recommendation with expected outcome. Automatically logged with a unique rec_id.
log_recommendation(rec_id="REC-0001")
2
Operator Records Action
Owner logs whether the recommendation was taken, partial, deferred, or rejected.
record_action_taken("REC-0001", "taken")
3
Outcome Measured
Next week: actual KPI change recorded vs what the agent predicted.
record_outcome("REC-0001", observed=24.1, expected=29.9)
4
Confidence Recalibrated
Confidence weights blended with observed accuracy. More data = more weight to observed vs prior.
recalibrate_confidence() → weights updated
5
Weekly Learning Report
Feedback summary included in next briefing. Shows action rates, accuracy by market, calibration status.
get_feedback_summary() → included in briefing
Confidence Weight Calibration — Live Example
pricing_misalignment
0.819
acc: 0.806
inventory_aging
0.780
prior only
funnel_deterioration
0.710
prior only
margin_compression
0.680
prior only
feedback_log.json — REC-0001
accuracy: 0.806
Recommendation: Charlotte pricing_misalignment
Predicted: +29.9% acceptance rate improvement
Action taken: taken — narrowed offer band −2.5%
Observed: +24.1% acceptance improvement
Accuracy: 0.806
Weight update: 0.820 → 0.819 (stable)
Status: Prediction accuracy above 0.75 — weights stable.