02 Understanding Comps Score
The comps score is the AI's assessment of how similar a comparable property is to your subject property. Higher scores indicate more reliable comparisons. Lower scores mean the comp has significant differences and should be used with caution.
How the Score Is Calculated
The AI evaluates each potential comparable across multiple dimensions:
- Location proximity — Properties closer to the subject score higher. A comp on the same block is weighted more heavily than one two miles away.
- Property type match — Same type (single family, condo, multi-family) scores higher. Different types reduce the score.
- Size similarity — Square footage and lot size within 20% of the subject score higher.
- Bed/bath match — Similar bedroom and bathroom counts improve the score.
- Sale recency — More recent sales score higher. Sales within the last 3 months are weighted most heavily.
- Condition and features — Similar age, condition, and notable features (pool, garage, lot features) contribute to the score.
Score Ranges and What They Mean
| Range | Label | What It Means |
|---|---|---|
| 80-100 | Strong comp | Very similar property, recent sale, close location. Use as a primary comparable with high confidence. |
| 60-79 | Good comp | Minor differences in size, age, or distance. Still a reliable comparable for your analysis. |
| 40-59 | Fair comp | Notable differences that may require value adjustments. Use as a secondary reference. |
| Below 40 | Weak comp | Significant differences. Consider excluding unless you have a specific reason to include it. |
Score Distribution
Look at the spread of scores across all comps, not just the highest one.
- Tight cluster (most scores within 10-15 points of each other) — High confidence in the comparable set. The market data is consistent.
- Wide spread (scores from 30 to 90) — Mixed inventory. Some comps are very relevant and others are not. Focus on the high-scoring group.
- All scores low (everything below 50) — The subject property may be unusual, or the area has very few recent sales. Manual analysis may be needed.
Confidence Level
The overall confidence indicator considers:
- Number of comps found — More comps generally mean higher confidence.
- Score distribution — Tight clustering indicates stronger market consensus.
- Data freshness — Comp data is updated regularly. Check the data age timestamp on each result.
A high-confidence result shows 6+ comps, most scoring above 70, with a narrow distribution.
When to Use Low-Scoring Comps
A low-scoring comp can still be useful if:
- Limited inventory in the area — any sale data is better than none.
- The low score is due to one factor (different property type) but location is ideal.
- You need to show a broader market trend rather than direct comparables.
Flag these as secondary comps and adjust their value manually in your analysis.
Next steps: - How to Run Comps on ReiSearch - Adjusting Comps Filters - Saving and Exporting Comps - What Is Underwriting?
Was this helpful?
Your feedback helps us improve the help center.