How adjustments work (the basic mechanic)
When an appraiser or AVM uses a comparable sale, the comp's sale price is the starting point. Then each line-item difference between the comp and the target gets added or subtracted.
Example. Target is a 4-bed / 3-bath / 2,400 ft² home. Comp sold for $520,000 but is a 3-bed / 2-bath / 2,000 ft² home. Adjustments:
- + Bedroom: target has 1 more bed → +$15,000
- + Bathroom: target has 1 more bath → +$12,000
- + Sqft: target has 400 more ft² → +$60,000 ($150/ft²)
Adjusted comp value: $520,000 + $15k + $12k + $60k = $607,000. That's the comp's contribution to the target's value estimate.
The same logic applies to every comp; the AVM averages the adjusted values across 5–8 comps to produce the central tendency.
The 10 factors, in detail
1. Living area (square footage)
The largest adjustment by dollar magnitude. Typical $/ft² adjustment is 40–60% of price-per-ft² in the local market — not 100%.
Why not 100%? Because adding 400 ft² to a small home doesn't increase the value by exactly 400 × the local $/ft²; there are diminishing returns. A 2,400 ft² home isn't worth 1.5× a 1,600 ft² home; it's worth maybe 1.3×.
Real-world ranges:
| Market | Local $/ft² | Adjustment $/ft² |
|---|---|---|
| Median midwest | $150 | $60–90 |
| Coastal CA | $700 | $280–$420 |
| Tech-hub Austin | $300 | $120–$180 |
2. Bedrooms
Smaller magnitude than sqft because beds and total area are correlated — adjusting for both is double-counting. Modern AVMs typically adjust for bed count only when the comp's bed count differs and the sqft is similar.
Typical: $8,000–$25,000 per bed, depending on home value and local conventions.
A 4-bed home with 1,800 ft² is worth less than a 4-bed home with 2,400 ft² — because the rooms are smaller. The AVM picks one or the other adjustment, not both.
3. Bathrooms
Half-baths and three-quarter baths are weighted lower than full baths. A common convention:
- Full bath: 1.0× value
- Three-quarter bath (no tub): 0.7×
- Half-bath: 0.5×
Typical adjustment: $5,000–$20,000 per full bath.
4. Lot size
Highly market-dependent. In dense urban markets, lot size barely moves the value (the structure dominates). In suburban / rural markets, lot size is a major driver.
- Urban townhomes: $1–$3 per ft² of lot difference
- Suburban single-family: $5–$15 per ft²
- Rural acreage: $5,000–$30,000 per acre
5. Age / year built
Newer homes earn a premium (mechanical systems newer, code compliance, energy efficiency). The adjustment is non-linear — the delta between a 5-year-old and a 15-year-old home is much smaller than between a 50-year-old and a 60-year-old home (because at 50+ years, every system is at end-of-life).
Typical:
| Age delta | Adjustment as % of value |
|---|---|
| 0–10 years | 0.1–0.3% per year |
| 10–30 years | 0.3–0.5% per year |
| 30–50 years | 0.5–0.8% per year |
| 50+ years | 0.8–1.5% per year (steeply discounted) |
6. Condition
The biggest adjustment in markets with mixed-stock comps. A recently renovated kitchen vs an original 1980s kitchen can be a 15–20% value swing. Typical condition adjustments:
- Excellent vs Good: +5%
- Good vs Fair: -5%
- Fair vs Poor: -10–15%
- Each major system at end-of-life: -2–5%
This is the adjustment most free AVMs miss because they don't have current photos. Modern AI-vision AVMs (Twellie, HouseCanary) infer condition from listing photos.
7. Garage / parking
- No garage: -$5–15k
- 1-car attached: baseline
- 2-car attached: +$5–15k
- 3-car attached: +$10–20k
- Detached vs attached: -$3–8k
8. View / location quality
The hardest factor to quantify. Two homes on the same street can differ 20% in value because one backs to a busy road and the other backs to a park.
Modern AVMs use:
- Walk Score / bike score / transit score
- GreatSchools rating
- Distance to amenities
- Crime data (FBI UCR)
- Noise (proximity to highways, airports)
- View (water, mountain, park) — manual flag
Typical: ±5–15% of value for location-quality differences within the same comparable set.
9. Time of sale (date adjustment)
Every comp must be adjusted for market movement between the comp's sale date and the target's valuation date. In a normal market:
- Stable market: 0.0–0.2% / month
- Appreciating market: 0.3–0.6% / month
- Hot market: 0.6–1.2% / month
- Declining market: -0.2–-0.8% / month
A 6-month-old comp in a 0.5%-per-month market gets a +3% upward adjustment. Skipping this is the most common AVM mistake — using stale comps in a moving market understates value.
10. Special features
Catch-all for non-standard amenities:
- In-ground pool: +$15–35k (regional)
- Fireplace: +$2–8k (counts more in cold markets)
- Finished basement: +$25–60/ft² of finished area
- Workshop / detached studio: +$10–40k
- Solar (owned): +$10–25k; solar (leased): -$5k typical
- Smart home / EV charger: +$2–8k
How to read total-adjustment magnitude
Every comp has a total adjustment (sum of all 10 factors, absolute values). This is the comparability score.
| Total adjustments | Comp quality |
|---|---|
| < 5% of sale price | Excellent — direct comparable |
| 5–10% | Good — standard comparable |
| 10–15% | Fair — usable but discount weight |
| > 15% | Poor — exclude or treat as last resort |
If the AVM's headline number is built on comps with 20%+ adjustments, the confidence band should be wide and you should trust the number less.
What the AVM hides (and the appraiser shows)
A formal appraisal makes every adjustment explicit and defensible: the appraiser writes "+$15,000 for second bath" and signs their license against it. An AVM does the same math but rolls it into a single output number; you have to dig into the methodology page to see the multipliers.
A good AVM report (Twellie's, HouseCanary's, Quantarium's) shows every adjustment for every comp. A free AVM (Zestimate's free tier) shows the headline only. The transparency difference is one of the main reasons paid AVMs exist.
What to do with this knowledge
When you're reading a valuation report, look for the comp with the smallest total adjustments — that's the strongest single data point in the report. The AVM averages across all comps, but your gut check should be: "what did the most-similar recent sale go for, with small adjustments?"
That number is your floor. The AVM headline is your central estimate. The upper-band is your ceiling. Together you have a narrow defensible range to offer in.
The full Twellie report shows all 10 adjustment factors per comp on the Comps page. Sample at /mockup/report, methodology at /methodology.