Certified appraisers adjust comp prices for condition differences. Here is the heuristic behind those adjustments and why automating it requires more than listing descriptions.
When a certified residential appraiser selects a comparable sale and adjusts its price for the subject property, the condition adjustment is often the largest single line item on the grid. A comp that closed at $420,000 might carry a -$18,000 condition adjustment because the subject is in superior condition, or a +$22,000 adjustment because the subject needs work the comp did not. These are not arbitrary numbers; they reflect a trained appraiser's judgment about how buyers in that market price condition differences.
How Manual Appraisers Rate Condition
The Fannie Mae condition rating system runs from C1 (newly constructed, never occupied) to C6 (in need of substantial repairs). Most residential properties cluster in C3 (well-maintained, minor updating needed) and C4 (adequate condition, some deferred maintenance). The transition from C3 to C4 -- or from C4 to C5 -- triggers the largest per-dollar adjustments in a typical residential appraisal.
A certified appraiser walks through the property and assigns a condition rating based on observed quality. They look at the roof condition, kitchen and bath finishes, flooring, mechanical systems, and the overall maintenance state of the structure. The rating is an informed judgment, not a formula. Two qualified appraisers sometimes disagree on the boundary between C3 and C4 for the same property, which is why appraiser calibration training exists.
The adjustment dollar amount the appraiser assigns to move from one condition tier to another is derived from paired sales analysis in the local market. A paired sales analysis finds two similar properties that sold within the same time window, where one difference -- in this case condition -- is isolated and the price differential attributed to that difference. In a market with enough transactions, the analysis is reasonably robust. In thin markets, it relies on appraiser experience and regional data from peer appraisers.
Why Listing Descriptions Are Insufficient
The first instinct in automating condition adjustment is to parse listing description text for condition signals. Listings that say "recently renovated," "updated kitchen," or "move-in ready" should score higher; listings that say "as-is," "needs TLC," or "investor special" should score lower. This approach is better than ignoring condition entirely, but it has several failure modes.
Listing agents have an incentive to describe condition favorably. An agent representing a C4 property will not write "deferred maintenance throughout" in the description; they will write "original character" and "priced to sell." Sentiment analysis on listing text picks up the optimistic framing, not the underlying condition. Properties sold off-market or via foreclosure sale often have no listing at all, leaving the algorithm with no text signal to parse.
MLS listing history also has a temporal problem. A property that sold in 2019 as a fixer-upper may have been fully renovated before a 2024 resale, but the AVM pulling the 2019 comp will see the old listing description unless it also ingests permit history and subsequent listing updates. Stale condition data produces systematic bias in markets with high renovation activity.
Permit History as a Condition Signal
Building permit data is a more reliable condition signal than listing descriptions because it represents documented improvement activity. A property with kitchen and bath permits pulled within the last five years has likely transacted from a C4 condition toward C3, regardless of what any listing said. A property with no permit activity in 15 years and a high age-adjusted depreciation calculation is more likely in C4 or C5 condition than one with recent mechanical permit pulls.
Permit data is available at the county level for most Colorado markets. It is not always clean -- permit records have varying levels of detail, not all work is permitted, and permit status (open versus closed) matters for interpreting the signal -- but it provides a factual anchor that listing descriptions cannot. Combining permit history with DOM patterns and listing-condition flags gets closer to what an appraiser's physical observation produces.
Days on Market as a Proxy
Properties in poor condition typically take longer to sell than comparable properties in good condition in the same market, holding price constant. Conversely, properties that are priced below market to account for condition sell quickly despite condition issues. Days-on-market carries a weak condition signal by itself, but it strengthens when paired with the price-to-list-price ratio and the revision history of the listing.
A property that sat on market for 90 days, had two price reductions, and ultimately sold at 93% of original list price in a market where similar properties closed at 99% of list price is showing signs of a condition-related pricing problem. An AVM that captures that pattern and weights it in the condition score is doing a better job of approximating appraiser judgment than one that treats all closed sales as equivalent inputs.
The Disclosure Requirement
Whatever method an automated valuation system uses to approximate condition adjustment, it needs to be disclosed in the output. A desk reviewer cannot evaluate a condition adjustment they cannot see. The report should show the condition rating assigned to each comp, the adjustment applied, and the basis for the adjustment -- even if the basis is "derived from permit history and listing signals" rather than "physical observation." Transparent approximation is more useful to a lender than opaque accuracy.
This is the standard Plotgleam holds to. The condition score in the output is not a black-box number; it shows the signals that drove it and the adjustment applied to each comp. When a desk appraiser reviews the report, they can agree with the condition assessment or override it -- the same workflow they use with any comp grid. The goal is to make the automated output reviewable, not to replace the reviewer.
Condition Adjustment and Local Market Calibration
The dollar value of a condition adjustment is not transferable across markets. In a high-cost Denver infill market, the premium for a fully renovated property over an unrenovated comparable might be $60,000 on a $650,000 home -- roughly 9%. In a lower-cost suburban market in the same metro, the renovation premium on a $380,000 home might be $28,000 -- roughly 7%. In a mountain resort market where buyers expect properties to reflect the premium of the location, condition adjustments may behave differently still.
An automated system that applies national or regional average condition adjustment factors to properties in specific Colorado markets is likely to be systematically wrong in one direction or the other for any given geography. Calibrating condition adjustment values against local paired sales analysis -- finding transaction pairs where condition is the isolatable variable -- is the methodological requirement for accuracy, and it requires enough local transaction data to support the paired analysis.
This is one of the reasons Plotgleam focuses on Colorado markets rather than claiming national coverage. Building a condition adjustment methodology that is calibrated against Colorado transaction data -- and recalibrated as new transactions close -- produces materially better accuracy on Colorado properties than applying a methodology built on national data. The same argument applies to any geographic market with distinctive housing stock characteristics. Local calibration is not a nice-to-have; it is a prerequisite for the kind of condition-adjusted accuracy that lender use cases require.
Request a demo and run a live report on a property from your own market during the walkthrough.