Methodology
The whole report turns on one comparison: what your home would sell for versus what the county taxes it as. Here is exactly how we build each side of that, and where it can be wrong.
We geocode the address you enter to a specific property rather than a ZIP or street centroid, then match it to property records so we are working with the right home: its size, bed and bath count, lot, and year built. The report is only as good as the match, so if the address is ambiguous, we say so.
We produce an automated valuation model (AVM) estimate: an algorithmic estimate of what the home would sell for today, based on its characteristics and recent sales of similar homes nearby. This is the same broad technique behind the price estimates you see on major real-estate sites.
We pull the assessed (taxable) value the county has on record for the property, from public assessor and tax data where it is available. This is the figure your bill is actually calculated from. Assessment practices differ by area, some assess at full market value, others at a fixed ratio of it, so we account for the local basis before comparing.
We compare the assessed value against the market estimate. If the assessment is meaningfully higher, that gap is the potential over-assessment. We then apply the local effective tax rate to estimate what that gap is costing you each year. Because it is built on an estimated value and a published rate, the overpayment figure is itself an estimate, and we present it as one.
We gather recent comparable sales, similar homes near yours that sold recently, because comps are the evidence assessment appeal boards actually weigh. We include enough context (size, sale date, distance) for you to judge each comp and drop any that do not fit your home.
When the gap is small, or the comps do not support a challenge, the report says the case is weak. We would rather tell you an appeal is not worth filing than sell you false hope.