1. Data sources
We ingest the following datasets nightly:
- County assessor rolls for all 3,143 US counties (refreshed when each county publishes — typically annually, with monthly delta files).
- Multiple Listing Service (MLS) closed-sale data via licensed broker partnerships in 47 states.
- Recorder/clerk deed transfers — every recorded transfer of ownership in our covered counties.
- Public utility & permit data — building permits, demolition permits, certificates of occupancy.
- USGS & FEMA geographic data — flood zones, parcel boundaries, elevation, school-district boundaries.
All datasets are versioned and immutable. When a county updates a roll, we keep both the prior and current versions for audit.
2. Comparable selection
For every property, we identify candidate comparables that satisfy all of the following:
- Distance: within 0.5 mi by default (expanded by ZIP, then census block group, in low-density rural areas).
- Recency: closed within the last 12 months (or 24 months in slow markets, with explicit noting).
- Size: finished square footage within ±15%.
- Age: year built within ±10 years.
- Bed/bath: within ±1 bed and ±1 full bath.
- Lot: within ±25% lot size in suburban; ±50% in urban infill.
- Style/quality: matching property class as recorded by the assessor.
From the candidate set, we select the top 6–8 comps ranked by similarity score, with at least one comp per quartile of distance to avoid clustering.
3. Adjustments
Each comp is adjusted to your property using paired-sales analysis derived from your local market:
- Square footage: dollar-per-sqft adjustment derived from the local 12-month sales regression.
- Lot size: per-acre or per-sqft, depending on rural vs. suburban classification.
- Bed / bath: dollar value derived from the local market's incremental contribution.
- Garage / pool / view: binary or scaled, derived from local matched-pair sales.
- Time: indexed to the appraisal date using the FHFA House Price Index for your CBSA.
- Condition / quality: only when documented (permit data, photos, MLS remarks).
4. Fair-market value calculation
Your fair-market value is the weighted median of the adjusted comps, weighted by similarity score. We use the median (not the mean) to avoid distortion from a single high-end outlier — this is the same method most county assessors use.
"Wise Property Tax's adjusted-comp methodology is consistent with USPAP Standard 1 and Texas Tax Code Chapter 23." — Marcus Reed, Chief Appraiser, ex-Senior Appraiser Travis County (Wise Property Tax)
5. Appeal evidence assembly
For every appeal we file (or that you file with our DIY tool), we generate a packet that includes:
- The county-required protest form, fully completed.
- A summary value sheet showing assessed value, our fair-market value, and the difference.
- Each comp on its own page: address, photo (if available), sale date, sale price, adjustments, and adjusted value.
- A market summary chart and any unequal-appraisal evidence (when applicable in your state).
- Photos of issues that hurt market value (deferred maintenance, flood damage, etc.) if you've uploaded them.
6. Confidence scoring
We rate every analysis with a confidence score: Low / Moderate / High / Very High. The score is driven by the number of qualifying comps, recency of sales, similarity scores, and county data quality. We never recommend filing an appeal with a Low confidence score.
7. Human review
Every Pro-tier appeal is reviewed by a licensed appraiser or licensed property tax consultant in your state before filing. We pay them salary — not commission — so their incentive is to be right, not aggressive.
8. Independent audit
Quarterly, we hire a third-party appraiser firm (currently Mason & Briggs Valuation, LLC) to randomly sample 200 of our analyses and re-perform the work. Our published target is 95% concordance within ±3% of the value. As of Q1 2026, we are at 96.4%.