Methodology · v1 · October 2026

How we get the number

Plumbroll is a transparent screen, not an appraisal. Every figure on a parcel page can be recomputed from public county files using the steps below.

1. Data

• Summit County: Fiscal Office CAMA data downloads (parcel, appraised value, sales, commercial building, commercial income-use, owner and mailing tables), tax year 2026, exported Oct 2, 2026.
• Butler County: Auditor public AA407 data export (same table family), latest certified tax year 2025, exported Oct 1, 2026. TY2026 values are not in the public export yet; Butler results will be refreshed.
• Montgomery County: parcel, use code, owner and 2025 value from the county's public ArcGIS parcel layer. Sales and building-size files were not publicly downloadable when we built v1, so Montgomery parcels are listed but not modeled.
• Tax rates: Summit Fiscal Office "25TY-26CY Full Rates" (Class II effective mills by taxing district); Butler: each parcel's TY2025 gross charge less reduction factors, divided by assessed value, from the county charge file.

2. Which parcels

Ohio land-use codes 401–403 (apartments, 5–50 units), 420/425/426/429 (retail and small centers, 1,500–60,000 sq ft) and 447–449 (office, 1,500–60,000 sq ft).

3. Comparable sales

Sales recorded Jan 1, 2023 – Jun 30, 2026 that the county coded valid (Summit code 1, Butler code 0, single parcel), priced at least $50,000, with a county value between 0.4× and 2.5× the price (to drop non-market transfers). Each sale becomes a price per sq ft (or per unit for apartments).

A parcel is compared against the narrowest cohort with at least 5 sales, excluding its own: county × use × size band × age band, then county × use × size band, then county × use. The model value is the cohort median times the parcel's size.

Income check

Where the county's own income model records a net operating income for the parcel, we divide it by a market cap rate derived from the same sales (median NOI ÷ price for the use) and average that with the sales value.

4. Flag and confidence

We flag a parcel when the county value is at least 10% above the model and cohort dispersion (coefficient of dispersion, COD) is 60% or less. Confidence (0–100) rises with the number of comps (up to 20), falls with dispersion, gains when the income check agrees within 25%, and is cut for broad cohorts or extreme gaps that usually mean a data issue (multiple buildings, partial interests).

5. Dollars a year

savings = (county value − model value) × 35% × Class II effective mills ÷ 1000 × 0.85. Ohio taxes 35% of market value. The 0.85 cushion reflects that reduction factors typically lower effective rates after a value update; certified 2026 rates are not out yet.

Limits

• Sales are not time-adjusted, and condition, location within a city, leases and deferred maintenance are not captured.
• Many small commercial sales are not arm's length; validity codes are the county's.
• A Board of Revision decides value on evidence (appraisal, rent roll, sale). Our number is a reason to look, not proof.
• A school board can file a counter-complaint seeking a higher value.

Found an error? Tell us on the contact form and we will correct it.