What goes into the benchmark.

Two lists, two different questions.

The residential comparison helps a homeowner examine customer-rating evidence alongside service fit and credential records. The commercial comparison examines a reproducible slice of historical electrical permit activity. A single blended score would hide those differences.

Inclusion starts with a recognizable business, evidence of relevant work in the Kansas City market and a matching credential record where one is available. The source package also contains pending and excluded candidates. A missing or unresolved record is not a finding that a company is unlicensed, unsafe or unqualified.

Commercial: historical KCMO permit activity.

The current metric counts distinct commercial-class sites with an issued electrical permit whose declared project value is at least $100,000. The application date window runs from October 1, 2021 through May 9, 2025, inclusive. Permits recorded as Cancelled, Voided-City, Revoked or Expired are excluded. A company needs at least three qualifying sites to appear in the ranking.

1. Start with the saved KCMO CPD electrical permit export and the edition’s company-to-filer alias table.
2. Filter to the application date window, commercial classification, general or service electrical permit types, a recorded issue date and declared-value threshold. Fire-alarm-only records and cancelled, voided, revoked or expired permits are excluded.
3. Normalize project addresses and count each qualifying site once per company, even if it appears on several permits.
4. Sort by the site count, highest first. Equal counts share a competition rank: 1, 2, 2, 4. Companies inside a tie are alphabetical.

Credential labels are separate factual fields. They do not add points. Filters narrow the visible list without recalculating ranks.

The geography and the dates matter.

A Kansas-based contractor can appear, but the scored activity comes from Kansas City, Missouri records. Kansas-side permits, permits in other Missouri cities, jobs below the threshold and applications after the window are not represented. The 2026 edition date describes the research package, not a 2026 construction window.

What the metric cannot establish.

The source records issue dates and, for closed permits, completion dates. It does not include inspection results. The data standard’s project-value field does not specify whether it covers the whole project or only the electrical scope. Treat it as an applicant estimate, not a verified electrical contract price. More qualifying permits do not establish better workmanship, safer work, lower prices or present-day staffing capacity.

Does the threshold change the result?

Yes. The source cohort is held constant while the value threshold changes. Individual commercial entries show their alternate position at $50,000 and $250,000, or state that they fall below the three-site floor.

The closed-only check keeps the same company cohort, application window and three-site floor but requires a Closed status and a completion date. It is a narrower view of recorded activity, not an inspection or workmanship score.

Identity corrections in this edition.

KC One Service is counted using its own company-name variants. Electric One LLC records are excluded from that entry and evaluated under Electric One’s separate operating identity. Both firms have three qualifying sites and share T-27 at the $100,000 threshold. Electrical Connections has two qualifying sites after a cancelled permit is excluded, so it falls below the ranking floor. The entire order and affected ties have been recalculated.

Residential: ratings with review-volume context.

Google listings were opened individually during an AI-assisted observation pass, with each direct link, rating, review count, observation date and matching rationale recorded. The method uses no paid API or bulk scraper. The listing must be a high-confidence match to the company and branch, with at least ten reviews, to enter the scored cohort. Multi-trade companies are labelled because their reviews can cover HVAC, plumbing or other services.

Every eligible candidate must have either a valid observation or an explicit, dated reason why a usable listing could not be matched. We do not calculate an edition from whichever companies happened to be checked first. An unobserved company is not given a zero rating.

The weighted value is B = (25 × C + n × r) / (25 + n), where C is the median rating of the complete qualifying cohort, n is the listing’s review count and r is its rating. The prior weight of 25 reduces the influence of very small review samples; it is an editorial choice, not a guarantee of statistical accuracy.

Order and ties use B rounded to three decimal places. Equal rounded values share a competition rank and are alphabetical inside the tie. The displayed 0–100 score is 100 × (B − 4), clipped to that range. Display rounding does not change the underlying tie rule.

Customer ratings are opinions about a listing. They are not inspections or proof of electrical competence. Neither a high rating nor an active credential record replaces a project-specific scope, insurance check or discussion with the contractor.

For this edition, 50 of 60 eligible companies met the rating and matching requirements; the cohort median C is 4.85. Observations were recorded on Oct 4, 2026. Eight listings had fewer than ten reviews, and two could not be confidently located. All ten remain visible without a rank.

Small score gaps should not be read as meaningful differences in contractor quality. The weight and review floor are editorial choices: using a prior weight of 50 instead of 25 preserves the top eight but shifts some smaller-review firms. Ratings may change after the observation date.

Sources and how to read them.

Each entry exposes its source links and observation dates. “Listed address” may be an address from a company website or license record; it is not necessarily a walk-in office. The illustrative category photographs do not depict a listed company or a verified project.

Download the limited public-record comparison data. This export includes the displayed company identifiers, corrected commercial counts and source links. It contains no Google ratings or review counts, no individual licensee details and no private research notes.

Editorial controls and corrections.

The formulas have no paid-placement input. A service filter does not buy or change a position. The publication uses a brand editorial byline; the methods, source dates and corrections supply the accountability for each edition.

Corrections must identify the affected claim and supporting source. A changed input triggers a recalculation of the entire relevant cohort, including ties, counts and sensitivity results. See the current correction process.