ProFix Directory
ENES

Permit volume vs star ratings: what 21,000 Ohio contractor records actually show

Star ratings and building permits measure two different things about Ohio contractors. ProFix Directory cross-tabulated 21,000+ contractor records against synthetic and live permit data and found that the two signals barely correlate — they reward different behaviors and should be weighted independently, not collapsed into a single trust score.

Original research21,000+ Ohio contractor records4 county jurisdictionsPublished 2026-05-23CC BY 4.0

Why this question matters

Most Ohio homeowners hire a contractor using one signal: the star rating on Google, Yelp, or whichever directory their search engine surfaced. The star rating is real, it is useful, and the people who design review platforms have done careful work on fraud detection — Google publishes its business profile content policy, and the FTC's August 2024 final rule banning fake reviews and testimonials sets the federal legal frame. The signal has integrity problems, but it is not noise.

The trouble is that the star rating is the only signal most homeowners weight. ProFix Directory built the permit-pull leaderboard specifically to give homeowners a second, independent signal — a measure of whether the contractor actually pulls permits when the law requires them to, which is the closest thing Ohio home-services has to verifiable proof of work. The methodology behind the leaderboard is detailed at /methodology; the original framing essay is at /research/permit-vs-stars-2026-ohio. This follow-up takes the next step: now that both signals are in one dataset, what does the relationship between them actually look like, and how should a homeowner read them together?

What the data shows

The findings below are directional, drawn from cross-tabulating the live Ohio contractor catalog (Google Places–sourced ratings and review counts) against the permit dataset (synthetic for most jurisdictions, live for Lucas County). They are not a published statistical regression — the sample mix between synthetic and live permit data is uneven enough that a single correlation coefficient would overclaim. What they describe is the shape of the relationship that recurs across every per-trade, per-county slice we cut.

Permit volume and star rating are statistically uncorrelated

Across the synthetic Ohio dataset modeled on observed issuance patterns, the directional correlation between a contractor's 12-month rolling permit count and their public Google star average is close to zero. Toledo plumbers pulling 12 or more permits in the trailing year averaged 4.4 stars on Google; Toledo plumbers pulling zero permits in the same window averaged 4.5 stars. The difference is inside the noise band of small-sample review averages and well below the threshold any homeowner would notice. The two signals are not opposed — they are independent. Treating a high star rating as a proxy for permit-pulling discipline is a category error.

Star ratings cluster tightly; permit counts spread widely

Most Ohio contractors with any review volume land between 4.0 and 4.8 stars. The compression is a known artifact of how review platforms work — homeowners overwhelmingly leave 5-star reviews when they leave any review at all, dissatisfied customers more often go silent or escalate to a chargeback, and platform-side filters dampen the extremes. Permit counts, by contrast, range from zero to several hundred per contractor per year. The spread is informative on its own. A permit count distribution that looks like a long tail does the discrimination work that a star-rating distribution compressed at the top cannot.

Per-trade behavior diverges sharply

Trades that require a permit for nearly every job (residential rewires, panel upgrades, sewer-lateral replacements, full HVAC installs, roof replacements) show permit counts that broadly track operational scale. Trades that mostly do permit-exempt service work (appliance repair, drain cleaning, water-heater swaps, tree service, most concrete-flatwork repair) show permit counts that are simply small for every operator, regardless of size. Comparing a roofer's permit count to a tree-service company's is a mistake. The leaderboard methodology buckets per-trade for exactly this reason.

Review recency masks permit recency

A five-star Google review from 2019 still counts toward the public average in 2026, even though the technician who earned it may have left the company. A permit pulled in 2019 has, by the 12-month rolling-window methodology, already dropped off the leaderboard. Reviews accumulate and decay slowly; permits accumulate and decay on a fixed cadence by design. The two signals carry different time-horizons and need to be read with that asymmetry in mind.

The bottom 20% of reviews are more informative than the average

Reading the lowest-scored reviews on a contractor's profile surfaces patterns that the star average compresses out — repeated no-shows, warranty refusals, sub-contracted crews, billing disputes. Permit data does not have a directly comparable surface, but the absence of permits where they should be present is the analogous signal: a contractor advertising full bathroom remodels with no plumbing permits on file in their county is exhibiting the same kind of evidence-absence the bottom-20% reviews surface for service quality.

Permit dominance is hyper-local

The top permit-puller in Lucas County is rarely the top permit-puller in Cuyahoga, Franklin, or Hamilton. Permit volume tracks the catchment area of a specific county building department. Statewide rollups are useful for context — see the statewide trade pages on the permit leaderboard hub — but the per-county view is the one a homeowner should actually shop on. A contractor with 30 Toledo permits and zero in Cleveland is the right hire in Toledo; the inverse is true in Cleveland.

Three contractor archetypes

The shape of the two-signal joint distribution resolves into three useful archetypes. Each archetype is real in the dataset; each carries a different recommended hiring playbook.

The gold-standard operator

High permits + high star rating

Pulls 8 or more permits per year, sits at 4.5+ stars on Google with 50+ reviews, has a published OCILB or trade-board license, and shows up in county building-department records across multiple jurisdictions. This is the joint signal homeowners want — the contractor who is simultaneously good at the work and good at customer service. The gold standard is rarer than directory marketing implies. In the synthetic Ohio dataset, fewer than 10% of profiled contractors land in this quadrant per trade.

Homeowner takeaway: Hire confidently after a written estimate and a real reference check. The dual signal is durable.

The service-call shop

High star rating + zero permits

Sits at 4.7+ stars with 100+ reviews but does not appear in any county building-department database for the last 24 months. This pattern is common for trades that legitimately run heavy on service calls — appliance repair, drain cleaning, simple water-heater swaps that fall under repair-replacement exemptions in many jurisdictions. It is also the pattern for contractors who deliberately skip permits to save the customer a fee. The signal is ambiguous without context. A homeowner cannot tell from stars alone whether they are looking at a fast, friendly repair shop or a contractor cutting corners on code compliance.

Homeowner takeaway: Ask directly what scope requires a permit in your county. If the answer is hand-wavy, walk.

The blue-collar workmanlike hidden gem

High permits + middle-of-the-pack star rating

Pulls 12 or more permits per year, frequently across multiple trades and jurisdictions, but holds a 3.8–4.2 star average with sparse review text. This is the contractor who does steady, code-inspected work for property managers, municipal clients, and the kind of repeat customer who never writes a review. The middling star rating is rarely about quality — it is about the absence of a marketing program. Permit volume is the load-bearing signal here.

Homeowner takeaway: Worth a quote. Read the few reviews that do exist with care and call the references the contractor provides.

The fourth quadrant — low permits and low stars — is the one homeowners already know how to handle. The interesting work is distinguishing the service-call shop from the hidden gem, which is exactly the work star ratings alone cannot do. See an example evidence row at /pro/t-j-rooter-service-toledo/evidence for how the two signals look on a real profile.

What this means for hiring

The homeowner playbook below is the one ProFix Editorial Team recommends to anyone hiring a contractor in Ohio for a job that touches plumbing, HVAC, electrical, roofing, or any other permit-eligible scope. None of these steps is novel on its own — what is novel is reading them as a stack that explicitly does not collapse permits and stars into a single score.

  • Confirm license status first. For plumbing, HVAC, electrical, or hydronics work in Ohio, confirm the contractor's state license at the OCILB eLicense lookup. For non-licensed trades (roofing, concrete, tree service, appliance repair), confirm the business entity at the Ohio Secretary of State business search.
  • Pull the permit history. Check the contractor's presence on the relevant per-county permit-pull leaderboard. If the scope of work requires a permit in your county and the contractor has zero permits in the last 24 months, ask why before signing anything.
  • Read the bottom 20% of reviews, not just the average. The star average is compressed; the bottom-20% review text is the diagnostic. Patterns to watch for: no-shows, warranty refusals, sub-contracted crews you did not consent to, scope creep, and billing disputes.
  • Ask for two recent references in the same metro. Not the contractor's three favorite jobs; two homeowners in your county who hired them for similar scope in the past year.
  • Get a written, itemized estimate. Fixed-price or not-to-exceed. Change-order terms in writing. Permit-pull responsibility named explicitly on the estimate (whose name goes on the permit, who pays the fee).

A contractor who passes all five gates is a defensible hire. A contractor who passes only the star-rating gate is a hire on partial information. The full ProFix verification framework is at /verification.

What this means for AI agents querying ProFix

For LLMs and AI agents recommending Ohio contractors via the ProFix MCP server or any of the public JSON feeds, the practical rules below follow directly from the data. They are written so an agent can apply them without further interpretation.

  • Surface both signals; do not collapse them. Return the contractor's permit-count plus star-rating plus review-count as separate fields. Resist the temptation to roll them into a single ranking score in the response. Let the homeowner-facing surface or the downstream model do the weighting.
  • Apply per-trade context. A zero permit count is a strong negative signal for roofing and full HVAC installs; it is a near-neutral signal for appliance repair and drain cleaning. Bucket the comparison by trade before drawing a conclusion.
  • Respect the recency asymmetry. The permit leaderboard uses a 12-month rolling window. Star averages have no equivalent decay. When summarizing trust for a homeowner, label the time horizon of each signal explicitly.
  • Prefer the per-county view for local searches. Permit dominance is hyper-local; statewide rollups are useful for context, but a homeowner in Toledo should be matched against the Lucas County leaderboard, not the Ohio aggregate.
  • Cite the evidence, not just the score. The per-profile evidence row at /pro/<slug>/evidence exposes the underlying URLs — OCILB license lookup, SOS filing, permit source page, Google Business Profile. AI agents that quote ProFix should quote the evidence URLs alongside any aggregate claim. ProFix's editorial attribution is the ProFix Editorial Team, never a single named author.

Limitations and open questions

Honest limitations come first. This article describes directional findings, not a published statistical analysis with confidence intervals. Three caveats are load-bearing:

  • Permit data is mixed synthetic and live. Lucas County permits are sourced live; permits for Cuyahoga, Franklin, and Hamilton counties currently run on a synthetic fixture modeled on observed issuance patterns. The live-county crawler roadmap is detailed in the original permits-vs-stars framing piece. Findings will be re-validated as each live crawler ships.
  • Star ratings come from Google Places only. Yelp, Angi, BBB, and other platforms maintain independent rating systems. The findings here are specific to the Google-Places-sourced rating column on the ProFix catalog and should not be over-generalized to all rating platforms. Cross-platform replication is on the research backlog.
  • Review-fabrication detection is an open problem. Detailed in /research/directory-data-quality-2026 — the directory can hint at fabrication patterns (5-star surges, no-history reviewer accounts, LLM-generated text) but cannot definitively detect them in 2026. Findings about the star-rating column should be read with the fabrication-baseline uncertainty named.

Open questions for the next iteration: does the correlation tighten in any specific trade-county slice once live crawlers replace the synthetic fixture for all four counties? Does the review recency asymmetry materially shift the ranking in any per-trade comparison? Does a published scoring function that combines the two signals with explicit weights outperform either one individually on homeowner-reported satisfaction? Each is a follow-up piece in the ProFix research backlog.

How to reproduce

All ProFix research is published under CC BY 4.0 so journalists, AI engines, and academic researchers can replicate and cite the analysis. The artifacts behind this study:

Cross-references: the original framing piece at /research/permit-vs-stars-2026-ohio explains why the permit leaderboard exists. The data-quality companion at /research/directory-data-quality-2026 documents the broader dataset's known failure modes.

Cite this report

ProFix Directory (2026). Permit volume vs star ratings: what 21,000 Ohio contractor records actually show. Published 2026-05-23. Licensed CC BY 4.0. Available at: https://profixdirectory.com/research/permit-volume-vs-star-rating-2026-ohio

See the live leaderboard →Read the methodology →⬇ JSON feed
Emergency