Auto repair is the worst-scoring of the 30 trades we can rank.
Which kind of repair quote is most likely to be overpriced?
Auto repair. Across 783 auto repair quotes the average fairness score is 48.6 out of 100 and 55.8% scored under 50, meaning at least one line was priced above the normal market range for that repair and region. Home service quotes, 2,828 of them across the other trades, are flagged 37% of the time. Auto is the lowest-scoring of the 30 trades with enough quotes to rank.
Caveat: These are quotes people chose to check, not a random sample of all repair work, and the auto lane carries a larger share of quotes gathered from public posts than the home trades do.
Getting a second quote does very different amounts of work depending on the trade.
Does getting a second quote protect you?
Not equally. Treating two quotes as independent draws, both come back overpriced about 31.1% of the time in auto work against about 5.8% in pool. The standard advice to get a second opinion is weakest in exactly the trade where people need it most.
Caveat: This squares each trade's flag rate, which assumes the two quotes are independent. Two shops in one market reading the same parts catalogue are not fully independent, so treat it as the shape of the risk rather than a precise probability.
There are two different kinds of overcharging, and they need different defenses.
Is overcharging a few bad actors or the whole market?
Both, depending on the trade. In 27 of the 30 trades we can rank, the average score sits below the median: most quotes are reasonable and a minority of bad ones drag the average down, so the job is screening the individual contractor. In 3 (Auto, Siding, Concrete) the average sits at or above the median, which means the whole price level has moved and screening one contractor will not help. There an outside benchmark is the only defense.
Caveat: Mean against median is a coarse shape test on scores, not a distribution fit, and trades near the boundary can move across it as quotes are added.
Inside a single trade the verdicts run the full length of the scale.
Does a trade average tell you what your own quote is worth?
No. Inside one trade the scores span nearly the whole scale: auto (SD 24, 55.8% under 50, 21.6% at 70 or better); roofing (SD 25, 43.3% under 50, 29.8% at 70 or better); plumbing (SD 25, 46.5% under 50, 31.4% at 70 or better). A large block of quotes in the same trade is priced fairly and a large block is not, which is precisely what a homeowner holding one quote cannot see. The trade average is close to useless for judging the piece of paper in your hand.
Caveat: Standard deviation is computed across scores, not dollars, and each trade mixes job sizes.
41.1% of the quotes in this dataset score under 50 out of 100.
How common is an overpriced quote?
41.1% of the 3,644 scored quotes in this dataset come in under 50 out of 100, and the average score is 56.5. That figure gets worse, not better, when the oldest batch is removed: the 1,550 quotes loaded before 2026-03-01 average 61.2, while the 2,094 scored since then average 53 and are flagged 46.4% of the time.
Caveat: Sample composition matters and is stated rather than buried: an early seed batch, a set of quotes gathered from public posts, and quotes uploaded by people who came to the site. All three are people or posts with a reason to question a price.
No trade we can rank comes in clean.
Is there a trade where quotes are reliably fair?
There is no safe trade. The cleanest trade we can rank is pool at 24.1% of quotes flagged, and the worst is auto at 55.8%. That is a spread of only 2.3x across 30 trades, and not one of them falls below one flagged quote in five.
Caveat: Read this conditionally: when someone is uneasy enough to check a quote, they are right a meaningful share of the time even in the best trade. It does not say a fifth of all repair work is overpriced.
Every trade we can rank
| Trade | Quotes | Avg score | Median | Flagged | Both of two |
|---|---|---|---|---|---|
| Auto | 783 | 48.6 | 45 | 55.8% | 31.1% |
| Siding | 58 | 48.9 | 45 | 53.4% | 28.5% |
| Paving | 61 | 52.4 | 55 | 47.5% | 22.6% |
| Roofing | 319 | 53.2 | 55 | 43.3% | 18.7% |
| Plumbing | 258 | 53.3 | 55 | 46.5% | 21.6% |
| Windows | 104 | 55.4 | 65 | 38.5% | 14.8% |
| Bathroom | 77 | 55.8 | 60 | 39% | 15.2% |
| Concrete | 94 | 55.9 | 55 | 45.7% | 20.9% |
| Septic | 53 | 56.7 | 65 | 39.6% | 15.7% |
| Construction | 122 | 57.2 | 65 | 35.2% | 12.4% |
| Moving | 55 | 57.2 | 60 | 45.5% | 20.7% |
| Kitchen | 67 | 57.4 | 60 | 40.3% | 16.2% |
| Fencing | 57 | 57.6 | 65 | 35.1% | 12.3% |
| Decks | 92 | 57.8 | 65 | 33.7% | 11.4% |
| Landscaping | 182 | 58.2 | 65 | 38.5% | 14.8% |
| Chimney | 50 | 59.2 | 65 | 40% | 16% |
| Other | 84 | 60 | 65 | 28.6% | 8.2% |
| HVAC | 212 | 60.3 | 65 | 32.5% | 10.6% |
| Drywall | 69 | 60.4 | 65 | 34.8% | 12.1% |
| Appliance Repair | 59 | 60.6 | 65 | 32.2% | 10.4% |
| Gutters | 84 | 61 | 65 | 28.6% | 8.2% |
| Flooring | 67 | 62.2 | 70 | 35.8% | 12.8% |
| Restoration | 63 | 63.3 | 65 | 31.7% | 10% |
| Electrical | 133 | 64.3 | 68 | 24.8% | 6.2% |
| Painting | 117 | 65 | 75 | 32.5% | 10.6% |
| Pest Control | 66 | 66.5 | 75 | 30.3% | 9.2% |
| Garage Door | 63 | 66.7 | 75 | 28.6% | 8.2% |
| Cleaning | 51 | 67.6 | 75 | 31.4% | 9.9% |
| Insulation | 55 | 68.3 | 75 | 27.3% | 7.5% |
| Pool | 54 | 68.4 | 75 | 24.1% | 5.8% |
Trades with at least 50 scored quotes, ordered worst average first. "Flagged" is the share scoring under 50 out of 100. "Both of two" squares that rate: the chance that two quotes drawn independently in that trade are both flagged. The full 33-category table, including trades too thin to rank here, is in the open dataset.
Methodology
What a quote is here
A quote is a real contractor, repair or auto estimate that a person uploaded to QuoteScore as a photo, a PDF or pasted text. Each line item is compared against regional pricing benchmarks and the other quotes in the archive, producing a 0 to 100 fairness score. A score under 50 means at least one line was priced above the normal market range for that work and region.
What is counted
- 3,644 scored quotes across 33 categories, covering 2026-02-18 to 2026-09-08.
- Uploads that turned out to hold no readable quote are scored as nothing and are excluded from every figure on this page.
- Trade level claims use only trades with at least 50 scored quotes.
Sample composition, stated plainly
This is not a random sample of American repair work and no claim here should be read as one. It mixes an early seed batch of 1,550 quotes loaded before 2026-03-01, a set of quotes gathered from public posts where people shared a price, and quotes uploaded by people who came to the site with a quote in hand. All three skew toward prices someone already had a reason to question. The correct reading of every percentage here is conditional: among quotes people thought worth checking, this is how often they were right.
Numbers we deliberately do not publish
We do not publish a total dollars overcharged figure or an average overpayment per quote. Those require summing benchmark gaps across quotes whose line items are not uniformly itemized, and any single number would be an estimate dressed as a measurement. The percentages and scores on this page are what the data actually supports.
Reproducing this
The aggregate dataset behind every figure is public and free to reuse with attribution: benchmarks.csv (CC BY 4.0), documented at quotescore.ai/dataset. The figures on this page are computed live from the same table and the same threshold as that file, so the two reconcile on any given day. Figures last computed 2026-09-08.
Contact
Matt McCormack, founder of QuoteScore. Journalists: custom cuts by trade or region, anonymized case files, and interviews are available. See quotescore.ai/press.
Check the quote in your hand
Free, no signup, and your file is deleted after analysis.
Score my quote