How We Check a Solar Quote

The service extracts only fields that appear in the uploaded proposal, such as total price, system size, equipment models, production claims, warranty terms, payment schedule, and financing disclosures. Missing information remains missing; the extractor is instructed not to infer or calculate absent values.

Price per watt uses the same-scope pre-incentive cash price divided by DC system watts. A battery, roof replacement, electrical upgrade, or other material non-PV scope is subtracted only when its price is separately itemized. A financed proposal is not screened without an explicit same-scope cash price. The quote's claimed post-incentive net price is displayed as extracted evidence, but is not converted into a comparable net price per watt because program eligibility has not been verified.

When the printed amount financed is at least 5% above the printed cash price for the same system, the report asks for a line-by-line reconciliation. A difference of 20% or more is treated as a high-priority itemization question before signing. The arithmetic difference is not automatically labeled a dealer fee and is not, by itself, described as improper.

The current price screen uses Berkeley Lab's published 2024 U.S. distribution for host-owned, stand-alone residential systems. It shows the 20th percentile, median, and 80th percentile as a historical national reference. The source distribution includes both cash and loan-financed transactions; our use of a same-scope cash basis is a conservative normalization choice, not a claim that the source is cash-only. It is not a state quote, a current installer bid, or a verdict on value. Roof work, electrical scope, equipment, financing, permitting, and local conditions can explain differences.

Production may use NREL PVWatts, Google Solar roof data, or a generic 1,400 kWh-per-kW placeholder. Reports name the source. The current PVWatts integration uses generic tilt, azimuth, array-type, and loss inputs, so it is a sensitivity scenario and cannot support a mismatch flag. Only Google Solar data tied to a full address and matching ZIP, an exact rooftop geocode, strong imagery, and the exact proposed array may be compared with an installer production claim. The array count must be stated in the quote or safely derived from the quote's system size and panel wattage; Google's assumed panel wattage is never used to infer it. Panel count, wattage, and system size must reconcile within the larger of 100 W or 2% to allow ordinary proposal rounding. A nearest or roof-capped Google configuration is not substituted into the comparison or retail-value arithmetic. The generic placeholder is rough arithmetic only.

The report does not calculate expected savings or a forecast payback period. It may show a full-retail-rate value scenario: scenario production, capped at annual usage, multiplied by a bill-derived rate or published state residential average. The corresponding scenario years always use the pre-incentive solar cash basis. This deliberately simplified screen omits self-consumption versus exports, time-of-use periods, fixed charges, degradation, financing, future tariffs, and unverified incentive eligibility, and is labeled accordingly.

A fixed rule set checks completeness, equipment transparency, production evidence, warranties, financing clarity, payment terms, and incentive assumptions. Flags are questions to resolve, not accusations. Category weights are: published price context 30%, production evidence 20%, equipment transparency 15%, warranty clarity 10%, financing clarity 10%, incentive clarity 10%, and document completeness 5%. A high-, medium-, or low-priority flag deducts 40, 20, or 10 points from its assessed category, with a floor of zero.

Price and production categories show “Not assessed” when the required evidence is unavailable; absence of evidence never earns 100. The overall score is normalized across assessed weights. Evidence coverage below 70% caps the score at 79, and below 50% caps it at 59. One high-priority flag caps the total at 79; two or more cap it at 59. The score is a document-completeness and clarity screen—not an endorsement of the installer, equipment, economics, or contract. Read it with the underlying findings. See the fictional sample for the format.