The SolarQna Installer Score (0–100)
An objective, mathematical scoring model built to eliminate pay-to-play review spam and sponsored rankings. Every score is derived from verified state licensing records, public interconnection filings, Bayesian customer sentiment, and hardware portfolios.
No contractor can purchase a higher ranking, sponsor placement, or buy removal of negative verified audits.
If a contractor’s C-10/C-46 license is expired, suspended, or uninsured, their score immediately drops to zero.
Prevents astroturfed 5-star ratings with few reviews by mathematically shrinking small samples toward the global prior.
Cross-checks self-reported install volumes against CPUC Rule 21, PM Surya Ghar, and MaStR open government registries.
The Mathematical Formulation
The SolarQna Installer Score utilizes a Weighted Geometric Mean rather than an arithmetic average. This ensures that serious deficiencies in any single sub-score cannot be masked by inflated metrics in another.
1. Track Record Sub-Score (s_track — Weight: 30%)
Evaluates log-scaled public interconnection volume (45%), company operating longevity up to 20 years (35%), and W-2 in-house installation labor vs subcontracting (20%).
2. Customer Sentiment Sub-Score (s_sentiment — Weight: 25%)
Calculated using Bayesian shrinkage: R_bayes = (C·m + n·R) / (C + n), with C = 10 prior reviews and m = 4.2 global median, preventing fake reviews from gaming rankings.
3. Equipment Quality Sub-Score (s_equipment — Weight: 25%)
Evaluates the contractor’s hardware portfolio across verified Tier-1 module manufacturers (50%), smart hybrid inverters (35%), and home storage units (15%).
4. Statutory Gate (T_i) & Verification Ratio (P_i)
T_i is a binary gate: if license is not active and insured, T_i = 0. P_i = min(1.0, N_verified / N_claimed) penalizes companies inflating marketing claims above public utility filings.