Every solar operator quotes one permitting number — "contract to PTO takes us about X weeks" — and it's the least useful number in the building. Total cycle time is an outcome: it moves when its inputs move, and it tells you nothing about which input to fix. The metrics below are the inputs. Track them by jurisdiction and by utility (blended averages hide everything), trend them monthly, and total cycle time takes care of itself.
The seven
1. First-pass approval rate — share of submittals (city and utility, tracked separately) approved without a correction or deficiency cycle. The single highest-leverage number in permitting: every avoided cycle returns one to three weeks. Segment by AHJ and by correction category; the categories tell you what to fix, the AHJ split tells you where.
2. Detection latency — hours from an actionable event (correction posted, deficiency issued, agreement ready, inspection result) to your team acting on it. This is the silent-aging killer, and it's entirely internal — the purest measure of whether the morning sweep is real. Target: same business day.
3. Signature turnaround — days customer-pending documents (tariff applications, interconnection agreements, HOA forms) sit awaiting signature. Cheapest metric to improve, most commonly ignored. Every day here is a day added to PTO with zero external cause.
4. Stage-aging distribution — for each pipeline stage, not the average but the tail: how many jobs sit past that stage's SLA right now, and the P90 duration. Averages flatter you; the tail is where homeowners churn and permits expire.
5. Authority-clock performance — external review durations per city and utility, measured against their own standards (the 45-day line, Oncor's 30-day PTO commitment). This is your escalation evidence and your market-entry intelligence — and it's the number that keeps you honest about whether "the city is slow" or you are.
6. Inspection pass rate — finals passed first attempt, with failure taxonomy. Re-inspections cost fees, truck rolls, and a week-plus of PTO chain each.
7. Rework and expiration incidents — the error ledger: expired permits, out-of-sequence utility filings, wrong-jurisdiction submittals, interconnection amendments filed late. Individually rare, collectively the signature of a process running on memory instead of system. Target: zero, reviewed monthly by name and cause.
Making them real
Three implementation rules separate a KPI program from a slide. Timestamps, not statuses: every metric above derives from stage-transition timestamps; if your tracker can't hold time, the KPIs are unbuildable. Segment or don't bother: a blended first-pass rate across Houston's SolarAPP+ lane and a strict co-op is noise; per-authority is signal. Close the loop: each monthly review ends with one process change aimed at one metric — the KPI program's output is checklist and profile updates, not commentary.
And benchmark honestly: your own trend beats industry folklore, because published "average" timelines rarely match any specific market's reality. This dashboard is, unsurprisingly, what TexPTO renders natively — every metric above computes from the stage timestamps the system already keeps, segmented by the authorities each job actually passed through, so the monthly review starts at "what do we change" instead of "who has the numbers."
FAQ
Which KPI first, if we can only track one? First-pass approval rate by AHJ — it drives the most calendar and its failure categories are directly fixable.
How do we measure detection latency without software? Painfully: portal-check logs against event timestamps. The difficulty of measuring it by hand is itself the argument for stage-level tooling.
What's a good contract-to-PTO number? The one that's shorter than your last quarter's, per market. Cross-company comparisons only mean something segmented by jurisdiction and utility mix.
Who reviews the KPIs? Ops monthly for process changes; leadership quarterly for the trend and the cost model it feeds.
Sources
- Metric mechanics and failure categories are documented across this library: rejections, inspections, deficiencies, expirations, delay economics
Operations guidance — definitions matter more than targets; fix the definitions first, then let the trend set the targets.