Your app shows you numbers. Here's how to turn them into answers, and spot the three failure patterns that matter.
Most monitoring apps show production, and most homeowners glance at it, see a number that isn't zero, and move on. Here's what a professional looks for instead:
Production drops to a new, lower plateau and stays there. This is the signature of a hardware event, a string going offline, a failed microinverter. It won't recover on its own, and every week it persists is money. Per-panel data makes it obvious: one panel dark or dramatically below its neighbors.
A gradual decline over weeks or months, often recovering after rain. That's soiling, dust, pollen, grime. It quietly costs 5–15% of production and a professional cleaning recovers it.
Missing days or weeks in the data usually mean your gateway was offline, not that the system stopped producing. But gaps matter: they blind you to the other two patterns, and long offline stretches can complicate manufacturer warranty eligibility.
Raw numbers mean little without context. Panels facing different directions produce different amounts, a west panel and an east panel should never be compared directly. The honest method compares each panel against neighbors facing the same way, and compares the whole system against a weather-adjusted model of what it should produce. That's exactly how a SolCare health report is built, and it's why our reports say things your app can't.