The honest answer, first
For most roofs, rain does most of the work. Panels sit at an angle, glass is deliberately smooth, and ordinary airborne dust washes off. If you live somewhere with regular rain, the case for climbing onto your roof with a bucket is weak.
It matters more when the dirt is not ordinary dust: a long dry season with no rain at all, farm or construction dust, wildfire ash, heavy bird droppings, or a very shallow tilt where water pools instead of running off. Those are real, and on those roofs cleaning can be worth doing.
⚠️ We are not going to give you a percentage. Every number you will find for "how much cleaning gains" comes from either a company that sells cleaning or a lab that controlled the conditions — neither is your roof. The useful thing is not a number from someone else. It is a measurement from yours.
Look at the shape of the loss, not the size
Dirt accumulates slowly and lands on everything, so soiling looks like a gradual, system-wide decline that recovers after heavy rain. That shape is the signature — not any single day's number.
A single panel falling while the ones beside it hold steady is not dirt. Dust does not pick one module. That shape means shade — a branch, a vent, a new neighbouring roofline — or the hardware itself.
⚠️ Be clear about what monitoring can and cannot do here, because plenty of marketing is not. From production figures alone, nobody can tell dirt from shade from a failing microinverter — the numbers look similar. What per-panel data does is tell you which of those shapes you have, which is the difference between cleaning your roof and calling an installer.
How to tell, without going anywhere near the roof
You do not need to clean anything to find out whether dirt is costing you. There are two signals that come free with the sky.
1. Output against available sunlight. A dirty array does not fail — it drifts. Measured against how much sun there actually was each day, soiling looks like a slow, steady slide over weeks, spread evenly across every panel. The weather is already accounted for in that comparison, which is what makes the drift visible instead of lost in the noise of cloudy days.
2. What rain does next. This is the one that settles it. Rain is a cleaning event you did not pay for, on a date you know. If production steps back up after a good storm, dirt was costing you. If it does not move, whatever you were losing was never dirt — a branch does not move in the rain, and a failing microinverter does not heal.
⚠️ That second signal is seasonal, and we should be honest about it: through a long dry summer there may be no rain to use at all. In that stretch the drift is what you have, and the storm that ends the season is what confirms it.
If you want to prove it on your own roof
There is a clean experiment available if you are already up there safely — and only then. Clean one or two panels on a face and leave the rest. The panels you did not touch are the control: same sun, same angle, same weather, same day. Any gap that opens between them is the cleaning and nothing else.
Compare each panel's total energy per day for a week either side — the daily total, not instantaneous power, which moves with every passing cloud.
⚠️ We are not suggesting you climb onto a roof to run an experiment. Falls are the real hazard on this page, and the two signals above answer the same question from the ground. This is here because some people will want the direct proof — and because it is the test we are running ourselves.
Before you get on the roof
The panels are rarely the risk — the roof is. Falls are the actual hazard here, and for a lot of owners that alone settles it.
Do not use a pressure washer. Forcing water past a panel's seals turns a cleaning job into a replacement. Check your warranty before anything abrasive goes near the glass, and do it on a cool panel rather than a hot one. Plain water, from the ground, is the version with almost no downside.
And if a company offers you a cleaning schedule, ask what they measured to arrive at it. A schedule is a product; a measurement is an answer.
Measuring it is the part we can help with
The WattHacker monitor logs every panel's energy every day, and the sunlight and rainfall at your address alongside it — which is what turns "are they dirty?" into something you can read instead of guess.
- Per-panel daily energy, kept and charted — every panel, not one number for the whole roof.
- The weather at your address, logged daily, so a drop can be compared against the sunlight that was actually available.
- The monitor is $50 once, with a 30-day money-back guarantee.
- ⚠️ It will not announce that your panels are dirty — from production alone nothing can. It shows whether the loss is system-wide or one panel, and whether rain took it away.
Still have questions about your own system? I'll go through it with you — your panels, your bill, and what's worth doing about it.
Book an energy reviewReal panels, real production, no signup.
Common questions
Does cleaning solar panels actually make a difference?
On most roofs, far less than the people selling it suggest. Rain removes ordinary dust, and panels are mounted at an angle that helps it run off. Cleaning makes a real difference in specific conditions — a long dry season, farm or construction dust, wildfire ash, bird droppings, or a very shallow tilt where water sits instead of running. The honest answer is that it depends on your roof, which is why measuring beats guessing.
How can I tell if my panels are dirty enough to matter?
Look at the shape of the loss, not the amount. Dirt builds slowly and affects the whole array, so it shows as a gradual system-wide decline that recovers after heavy rain. A single panel dropping while the ones either side of it hold steady is not dirt — that is shade or hardware. No monitoring can name the cause from production alone; what it can do is tell you which of those two shapes you are looking at.
How do I know if dirt is costing me anything, without getting on the roof?
Two signals, and neither needs a ladder. First, compare what your panels produce against how much sunlight was actually available that day: soiling shows up as a slow drift downward over a dry spell, with the weather already accounted for. Second, watch what heavy rain does. Rain is a free cleaning event on a known date — if output steps back up afterwards, dirt was costing you something. If it does not move, the loss was never dirt.
Can monitoring tell dirt from shade or a broken panel?
Not from a production number on its own — they look alike, and anything claiming otherwise is overselling. What separates them is time and rain. Soiling is gradual, affects the whole array, and RECOVERS after a storm. Shade is gradual and seasonal but never recovers from rain. A failing microinverter is usually one panel, often sudden, and never recovers at all. Per-panel daily energy plus the rainfall at your address is enough to tell those three shapes apart.
Is it safe to clean them myself?
The panels are usually fine; the roof is the risk. Falls are the real hazard, and they are why many owners should not be up there at all. Do not use a pressure washer — forcing water past the seals is a way to turn a cleaning job into a replacement — and check your warranty before anything abrasive touches the glass. Plain water from the ground, on a cool panel, is the low-risk version.
Should I pay a company to clean them?
Sometimes, and it is worth knowing what you are buying first. If you have real soiling and no safe roof access, paying someone is reasonable. What is not reasonable is paying on a schedule for a problem nobody has measured. Ask for before-and-after production data, and if they cannot provide it, you can measure it yourself with the control-panel method above.
See your own panels, live.
See plans Try the demo Back to home