How much does solar fleet monitoring cost? The pricing models, explained by a vendor

Monitoring pricing is oddly hard to research. Most vendors publish a demo form instead of a price, so an installer trying to budget for fleet oversight ends up collecting quotes just to learn what the market charges. We publish our prices, so this piece uses them as a worked example — not because ours are the only reasonable numbers, but because a concrete number you can check beats a range nobody will commit to.

Start with what is genuinely free. Every major inverter manufacturer ships a monitoring portal with its hardware, and per system, those portals are good. If you run twenty systems on one brand, the free portal is probably the right answer and you should not pay anyone. The cost appears at fleet scale, and it appears twice: once as labour, because somebody has to open one portal per brand every week, and once as the faults nobody catches, because in practice coverage collapses to whichever systems a customer complained about. That cost is real but it never shows up as a subscription line, which is why fleets underestimate it.

Paid fleet monitoring is priced three ways. Per system per month is the most common for residential and commercial fleets — the fee scales with your install base. Per kilowatt shows up at utility scale, where a plant is the unit that matters. And some platforms charge a flat monthly fee with a capacity allowance. The differences sound cosmetic and are not: per-seat surcharges, brand-adapter fees, and overage terms can double an apparently cheap quote once a real fleet hits them.

Our own structure, as the example: Sol AI is $1.50 per system per month, across the whole installed fleet — no tiers, no capacity allowances, and unlimited users because we do not charge per seat. Priced per system because the product only works fleet-wide: every system you've installed, watched, not a subset. Rae, our AI lead-response product, is priced separately ($749 a month flat, or $99 per AI-booked appointment with no monthly fee) — worth naming here only so its tiers are not mistaken for monitoring prices.

Whether roughly three dollars a system is expensive depends entirely on one question: does the platform find losses that pay for it? The case we keep returning to is a 29 kW commercial system that ran for 22 days with one of its three strings disconnected — a third of the system producing nothing — while its manufacturer portal showed green the whole time. Nobody needs a model to price a third of a commercial system's output for three weeks against three dollars a month. A platform that only aggregates dashboards cannot make that argument; a platform that compares each system against a weather-adjusted baseline can. When you evaluate cost, that is the division to do — fee per system against the production the caught faults actually recover.

The number quotes hide most often is implementation. Reading a mixed fleet means connecting each manufacturer's cloud, importing history, and building the baselines that make alerts mean something — real work, usually billed separately, sometimes surfaced only at contract time. Ours is $10,000 standard, and $5,000 for founding partners who register before August 31, with the first 90 days of the platform included. Whatever vendor you talk to, ask for the implementation figure in writing before comparing monthly rates; a low subscription with an undisclosed setup fee is not a low price.

To compare quotes like for like, normalise everything to cost per system per month at your fleet size: monthly fee, plus any per-seat and overage charges at your actual headcount and growth, plus implementation spread over the contract length. Then ask each vendor the questions that expose the differences pricing pages blur: is there a per-seat fee, what happens at the capacity boundary, which brands are actually read in production today, and what did the last customer pay in total in year one. Vendors with clean answers answer quickly.