Solar quotes are hard to compare because they are not really price lists. Each one contains a forecast of what the system will produce and what that will be worth, and those forecasts are built on assumptions that mostly do not appear on the page.
Two quotes for the same roof can differ by thirty per cent in projected savings without either one being dishonest. They just assumed different things.
Here is what to check, roughly in the order that it changes the answer.
1. Yield per kWp: the thirty-second test
Take the quoted annual generation in kWh and divide it by the array size in kWp. For a south-facing array at a normal pitch in central England, that ratio should land somewhere around 900 to 1,000.
Materially above 1,000 means one of three things: a favourable orientation assumption, a low system-loss assumption, or no allowance for shading. Any of those might be legitimate. All of them should be stated.
Materially below 900 is not automatically a problem either. A shaded or east-west roof genuinely produces less, and a quote that says so is being honest rather than pessimistic. What matters is whether the quote explains which.
The full loss chain behind that ratio, term by term, is in why your panels produce less than the quote said.
2. Array size against your consumption
This is where oversizing shows up, and it is the most expensive error in domestic solar because it is invisible in the projected savings if self-consumption has been assumed generously.
The logic is short. A unit you consume is worth the retail price you did not pay. A unit you export is worth your export rate, which under the Smart Export Guarantee is set by your supplier and is typically a fraction of retail. So the value of an extra panel falls as the array grows past what the household can absorb.
| Situation | Where the extra generation goes | What it is worth |
|---|---|---|
| Array well under annual consumption | Mostly consumed on site, especially in winter | Close to the retail import price |
| Array near annual consumption | Split, with summer midday surplus exported | Blended, falling as the array grows |
| Array well over annual consumption | Overwhelmingly exported at midday in summer | Close to the export rate |
Annual totals hide this. A system can match annual consumption exactly and still export most of its summer output, because generation and consumption do not line up hour by hour.
Ask for the quote to state annual generation and your annual consumption side by side. If generation exceeds consumption on a household without an electric vehicle, a heat pump or a battery, the marginal panels are being sold on export revenue, and that should be said out loud rather than folded into a savings total.
3. The self-consumption assumption
Every savings figure in every solar quote depends on what share of generation is assumed to be used in the house rather than exported. It moves the answer more than any other input, and it is the input most often omitted.
For a household with no storage, generation peaks in the middle of the day and consumption does not. The share actually self-consumed is commonly well below half, and it falls as the array gets bigger. A quote assuming a high self-consumption share for a house that is empty on weekdays is assuming away the central problem.
The underlying economics are in why a unit you use is worth more than a unit you export.
4. Inverter sizing
The inverter is usually rated below the array, and that is deliberate rather than a mistake. An array almost never reaches its rated output, so an inverter sized at 100% of array kWp spends its life underloaded, and inverters are least efficient at low input.
A DC to AC ratio somewhere around 1.0 to 1.25 is conventional. Well above that starts clipping real generation on the best days. Well below it means paying for inverter capacity that never gets used.
What matters more than the ratio is whether the quote mentions it at all, and whether the inverter choice matches the roof. On an array with shading or multiple orientations, a single string inverter behaves differently from per-module electronics, and that decision is worked through in microinverters against a string inverter.
5. What is missing from the price
A quote that is cheaper than the others is sometimes cheaper because it does not include things the others do.
- Scaffolding. Required on most installations and a substantial line item. If it is absent, ask whether it is excluded or genuinely not needed.
- The shading survey. If nobody has measured the skyline from the roof plane, the yield figure is modelled on an unobstructed roof. See what shade costs.
- Network operator notification. In Great Britain, installations up to 16 A per phase, roughly 3.68 kW single phase, connect under G98 and are notified afterwards. Anything larger falls under G99 and requires the distribution network operator to approve it before connection. A quote for a system above that threshold that does not mention an application has skipped a step that can delay or limit the installation.
- MCS certification. Needed to claim under the Smart Export Guarantee. Without it, the export payments in the savings projection are not available.
- Export metering setup. Payment under the Smart Export Guarantee is made on metered export readings, so the export arrangement has to actually exist.
- Bird protection, cable management, isolators, and making good. Small individually and the usual source of a final invoice that does not match the quote.
6. The financial projection
Most quotes project savings over 20 or 25 years. Three things are commonly wrong with the projection, and all three flatter it.
- 1Flat generation across the whole term. Panels degrade. At the measured median rate an array is down about 12% by year 25, and using year-one output for all 25 years overstates the total. See how much output panels lose.
- 2An assumed electricity price rise compounded annually. A projection that compounds energy price inflation for 25 years is largely reporting that assumption back to you. Ask what rate was used and what the total looks like at 0%.
- 3A single number rather than a range. Weather alone moves annual output by around 3.5% either side, year to year, before any other uncertainty. A payback period quoted to the month is claiming precision that is not there.
A battery, if one is included, needs its own scrutiny, because it is usually the largest single line and the one whose payback is most sensitive to consumption pattern. Whether it earns its place is covered in do you need a home battery, and the cases where it does not are in when a home battery is not worth buying.
A short list to send back
Five questions, answerable in a paragraph by anyone who has actually modelled the job.
- 1What annual generation, and what yield per kWp does that imply?
- 2What self-consumption percentage did the savings figure assume?
- 3Was shading surveyed from the roof plane, and what did it measure?
- 4What export rate was assumed, and is that my current rate or an assumed one?
- 5What does the payback look like with no energy price inflation, and with panel degradation applied?
A quote that can answer those has done the work. Our calculator produces the same figures independently from your own roof and consumption, which makes it a way to check a quote rather than a replacement for one, and it reports a range with the lower bound stated rather than a single number.
Great Britain connection standards G98 and G99 are published by the Energy Networks Association. Smart Export Guarantee eligibility and metering requirements are administered by Ofgem. Checked 2026. Source