What do you need to ask before quoting a solar system?
What the customer wants to keep running, for how long without mains power, and what they can spend. Everything technical follows from those three answers: the load determines inverter size, the runtime determines battery capacity, and the budget determines which of the two gets compromised. Customers cannot state a load, so the job is translating what they describe into one.
Key facts
| PARAMETER | VALUE |
|---|---|
| Product | Solar AI (SA-200) |
| Client | Joshville Global Auto & Energy Ltd |
| Inputs | Load · Runtime · Budget |
| Output | Specification tied to real stock |
Nobody knows what size they need
Ask a customer what solar system they want and the answer is almost never useful. They will name a figure they heard from a neighbour, or a budget, or simply say they want their house to work when the power goes.
This is not the customer being unhelpful. System size is expressed in units that have no meaning outside the industry. Nobody experiences kilowatt-hours. They experience the fridge going off, the router dying at nine, and the fan stopping in the middle of the night.
So the qualifying conversation is a translation problem. Three questions carry almost all of it.
What do you want to keep running?
This produces the load, and the honest version of the answer is a list of appliances, not a number.
The useful technique is to ask what must survive an outage rather than what the customer owns. Those are different lists, and the gap between them is usually where the budget is won or lost. A customer who says "everything" is describing an aspiration; a customer who says "the fridge, the lights, the router and the fan" has described a system that can actually be quoted.
Two things reliably distort this list. Air conditioning, which is often assumed into the requirement and multiplies the cost of everything downstream. And appliances with high starting loads — pumps, compressors, freezers — which draw far more at the moment of switching on than their running figure suggests. A system sized to running loads will trip the first time the borehole pump starts.
How long without mains?
This is the question most often skipped, and it determines battery capacity, which is usually the largest single cost in the system.
Two customers with identical appliance lists need very different systems depending on whether they need four hours of cover in the evening or full overnight autonomy. Same load, substantially different battery bank, substantially different price.
It also has to be asked about the worst case rather than the typical one. Everyone describes their normal outage. The system has to survive the bad week, and the customer's tolerance for that bad week is a commercial decision they should be making consciously rather than discovering later.
What can you spend?
Engineers resist putting budget in a technical conversation. It belongs there, because budget decides which constraint gives way.
A specification produced without a budget is frequently correct and unsellable. The customer sees a number far above what they had in mind, and the conversation restarts from zero — except now they have been told the honest answer is unaffordable, which makes them more likely to buy something inadequate from someone less scrupulous.
Knowing the ceiling early lets the conversation become an explicit trade: this budget covers the fridge, lights and router overnight, but not the air conditioning. That is a proposal a customer can accept or adjust. It is also honest about what the system will not do, which is the thing that prevents the angriest possible customer six months later.
Then, and only then, the catalogue
Once load, runtime and budget are known, specification is largely mechanical: load determines inverter capacity, runtime determines storage, budget arbitrates.
The part that must not be mechanical is what fills the specification. A quote is only useful if it names equipment the business actually holds, at the price it actually sells for in that market. A theoretically correct system built from components nobody can source is not a quote — it is a physics exercise.
This is why the assistant in Solar AI selects from the live catalogue rather than generating a recommendation from general knowledge. The conversation above is the part a language model does genuinely well: asking follow-ups, handling vague answers, converting "my fridge, some lights and the TV" into a load profile. What it must not do is invent the products or the prices at the end of it.
The shape of a good qualifying conversation
Three answers, in order, each constraining the next:
- Load — what must keep running, with starting loads accounted for
- Runtime — how long without mains, measured against the bad week
- Budget — the ceiling, stated before the specification is built
Everything after that is selection from stock. Everything before it is guesswork dressed up as engineering.
Frequently asked
- Why can't customers just say what size system they need?
- Because system size is expressed in units nobody encounters outside the industry. Customers know which appliances matter to them and roughly how long the power stays off. Converting that into a load profile and a capacity figure is the expertise they are paying for.
- What is the most common mistake when sizing solar?
- Sizing to the appliance list without asking about runtime. Two customers with identical appliances need very different battery capacity depending on whether they need four hours of backup or overnight autonomy, and battery is usually the largest cost in the system.
- Why does budget belong in the technical conversation?
- Because it decides which constraint gives way. A quote that ignores budget produces a correct specification the customer will not buy, and the sales conversation restarts from nothing. Knowing the ceiling early lets you propose a system that is honest about what it will and will not carry.