Three quotes, one scope, and an 80% spread
Last March I sent the same one-page scope to three installers for a 20 kWh battery backup system at our 18,000-square-foot facility. Same critical-load panel, same four-hour backup window, same everything. The bids came back at $21,400, $27,900, and $38,900.
Finance asked the obvious question: why wouldn't we take the low bid?
I didn't have an answer that fit on one line. And that's usually the first sign something's off with how you're comparing them. When three numbers for “the same thing” spread 80% apart, at least two of them aren't actually priced for the same thing.
I've managed facilities and vendor spend for a 140-person company since 2020—roughly $600K a year across 14 vendors, reporting to both operations and finance. I've bought plenty of equipment where the sticker price turned out to be the least important number on the page. Battery backup is where it finally became impossible to ignore.
The surface problem: we compare prices, not scopes
Everybody does it—honestly, it's not laziness. The quote is one number. The scope is thirty pages of line items that mostly look identical. So your eye goes to the total, and the total says the low bid saves $6,500.
Except it doesn't, because the three proposals aren't three prices for one system. They're three different systems with three different assumptions, and only one of them wrote those assumptions down where a buyer would actually see them.
It took me four years and close to 30 vendor quotes across battery, EV charging, and fleet equipment to understand that the number at the bottom of a proposal tells you almost nothing on its own.
What's actually different between two battery quotes
When I finally sat down with the three proposals and read past the totals, the gap stopped being mysterious. It came down to three things.
1. You're buying cycles, not kilowatt-hours
Battery quotes are usually priced per kWh, which makes them look comparable. They aren't. What you're really buying is usable kilowatt-hours multiplied by the number of cycles the battery will deliver before it hits end of life.
The low bid used a cell rated for about 3,000 cycles to 80% capacity. The premium bid used an LFP cell rated for 6,000. Same 20 kWh nameplate, roughly double the lifetime throughput on the second one—which means the $6,500 “cheaper” option was actually more expensive per delivered kilowatt-hour. The spreadsheet said take the low bid. My gut said the quote looked too clean. The cycle-count line is what settled it.
2. Nameplate capacity isn't usable capacity
One proposal quoted 20 kWh and delivered 16 kWh of usable storage, because it limited depth of discharge to 80% to protect the cells. The other delivered about 19 kWh at 95% depth of discharge. Same headline number, 15–20% difference in what you actually get when the grid goes down.
On paper that's a spec detail. In an outage, that's the difference between riding out a four-hour event and dropping load at hour three. (Note to self: never size a system off nameplate again.)
3. The inverter question nobody asks
This was the real red flag. The low bid assumed we already had a compatible hybrid inverter. We didn't. Adding one came to about $6,800, which erased most of the savings. The premium bid included it.
There's a second layer, too: most battery warranties only hold if the battery is paired with an inverter on the manufacturer's approved list. If your inverter isn't on that list, the warranty is void—and you won't find that out until year four, when you need it. On top of that, 48V systems and high-voltage systems aren't interchangeable. Smaller, modular installs tend to run 48V; bigger ones usually go high-voltage for efficiency and fewer parallel strings. A quote that doesn't say which architecture it's pricing is incomplete, full stop.
What the cheap system cost us
I learned most of this the hard way. In 2022 we went with the low bid on a smaller 10 kWh system for our second location—roughly $9,000 versus $13,500. It passed commissioning fine. Year one was fine. By year three the capacity had faded noticeably, and during a two-day storm outage it dropped the critical load at about the 90-minute mark.
What that cost: roughly $4,200 in emergency generator rental, a replacement system in 2025 at full price, and a round of complaints from the two tenants who share that building. Total cost of the “cheap” system over three years came out meaningfully higher than the premium bid would have been.
The part I didn't budget for was the impression. When the power goes out, the backup system is the only piece of our operation anyone experiences. It either holds or it doesn't, and that's the whole story as far as the people in that building are concerned. We spent years building a reputation for being reliable, and one failed outage did more damage to it than any marketing ever did for it. The money we saved on the system got spent back on the impression it made.
How I compare battery quotes now
I don't compare totals anymore. I compare five lines, and if a vendor can't supply them, the quote goes in the “not comparable” pile:
- Cycle life at a stated depth of discharge (not “long life”)
- Usable capacity, not nameplate
- Chemistry (LFP vs. lead-acid vs. NMC) and the C-rate it's rated for
- Approved inverter list, in writing
- Warranty terms, including what voids them
As a benchmark, the spec sheet I keep coming back to is Pylontech's US5000—a 48V LiFePO4 module rated at 4.8 kWh nominal, around 95% usable capacity, and 6,000 cycles at 90% depth of discharge, in a modular format you stack rather than replace as a cabinet. Pylontech's Force H2/L2 line covers the high-voltage side for bigger installs. I'm not saying it's the only option on the market. I'm saying those are the numbers I now require from every quote, because a proposal that doesn't list them isn't something I can compare to anything.
Also worth checking: UL 1973 and UL 9540A listings. Most AHJs ask about them now, and a battery without them can stall your permit for weeks.
The same trap shows up everywhere else
Once I saw the pattern, I couldn't unsee it. “How much does EV charger installation cost” is the same question in a different costume. The honest answer is that it depends—on panel capacity, conduit runs, permitting, and whether you need a service upgrade. When we priced six Level 2 stalls for the staff lot, three quotes for “the same job” came back more than 40% apart. Two of them had excluded the panel upgrade. Same story as the batteries.
Our fleet manager hit it too, pricing a remote tire pressure monitoring system. The cheapest per-sensor option carried a monthly subscription and sensors with two-year batteries. Five-year total cost came out higher than the system that looked pricier on day one. Not a no-brainer after all.
It even applies to smaller projects. If you're pricing a DIY whole-home battery backup for yourself, the listing pages quote nameplate kWh and skip the depth-of-discharge line—same trap, smaller scale.
Bottom line
The low bid isn't a price. It's a scope with the expensive parts left out.
A quote is a number. A spec sheet is a decision. If you're comparing numbers without the sheets, you're not shopping—you're guessing with someone else's money.
That's the whole lesson from four years of this. The cheapest quote is a hypothesis. The spec sheet is what turns it into a fact.