The Client Never Sees Your Spreadsheet — They See What You Installed

A procurement manager's argument for paying for component quality: why cheap inverters, controllers and panels cost more in callbacks, why a Victron Energy solar system needs surplus routing (battery full, heat hot water) scoped up front, and what a North Carolina pallet racking system taught me about hidden costs on an EV charging station build.

I'll say the part that earns me side-eye in procurement meetings: the cheapest system you can quote is almost never the cheapest one you can own — and it's never the one your client judges you by.

Some context on where that opinion comes from. I'm the procurement manager at a 40-person energy systems integrator. I've owned our equipment budget — roughly $2.4M a year — for six years, negotiated with 30+ vendors, and logged every purchase order in the same tracking sheet since 2019. When I pulled the 2023 numbers for our ops review, 61% of our unplanned field labor traced back to line items where we'd deliberately picked the cheaper option to protect a margin.

Those "savings" came to about $19,000 across the year. The rework came to roughly $71,000 in labor, travel and lost billable hours. I didn't need a sophisticated model to see which side of that trade I was on.

The number on the PO isn't the number on the job

Most buyers I deal with focus on the per-unit price and completely miss the three costs that follow it: commissioning time, callback labor, and the customer conversation you have to have in month four. The question everyone asks is "what's your best price on this controller?" The question they should be asking is "what does this controller do to my field hours?"

A real example from last spring. We quoted a warehouse retrofit outside Greensboro, North Carolina. The scope had a North Carolina pallet racking system reconfiguration on it — aisle widths changing, rack being moved — plus 96 modules going on the roof. Racking went first. That moved two of our conduit runs, added roughly 400 feet of conduit and pipe, and put eleven hours of labor into a job where the electrical estimate had been signed off six weeks earlier. Nothing in that job was technically hard. It was a sequencing problem, and sequencing problems don't show up in a unit price.

I've watched the same thing happen on hardware. In 2022 we trialed a lower-cost controller line across 14 sites to see whether the $34-per-unit delta was real money. It wasn't. We spent the following 18 months chasing firmware quirks, explaining reading discrepancies to clients, and eating two truck rolls per site on average. We didn't finish that trial with a cheaper system. We finished it with a more expensive one.

Clients judge the system by the three things they touch

Here's the thing I was slow to internalize: end users don't evaluate your engineering. They can't. They evaluate the system by the parts they actually interact with — the display on the wall, the app on their phone, and the hot water in their shower.

That's an uncomfortable realization for people who prefer to win arguments with spec sheets. But the perception is the product. If the client's only physical contact with a $60,000 installation is a wall unit with a scratched screen and an app that takes nine seconds to connect, that's their whole impression. Nobody walks out to the array with a clamp meter to check your string sizing.

The client never reads your commissioning report. They look at the box on the wall and decide whether they got their money's worth.

A full battery and cold water is a perception failure

This is where the surplus question matters, and it's the one most quotes skip entirely.

Say you spec a 3.6 kW array into a 48 V bank with a healthy lithium pack. By ten in the morning in June, the bank is at absorption and the array is producing into a system that has nowhere to put the energy. From the client's side, the panels look like they stopped working. They paid for twelve panels and got four hours of visible production.

The fix is planning, not hardware heroics. On a Victron Energy solar system, surplus routing is a documented configuration — typically a programmable relay on the Cerbo GX or a dedicated diversion controller pushing excess PV into a hot water element once the bank is full. On a Victron Energy battery setup, the sequence most clients actually want — battery full, heat hot water — should happen without them ever thinking about it. It's not exotic technology. But it has to be in the design and in the budget, and it usually isn't, because it doesn't show up as a line item on a comparison spreadsheet.

The question at the six-month review won't be "what's your MPPT efficiency curve?" It'll be "my battery is full by ten and my shower is still cold — why did I pay for panels?" That's a perception problem created entirely by a procurement decision.

Where cheap really breaks: EV charging and load math

If you want to see the cost of underspecifying in one place, look at how to set up a EV charging station on an existing commercial service.

Everyone starts with the charger. Almost nobody starts with the service. A 48 A continuous Level 2 load needs a 60 A branch circuit under NEC 625, and whether you can actually land it depends on existing service capacity, available breaker spaces, and whether the transformer has headroom. Half the projects I've been pulled into had a quote for chargers and no load calculation at all.

Load management is where the engineering lives — and where the cheap option quietly becomes expensive. An inverter-based system that can shave peak demand changes the economics of a station, because demand charges are often the biggest line on a commercial utility bill after year one. That's a design decision. It isn't something you bolt on after the slab is poured.

Commercial Level 2 installation typically lands somewhere in the $2,000–$8,000 per port range depending on trenching, panel capacity and whether you're core-drilling a wall (based on contractor quotes I collected across 2023–2024; verify locally, it moves fast). The spread inside that range is almost never the charger. It's the site work.

The 300 watt panel is the wrong line item to squeeze

The 300 watts solar panel is one of the most commoditized pieces of the system — which is exactly why people negotiate it hardest, and exactly why it's the wrong place to fight.

Modules in that class are broadly similar in price. What isn't similar is how they behave in a string. Take twelve panels at 300 W into a 48 V bank: 3.6 kW of array. A Victron MPPT 150/70 is rated for 4,000 W at 48 V per the datasheet, so the controller isn't your constraint. Voltage tolerance and temperature behavior are. Mix two module models with different temperature coefficients on one string and you'll donate harvest every hot afternoon — silently, with no alarm, for years.

What I do now: hold the module model fixed across a string, buy the extra headroom on the controller, and refuse substitutions on the panel spec without a re-check. That's not a purity thing. It's that a substituted module costs me a design review, and design reviews are billable hours I can't recover.

What I'd push back on if I were reading this

To be fair, budget pressure is real. I get why a project engineer picks the cheaper unit when the client has a fixed number and the margin's already thin — I've signed that PO myself more than once. Granted, some of those calls worked out fine. Not every lower-cost component fails.

The numbers said go with the cheaper line in early 2022. Something felt off — the spec sheet was immaculate but their technical support line took three days to answer a simple question. I went with the numbers anyway. That three-day response was a preview of a three-week lead time, which became a six-week one by fall.

Even after we standardized on a single controller line in 2024, I second-guessed it for a full quarter. What if we'd just locked ourselves into a price we couldn't renegotiate? It took two clean quarters of callbacks dropping by 40% before I stopped checking that spreadsheet every Monday morning.

What we actually do now

Three rules, and none of them are clever.

  • Every component comparison carries a callback estimate, not just a unit price. If a vendor can't give me documented field failure rates or a support SLA, the cheaper unit doesn't get the order — it gets a trial, on one site, with an exit clause.
  • Surplus and load management get scoped in the original design. If a system ends up with nowhere to put its own production, that's a design failure, and the client will blame the equipment, not the design.
  • Sequencing gets priced. Racking, roof work, trenching, commissioning — every trade that touches the electrical scope gets a coordination allowance, because that's where the real overruns live.

The legacy myth I keep running into is that quality is a luxury you buy when the budget allows. That thinking comes from an era when modules cost several dollars a watt and a 20% hardware premium was genuine money. Module prices fell by more than 80% between 2011 and 2023 (Source: NREL cost benchmark series, 2023) — the hardware delta shrank while labor and callback costs kept climbing. The math changed. A lot of quoting habits didn't.

So no, the good display and the relay that dumps surplus into the water heater don't show up in your efficiency numbers. They show up in whether the client calls you back next year. In my experience, that's the figure that actually compounds.

Pricing figures are general reference only and vary by vendor, site and time of order; verify current rates before quoting. Code references should be checked against the locally adopted edition of the NEC.

Renata Silva

Renata Silva

Renata Silva is a photovoltaic module analyst covering monocrystalline solar panels, bifacial modules, TOPCon and heterojunction designs, glass-glass construction, junction boxes, and module warranties. She interprets IEC 61215 and IEC 61730 evidence while comparing rated power, conversion efficiency, temperature coefficient, bifaciality, insulation, mechanical-load results, degradation assumptions, and tolerance. Her technical guides help EPC engineers, distributors, and project buyers separate qualification evidence from site-specific energy yield, climate exposure, installation constraints, and long-term performance risk.