The Invoice That Should Have Been $4,200. It Was $6,850.
I still have the paperwork on this one. Q3 2023, we spec'd out a lighting refresh for two office floors. Three vendors, similar fixture lists, quotes ranging from $4,200 to $5,100. We went with the $4,200 bid. Standard play — lowest responsible quote wins, procurement policy signed off.
The final landed cost was $6,850. Not because the vendor played games. Because the quote never captured what the project actually required.
Since then I've tracked every lighting-related invoice across 9 years and roughly $310,000 in cumulative facility spend. I've negotiated with 40+ electrical and lighting suppliers. And I'll tell you what I've learned: the fixture price is almost never the problem. The problem is everything the quote leaves out.
What You Think the Problem Is (The Quoted Price)
When a lighting project goes sideways, the first place everyone looks is the unit price. Did we overpay? Did that acuity-brands product line really justify the premium?
Here's what I can tell you from my own tracking — and I should mention this is based on mid-size commercial builds, not massive industrial plants. If you're in the 200,000+ sq ft category, your experience might differ significantly.
In my spreadsheets, the quoted unit price of fixtures accounted for about 38% of the variance between budget and actual spend. The other 62% came from things that never appeared on the initial estimate. Installation labor on a fixture you thought was plug-and-play. Control integration you didn't know was required. Replacement parts you needed two months later because the quantity shipped didn't match the quantity spec'd.
I remember thinking — back in maybe 2021 — that if we just got better unit pricing, the overruns would dry up. That was wrong. The unit price matters, but it's rarely the culprit.
Why the Real Problem Is Specification Drift
Here's the pattern I kept seeing across our projects, and it took me about three years to name it properly: specification drift.
Someone defines the scope. "We need recessed lighting cans in the open office areas." Sounds clear. Everyone nods.
Three weeks later, the installer says the ceiling cavity won't accommodate the cans. So we switch to flush mount fixtures — which, by the way, are not a 1:1 replacement. Flush mounts throw light differently. Coverage changes. The light area per fixture (how much floor or surface space it actually illuminates) shrinks. Now you need more fixtures. Now you need to redo the photometric layout. Now the installation crew is on hour 14 instead of hour 9.
That "minor" spec change cascaded into $1,600 in additional costs on that specific project. I want to say it was closer to $1,800, but I'd have to pull the actual invoice to be sure.
The assumption is that better unit pricing solves lighting budget problems. The reality is that better specification accuracy solves them.
I've seen this same dynamic with decorative fixtures, too. A previous employer — hotel renovation context, not pure office — had a flower chandelier spec'd for a lobby feature. Gorgeous fixture. Wrong ceiling load rating. The structural reinforcement alone added about 30% to that line item, and no one had priced it in because "the chandelier quote" only covered the chandelier.
What It Actually Costs You
I want to quantify this, because "budget overruns are bad" is not a useful insight. Here's what my records show over roughly 60 lighting-related purchase events:
- Average spec-drift cost: +22% over the approved budget, with a range from 8% (simple fixture swap, same ceiling) to 61% (control system change mid-install).
- Re-work incidents: 1 in 5 projects required some fixture relocation or return. Return freight averaged $380 per event.
- Timeline slippage: When spec drift hit, install time ran 1.4x the estimate on average.
On a $50,000 annual lighting/fixture budget line, that 22% average drift means you're either eating $11,000 in unplanned spend or you're going back to finance mid-cycle asking for more. Neither is fun.
There's a second cost that's harder to put a number on: relationship erosion. When you come back to a vendor three times with revised specs, some of them start padding their next quote. Not maliciously — they just add a "chaos buffer" because they've been burned. When I audited vendor behavior after two messy projects, I found our subsequent quotes from those same suppliers were running 12-15% higher than comparable quotes from vendors we hadn't worked with. Correlation, not causation, but it made me think.
The Fix (Which Is Really a Process Fix)
After getting burned twice the same way, I built a 30-line pre-bid specification checklist. Nothing fancy — just a structured way to force the conversation before quotes go out. It covers ceiling cavity dimensions, load ratings, control compatibility (including dtl dark-to-light photocontrols if the space has daylight-responsive requirements), fixture coverage calculations, and a "replacement parts availability" lookup for anything in a sealed ceiling detail.
I'll be honest: it slows down the front end of a project by about two days. Quotes take slightly longer because vendors have to answer more questions. But across the 11 projects since I implemented it, average spec-drift cost dropped to 6% — down from 22%. That's $8,000 avoided on a typical $50,000 annual line.
One more thing I should add, because this matters for smaller facilities: if you're a 20-person office replacing a handful of fixtures, you may not get much attention from large distributors. I've been on both sides of that. My advice — and this comes from being the small order in the room — is to build the specification accuracy habit early, even on tiny orders. The vendors who take a 4-fixture order seriously, who ask about ceiling height and control voltage and replacement availability — those are the ones I still call for the $20,000 projects. The ones who brush off small buyers? I've watched them lose the loyalty of growing companies. I know which side of that equation I'd rather be on.
It's not about spending more. It's about describing precisely what you need.