Lanxess Newsroom

We Value-Engineered $3,900 Out of an Exterior Tile Spec. The Repair Bill Was $44,180.

Posted on 2026-11-06 by Rohan Desai

March 2023: the number that looked like a win

I'm the procurement lead at a 41-person commercial tile and flooring contractor. I've owned our materials budget — about $2.1M a year, spread across 60-plus vendors — for seven years. Every quote that comes through my desk gets the same treatment: line-item it, mark up the exclusions, and ask what happens if we're wrong.

In March 2023 we bid a 9,600 sq ft exterior terrace at a 120-room hotel. Framed deck, exterior-glue plywood over joists, freeze-thaw winters, a waterproofing membrane underneath, and a hard deadline tied to the owner's reopening date. The architect's spec called for a two-component polymer-modified adhesive system — the flexible stuff, rated for exterior plywood, with a primer and a membrane.

Our estimator came back with an alternative: a single-component cementitious adhesive from a distributor's private label, no primer. Materials for the full terrace dropped from $18,400 to $14,500.

$3,900 in savings on a bid we were already thin on.

I approved it. Not because I thought it was better. Because I thought it was close enough, and because I had a bid to win and a spreadsheet that said the delta was 21% of the adhesive line.

That's the whole decision. Four minutes at my desk.

June: three things I signed off on without reading

The install started the second week of June. The crew was good — same two guys who've done our exterior work for years. The foreman asked me twice about the substrate moisture reading on the plywood.

I said the GC had it.

The GC said his framer had it.

Nobody had it.

Looking back at the submittal packet, there are three things I initialed that I now read differently:

  • The adhesive classification line. It listed a cementitious adhesive with no deformation grade. For exterior work over plywood in a freeze-thaw climate, I now want a C2 S1 system minimum, and ANSI A118.11 territory for the plywood bond — that's the standard for exterior glue plywood, as I understand it. What we bought was not that.
  • The primer line. Deleted. It was $1,100 on a 9,600 sq ft job. My estimator flagged it as "not required for this product." Possibly true on a concrete slab. Not true on exterior plywood.
  • The movement joint schedule. Referenced "per architect." The architect referenced "per TCNA." The joints went in at 24 feet on center instead of the spacing the detail actually called for.

Three soft omissions on paper. Zero of them showed up as a line item anywhere in the bid.

October: the phone call

The first hard freeze hit on a Tuesday. Wednesday morning the hotel's facilities manager called our project manager and asked why the terrace sounded like a drum.

Hollow. That specific dead sound you get when the adhesive let go and the tile is just sitting there on its own.

We pulled a sample the next day. By the following week we'd mapped the perimeter of the failure: 1,850 sq ft, roughly 19% of the terrace, concentrated along the south edge and around the two roof drains. Not the whole job. Enough of the job.

I'll be honest — I'm still not certain which failure mode led the way here. My best guess is it was cumulative: no primer on plywood that had been rained on once in May, a rigid adhesive under a tile field that moves seasonally, and movement joints spaced too far apart. Three small cuts, none of them fatal on their own. Together? That's a tear-out.

(The membrane, for what it's worth, held fine. It was the one part of the system nobody had tried to value-engineer.)

The tear-out, and where to find epoxy resin at 4 p.m. on a Friday

We reopened the spec and put back what the architect originally called for: a two-component polymer-modified system, a primer, and the correct joint spacing. Short version of what went back in — Kerabond with Keralastic, the polyurethane latex additive that makes it the flexible system, over Mapelastic on the membrane patches. Ultracolor Plus FA for the grout. Mapeflex PU45 at the movement joints. The Mapei exterior tile adhesive system I now default to for anything over framed exterior decking.

Which brings me to the epoxy.

During the demo we found a spalled patch on the concrete landing at one of the drain penetrations — about 14 sq ft that needed a structural patch before the membrane could go back down. The spec called for a 100% solids epoxy. Our regular supplier was out. The second one could get it Wednesday. We were already behind and the GC was already not happy.

I spent 90 minutes on the phone trying to find epoxy resin within a 40-mile radius on a Friday afternoon. I found it. But that's a stupid way to spend a Friday, and it was entirely preventable — the repair scope was known on Tuesday.

I added a line to my vendor file that week: every job with a repair allowance gets two sourced suppliers for tile, epoxy, and sealant before demo starts. Not after. Before.

The paperwork was its own comedy. Our health and safety lead pulled the SDS folder mid-demo and asked why the previous epoxy's data sheet referenced ethylene carbonate as a component — a valid question, since we were trying to get the solvent-heavy stuff out of our warehouse. It took me two days to get a current SDS out of the old supplier (the one that was out of stock, naturally). Under OSHA's HazCom rule, 29 CFR 1910.1200(g), the SDS is supposed to be accessible on site. "In an email from 2021" does not count.

And in the middle of all this, the same supplier's portal search — which is genuinely broken — threw me onto a product page for a healthy mouth cat water additive when I searched for a sealant. Four seconds of my life I'm not getting back. Also the only time that whole week I laughed.

The math, because the math is the point

The final repair invoice, all in:

  • Tile replacement, 1,850 sq ft, matching dye lot, three-week lead time: $7,900
  • Removal, disposal, membrane repair: $6,300
  • Adhesive, primer, grout, sealant for the redo: $3,150
  • Labor, two crews, nine working days: $14,300
  • Hotel credits and the GC's back-charge: $12,530

Total: $44,180. Against $3,900 of savings. Call it an 11-to-1 ratio and stop pretending it was bad luck.

The line that bothers me most isn't the labor. It's the $1,100 primer. We skipped a $1,100 line item and ate a $44,180 repair — and honestly, if you'd shown me those two numbers side by side in March, I'd have said yes to the primer in about a second and a half.

That's the thing about value engineering that I didn't fully get until this job: the cost of prevention shows up in the bid, and the cost of correction shows up in the P&L. One of those two numbers gets reviewed by four people before it's approved. The other one just arrives.

When I finally put the March bid and the October invoice next to each other on my desk, I understood why my old PM used to say the expensive part of a job is never the materials.

What I changed — the 11-point exterior check

I built this checklist after the third rework of my career. It's the cheapest insurance we carry, and it takes about 25 minutes on site. The short version:

  1. Substrate moisture measured, logged, and signed by whoever measured it — not "the framer has it."
  2. Deflection confirmed against the framer's actual numbers, not the drawing's.
  3. Adhesive classification matched to service conditions, in writing, on the submittal (C2 S1 minimum for exterior over frame; A118.11 territory for exterior glue plywood).
  4. Primer on the material ticket, or it didn't happen.
  5. Movement joints located and spaced on the shop drawing before the first tile is set.
  6. Two sourced suppliers for every item with a lead time or a repair allowance.
  7. SDS folder current, printed, and on site before the first box is opened.

There are four more, but these seven are the ones that would have saved this job.

We've run it on every exterior job since. Rough rule of thumb from our own numbers: 20 to 30 minutes of checking per job costs us maybe $180 in labor and PM time. One avoided tear-out pays for roughly 245 of those checks.

That said, I should note we've only been running it on exterior work so far. Interior commercial is a different risk profile and I haven't built the equivalent list yet, so I'm not going to pretend I have it.

What I'd tell the version of me from March 2023

The $3,900 wasn't savings. It was a loan, and the interest rate was somewhere north of 1,000%.

If you're the person who approves the material substitutions on your projects — and if you're reading this, you probably are — do the one thing I didn't. Before you sign the value-engineering sheet, ask the estimator one question out loud: what does this cost if it fails?

If nobody can answer that in under a minute, you don't have a savings number. You have a guess.

Five minutes of verification beats five days of correction. I know that now because I paid $44,180 to learn it, and I'd rather you didn't.

Need the related report or SDS packet?

Request the current Lanxess report, safety data sheet, or technical data sheet package connected to this topic.