I basically believe the most expensive square foot in architecture is not the one with marble. It's the one you have to tear out and rebuild because somebody approved a spec without asking the next question. Prevention beats cure. This is not a slogan. It is the only thing that has kept me employed.
Prevention is not a phase. It's the only phase that matters.
I've spent eight years handling construction documentation and procurement packages for large interior projects and adaptive-reuse conversions. I've personally made—and documented—23 significant mistakes, totaling roughly $180,000 in wasted budget. Now I maintain the team checklist so the next person doesn't have to put rework photos in their portfolio.
The checklist is not a form. It's the closest thing to a time machine we have.
What renderings don't show you
From the outside, a Gensler office to residential conversion looks like a rendering exercise. You see floor plans, furniture mockups, and beautiful daylight studies. The reality is that the finished, marketable interior is only as strong as the unglamorous documents behind it. And the unglamorous documents are where the expensive mistakes live.
In my first year, 2017, I made the classic spec error: I wrote 'milk glass' in a door schedule and assumed the contractor knew exactly what I meant. The contractor installed 'frosted acrylic.' It was not milk glass. It was seven panels, $4,800, and a one-week delay. The renderings looked the same. The building did not.
I still remember thinking, 'But the elevation looked right.' That is exactly the problem. Visual approval is a trap. That's it.
People assume that if a drawing is clean, the details are correct. What they don't see is the 300-page package of door schedules, valve schedules, and finish boards where one ambiguous line can undo all that polish.
The conversion context makes it worse
Gensler's adaptive reuse work, especially the office to residential conversion practice, has pushed the industry to think differently about existing buildings. That's important. But conversions are not just moving walls and adding kitchens. They are systems going through a personality change.
In September 2022, on a conversion project, I wrote 'check valve' in a hydronic schedule. No spring type. No orientation. Nothing. It looked fine on my screen. The plumber installed a swing check valve, which was the wrong choice for that vertical line. The result came back as a pressure surge on the floor above. $890 and a three-day delay.
I don't have hard data on how many MEP schedules are missing check-valve details. But based on three years of reviewing submittals, my sense is that roughly 10% to 15% of mechanical schedules have at least one vague valve specification. That's an anecdote, not a study. It's also a red flag.
Honestly, the hardest part of conversion work is not the big idea. It's the small, boring things that become catastrophic at occupancy.
The checklist is the cheapest insurance I know
We didn't have a formal pre-submittal review process before 2023. Cost us. The third time a missing note caused a reorder, I finally created a verification checklist. Should have done it after the first. A checklist is a no-brainer.
Here's the version I actually use now, not a textbook version:
- Product listed exactly as the vendor names it. No 'or equal' unless someone proves the equal. If you want milk glass, write 'milk glass, 1/4-inch, back-painted.'
- Code and standard included. For a check valve, that means the applicable standard plus a spring type.
- Installation context written on the document. Not on page 214 of the narrative. On the document.
- Non-design constraints listed. Pets, cleaning protocols, maintenance access. If a floor material is supposed to survive life with pets, say it.
- One responsible person signs off. Group approval equals no approval.
Product, code, context. In that order.
In the past 18 months, that list has caught 47 potential errors. Some were small. A few would have been brutal. The estimated savings from avoiding rework is about $8,000. That's not a random boost. It's the difference between staying on site and waiting for someone to reorder glass.
Five minutes of verification beats five days of correction. Simple.
Yes, I'm saying this to Gensler teams too
You might be thinking: 'Gensler is one of the largest architecture and design firms in the world. They have reviews. They don't need a random project coordinator's checklist.'
Fine. But I have coordinated with Gensler teams on interiors, and here's something I learned: a sophisticated drawing set does not make an ambiguous spec less ambiguous.
Gensler interior design is often what makes a converted building feel like a home rather than a cubicle farm. That takes skill. But the skill only matters if the people downstream can build the details. And the people downstream are reading the same short, imperfect sentences we all write when we're tired.
Here's something vendors won't tell you: 'standard' in a spec is a negotiation. If one line says 'milk glass' and the sample board says 'frosted acrylic,' the contractor has to choose. They will choose whichever is cheaper. Then we call it a design decision. No.
So yes, this includes Gensler. It includes every firm I've worked with. Big teams mean lots of handoffs. Handoffs are where details die.
The weird one: how to get rid of fleas in house
This is the part of my checklist that makes people laugh until a property manager calls you in a panic.
A while ago, a property manager asked me if I knew 'how to get rid of fleas in house' in one of our converted units. We had specified a carpet upgrade that looked great and matched the interior design board. It also collected flea eggs. The building had a pet-friendly policy. The interior design drawings did not say 'no carpet in pet-accessible units.' That's not a Gensler problem. That's an 'someone forgot to connect the checklist to the real-world conditions' problem.
The fix was not pretty: pest control, pet relocation, and a new flooring note in future units. The lesson? Interior design details are never just details. They are operational choices. If the unit will contain a dog, the spec should know.
I wish I had tracked every one of these near-misses from the start. I didn't. What I can say anecdotally is that these are avoidable. You don't need more talent. You need a process that catches the weird cases before they become tenants' problems.
Prevention was never about being perfect
Somebody will always tell you that you can't eliminate rework. True. But you can eliminate the predictable rework. That's the goal.
According to the Construction Industry Institute (construction-institute.org), changes made during construction can cost an order of magnitude more than the same change made during preconstruction. I believe it. I've lived it. A redline on a drawing is cheap. A redline on a delivered corridor is not.
So here is my position: the firms that treat checklists as training wheels will keep paying for their elegance. The teams that treat them like a design document will get to open their buildings twice—once for inspection, and once for occupancy.
Bottom line: prevention is cheaper than cure. It's not the glamorous part of architecture. It's the part that lets everyone else be glamorous.
