Contractor Prequalification: Everything the GC Asked For Already Existed

The email lands on a Thursday at 4:40 in the afternoon. The general contractor's risk group needs a document package before your crew mobilizes Monday morning.

Three years of loss runs and EMR letters. The written safety program. OSHA 300 logs for the last three years. Training records for every person on the crew list, including the two you added yesterday when somebody called out. Competent person designations for the scope of work. Signed JHAs from similar work you've done. A completed subcontractor prequalification questionnaire that runs eleven pages and asks four different ways whether you've had a fatality.

The safety manager reads it and doesn't panic, because none of it is unreasonable. All of it exists. Every one of those things actually happened. The training was delivered. The JHAs were signed. The inspections were done and the findings were corrected.

That's what makes the next three days so miserable. The work was done. The proof of it went somewhere else.

Where EMR Letters Actually Live

The EMR letters are in an email from the broker, somewhere in a mailbox with 40,000 messages in it. The 300 logs are fine, because those get built once a year and somebody owns them.

Training is where it starts coming apart. There's a spreadsheet the training coordinator kept, and it was accurate up until she left in March. After that it's cards in wallets, a PDF certificate the OSHA 30 provider emailed to whoever registered the class, and a foreman who will tell you with total confidence that everybody on his crew is current, and who is probably right.

The JHAs are in truck cabs. Some of them made it back to the trailer and into a folder. The ones from the job that finished in April went into a box, and the box went into the storage unit with the rest of that job.

The inspections were done every week on a paper form. Somebody photographed a stack of them once and put the photos in a shared folder, organized by nothing.

None of that is a discipline problem. Nobody was cutting corners. The company runs a real safety program and the field does the work. What the company doesn't have is a way to get the record of that work out of the field and into a place where anyone can find it later.

So the safety manager spends Thursday night, all of Friday, and most of Saturday assembling a document package by hand for a job that hasn't started yet. And the whole time, in the back of his mind, is the awareness that the GC is going to ask again in six weeks for the next phase.

The GC Isn't Being Difficult

Worth saying plainly, because it gets forgotten around the third hour of digging.

The general contractor is carrying your risk. If somebody on your crew gets hurt on their site, it becomes their incident, their OSHA visit, their conversation with the owner, and eventually their number. They're asking for documentation because their insurer asks them, because the owner asks them, and because the alternative is taking your word for it.

The prequalification packet is a proxy. It isn't really asking whether your program is good. It's asking whether you can prove your program is good on three days' notice, because a company that can prove it on demand is usually a company that's actually running it.

That's an uncomfortable thing to sit with, but it's a fair inference. And it means the retrieval problem isn't administrative. It's part of how you get judged.

The Software That Was Supposed to Fix This

Here's the part that surprises people. A lot of the companies living through that Thursday afternoon already bought software to solve it. Sometimes years ago. It's still in the budget.

What happened is almost always the same. The platform got selected by people who evaluated it on a laptop, in an office, on a conference room screen with good wifi. It looked capable, because it was. Then it went to the field.

The foreman opens it on a phone with a cracked screen. He has to log in, and he can't remember whether it's the one with the capital letter. The form has fourteen required fields and three of them are dropdowns with sixty entries. There's a signature box that needs a fingertip, and he's wearing gloves, and it's twenty-nine degrees. The photo has to be uploaded separately through a different screen. And when he finally hits submit, there's a spinner, because the site is in a river valley with no service and the app was built assuming a connection.

He does that twice. Then he goes back to paper.

At that point the software hasn't failed loudly. It's worse than that. It becomes a place where somebody in the office retypes paper forms two weeks late, which produces a database that looks populated and is mostly fiction. The company is paying for a system, the field is still on paper, and the Thursday afternoon problem is exactly where it was.

What Field-Ready Actually Means

If you're evaluating this category right now, the demo will not tell you what you need to know. Every product in this space demos well. They're all built to demo well.

The questions below are the ones worth asking instead. They're not features. They're conditions, and a vendor either meets them or talks around them, which is itself an answer.

Can it be completed with gloves on? Ask them to show you, not tell you. Small tap targets, tight dropdowns, and fingertip signature boxes are how you find out in January that nobody is using it.

How many taps from opening the app to submitting the most common form? Count them yourself during the demo. If the answer for a routine inspection is more than a handful, the field will find a faster way, and the faster way is paper.

What happens with no signal? This is the one that matters most and gets answered with a slide most often. The right answer is that the form completes fully offline, stores locally, and syncs when the device comes back into range, without the user doing anything to make that happen and without any risk of losing what they entered. Make them turn on airplane mode in front of you.

Does a photo attach without leaving the form? Most field documentation is a photo with two sentences attached. If capturing that takes three screens, it doesn't get captured.

Who has to do something for a completed record to become a report? If the answer involves anyone at a desk, you've bought a data entry job rather than a system.

Can the person who filled it out see what happened to it? Field adoption dies when submissions disappear into a void. If a foreman flags a hazard and never learns whether it got fixed, he stops flagging hazards.

Does a person's training record travel with the person? Crews move between jobs and between employers. If qualifications are tied to a project rather than a person, you'll rebuild the same records every time somebody transfers.

And the one that started all of this: when a GC asks, how long does it take to produce a filtered package by project, by person, and by date range? Have them do it live during the demo, using their sample data, while you time it. If it takes them more than a couple of minutes with a system they built and know intimately, consider what it will take you on a Thursday afternoon in year two.

The Part Worth Sitting With

The company in that story is not a bad company. It's a good contractor with a real program and crews who do the work properly. Nobody falsified anything. Nobody skipped a step. The training happened, the JHAs got signed, the hazards got corrected.

They just couldn't prove any of it quickly, and in this industry the ability to prove it has quietly become part of the work itself. Not because a regulator decided so, but because the people who hire you have decided so, and they're the ones writing the contracts.

The gap isn't between companies that run good safety programs and companies that don't. It's between companies whose programs leave a retrievable trail and companies whose programs live in truck cabs, storage units, and the memory of a foreman who is probably right.

Both of those companies are doing the work. Only one of them can show it on a Thursday.