There is a version of AI adoption that feels productive but produces almost nothing. A firm buys a tool, or starts using one of the widely available platforms, and people begin plugging tasks into it. Drafting emails. Summarizing documents. Generating boilerplate language. The tool works. The outputs are fine. And six months later, the firm is roughly in the same operational position it was before, just with a few people who type prompts instead of doing certain tasks manually.
This is not a technology failure. It is a sequencing failure. And it is far more common than most firms want to admit.
The core issue is straightforward. AI amplifies what already exists. When a firm has clear workflows, consistent data, and well-defined processes, AI can accelerate those things meaningfully. When a firm has fragmented workflows, inconsistent documentation, and processes that live in people’s heads rather than in any structured form, AI accelerates the fragmentation. It does not resolve it. It cannot resolve it. That is not what the technology does.
For contracting firms, this tension shows up in some very specific places.
Estimating is one of them. The firms that get real value from AI in estimating are generally the ones that already have some discipline around how estimates are built, where historical data lives, and how scope gets defined before the numbers start. Firms that have always relied on one senior estimator’s instincts and institutional memory, with no real structure underneath it, find that AI tools give them faster outputs that are built on the same shaky foundation. Speed without structure is not an improvement.
Project documentation is another. AI can help with writing, organizing, and summarizing documentation. But if the underlying problem is that documentation doesn’t get captured consistently in the field, or that the information coming in is incomplete and unreliable, AI produces better-looking documents that still don’t reflect what actually happened on the job. The output quality is a function of the input quality. That relationship does not change because the tool is sophisticated.
Business development runs into the same wall. AI can help a firm produce more content, write better proposals, and communicate more clearly. But if the firm has never defined its positioning, never captured its project history in a usable form, and has no real strategy behind its BD activity, AI helps them do more of the wrong things more efficiently.
None of this is an argument against using AI. It is an argument for being honest about what AI actually needs in order to deliver value.
The firms getting the most out of AI right now are not necessarily the ones with the most advanced tools or the highest budgets. They are the ones that did some foundational work first. They clarified their processes enough that AI had something coherent to plug into. They defined where human judgment still needed to lead and where AI could handle the load. They thought about governance, which is just a formal word for knowing who is responsible for what and how outputs get reviewed before they go anywhere that matters.
That foundational work is not glamorous. It does not have the appeal of demoing a new tool or watching AI generate something impressive in thirty seconds. But it is the difference between AI that produces a moment of interest and AI that produces a measurable business result.
The good news for contracting firms is that the bar here is not as high as it sounds. Most firms do not need a complete operational overhaul before AI becomes useful. They need enough clarity in a specific workflow to give AI something real to work with. Start there. Pick one area, define the process around it with enough specificity that the inputs are reliable, and then build AI into that workflow intentionally. That is how individual momentum becomes institutional capability.
The firms waiting for a tool to solve a structural problem will keep waiting. The firms willing to do the structural work first will find that the tools they already have access to start producing something worth talking about.