What You Need to Know: Precon World 2026 centered on one big problem - demand for preconstruction is skyrocketing, driven by data center work and rising bid volumes, while the pool of experienced estimators shrinks as a generation nears retirement. Speakers agreed AI can help close that gap, but only if it is built on trustworthy data.
Preconstruction has never had more work on its plate. Data center projects are surging, bid volumes keep climbing, and the pool of experienced estimators is shrinking as a generation heads toward retirement.
That tension ran through nearly every session at Precon World 2026. Whether the topic was AI, hiring, or career paths, the conversation kept circling back to one question: how do preconstruction teams take on more work without burning out the people doing it?
Here are three of the themes we heard the most, as well as some thoughtful feedback to take into consideration as we continue exploring these topics and what they mean for the industry as a whole.
1. AI you can trust starts with data you can see
The loudest message about AI was about whether estimators can trust AI.
Several speakers drew a sharp line between black-box tools and transparent ones. If a tool hands you a number and you can't see how it got there, you have no idea where to start adjusting it. Estimators need predictions built on concrete, traceable data, so they can apply their own judgment on top.
That transparency depends on the data underneath. Teams are investing in:
- Attributes that give each project context, from a handful of fields to several hundred, including formulas and ratios
- Controlled similarity, where estimators choose a baseline project and decide which factors (date, size, type) define a comparable job
- Curated subsets, such as a standing collection of healthcare projects for teams that bid mostly in that sector
- Outlier flags with adjustable thresholds, so anomalies surface before they skew an estimate
The takeaway: automation amplifies whatever you feed it. Good data makes it a multiplier. Bad data makes it a liability.
2. Preconstruction is a career, not an accident
A panel on professionalizing preconstruction opened with the observation that many construction students don't even know what preconstruction is. Those who do often think the job ends once the takeoff is done and the numbers are in a spreadsheet.
In reality, that's only where the job begins. Panelists described preconstruction as planning, risk management, and client communication. It's where estimators build trust with owners and get the most exposure to design teams and leadership. Communicating a number well takes emotional intelligence, not just technical skill.
The panel's ideas for closing the gap:
- Earlier exposure. Too many people find preconstruction by accident, or because field work wasn't a fit. First contact shouldn't be bid day.
- Elective tracks, not new degrees. Academia moves slowly, so a preconstruction emphasis, certificates, or graduate programs may arrive faster than a full degree.
- Soft skills over software. Knowing which questions to is often more important than knowing which buttons to click.
- Look inside first. The talent often already exists in the company. It just needs to be in the right seat. Field experience, in particular, makes for stronger senior estimators.
- Give credit where it's due. Precon may have the biggest influence on profitability of any department. Estimators hear about every job that loses money and rarely about the ones that make it. That has to change if companies want to keep them.
3. Agentic AI as a capacity multiplier
Demand on preconstruction has compounded, but headcount hasn't kept up. The gap between construction volume and the skilled workforce widens every year.
One session framed agentic AI as the answer, and drew a useful distinction. An assistant makes a task faster - an agent takes the task off your desk. Instead of asking AI to help with one estimate, a team might ask it to monitor every active project for risk, assess the impact when something changes, alert the right person, and track the issue until it's resolved. A human still makes the call.
Speakers were also clear that you can't skip steps. Autonomy is built from the bottom up: standardized workflows first, then connected project data, then organizational knowledge, and only then agents. They warned against common shortcuts, including DIY prompting by individuals, a single connector with nothing behind it, yet another point solution, and pilots run on tidy sample data that looks nothing like real projects.
The most practical near-term uses mentioned were updating unit prices, maintaining cost libraries, and building scope sheets.
Additional thoughts and considerations
The best moments came when estimators in the audience challenged the optimism on stage.
Not everyone wants automation. Estimators who learned the trade by hand often want to see every number going in. Tools that force automation on them will lose their trust. The vendors getting this right build in flexibility, so veterans and digital natives can work their own way.
Review is still critical. One argument held that AI finally gives teams the bandwidth to review estimates properly, something they rarely had time for before. An estimator countered that unreviewed AI output is obvious and embarrassing to clients. Both points can be true: AI creates time for review, but only if teams actually spend it there.
Agents can feel isolating. One estimator said agentic tools made their work feel more siloed. That’s an important consideration in a field that has become far more collaborative, with estimators, operations, business development, and designers planning bids together. If AI pulls people apart instead of giving them a shared view of the project, it's solving the wrong problem.
What it means for preconstruction leaders
The three themes point to the same place. Preconstruction teams can't hire their way out of the capacity crunch, and AI won't solve that problem on its own. The teams that pull ahead will do three things at once:
- Get their data in order before layering on automation.
- Choose tools that show their work, so estimators stay in control of the number.
- Invest in their people, from how they recruit and train to how they recognize precon's impact on the bottom line.
Learn More About DESTINI Estimator →
