The do’s and don’ts of running a tech pilot

Listen to the article
6 min

This audio is auto-generated. Please let us know if you have feedback.

Construction has a reputation of being a Luddite industry. Extra costs and an aversion to risk-taking can make builders hesitant to try new technology.

But tales of success can be found, if one knows where to look.

Take, for example, Providence, Rhode Island-based Gilbane Building Co., which used New York City-based Trunk Tools to track 21,000 documents on a large stadium project in Wisconsin. The software saved Gilbane’s project team 20 to 40 minutes in travel and in searching for answers to each question, and more than $100,000 in avoided rework.

The firm thought the results were so successful, it went on to sign an enterprise agreement with Trunk Tools to roll out the tech nationally. 

Meanwhile, Edmonton, Alberta-based PCL, whose U.S. headquarters are in Denver, trialed Broomfield, Colorado-based Vita Industrial’s Load Navigator, which helped the builder work through high-wind days while hauling giant pieces of form work through the air. The team saw a 30% increase in efficiency, and one project manager said it could potentially save anywhere from thousands to tens of thousands of dollars. 

And LeChase Construction Services, headquartered in Rochester, New York, adopted Oakland, California-based Join’s preconstruction tech planning tools to help streamline its work and meet a key funding deadline on a $13.5 million school expansion in New York. The outcome? The team maintained constant forward momentum during the preconstruction process, and the school is slated for completion in spring 2025. 

Key to all three of these success stories is the deliberate measures that these contractors took to test these technologies before committing wholesale. It’s not simple — trialing new technologies and services on a job or a jobsite can be a stressful and complicated task. On top of the pressure to make the investment work, there may be a compounding drive to make a profit or see immediate results.

Below, construction technology experts provide their list of “do’s” and “don’ts” to help builders make sense of how best to pilot their next technology.

DO: Pilot for the right reasons

When piloting a new piece of technology, make sure you’re doing it for the right reasons. A strong correlation between what the business needs and the goals of the pilot committee are important, said Alex Belkofer, senior director of VDC for St. Louis-based McCarthy Building Cos.

“I’ve seen a lot of folks go down the route of doing a pilot with a small group of people, and they come out with an end result, but if it doesn’t meet the needs of the business, sometimes those pilots dissipate,” Belkofer said.

McCarthy has stakeholders that come together to make these pilots happen — its innovation team, emerging technology team and VDC team, Belkofer said. That group, which he calls a community of practice, loops in the teams on the jobsite, such as field operations and preconstruction, who implement the pilot.

DON’T: Skimp on investment

Companies that talk the talk, but don’t walk the walk, on innovation can potentially set their teams up for failure. It’s up to builders to make sure the teams in charge of innovation can get what they need to make their pilots work, said Tim Gaylord, corporate director of innovation for Redwood City, California-based DPR Construction.

While Gaylord acknowledged that margins in construction are thin, he noted how vital it is that innovation teams be able to work on these types of technology.

“If you don’t have dedicated budgets that can help people try new things, and folks to help document and set these up, it’s really hard to innovate when you’re on a project,” Gaylord said.

DO: Know exactly what your tech does

Builders need to know what, exactly, they’re hoping to get out of it, said Jit Kee Chin, chief technology officer for Boston-based Suffolk Construction.

Chin used the example of Microsoft’s Copilot technology, the tech giant’s AI assistant that integrates with Microsoft 365. She said it operates best when given a specific set of parameters in the questions users ask. If someone types out a random question and receives a hit-or-miss answer, Chin says, then that user isn’t using the technology to its full potential.

“Understanding that and then having a way to know what it takes to be successful, to use and then actually get through that adoption barrier is super important,” Chin said.

DON’T: Fear failure.

Experiments are experiments. Don’t be afraid to fail fast, Gaylord said.

“I think there’s a lot that can be learned from a tool that doesn’t work out, but you have to have a structured way of capturing that and sharing that, which is tough. Let your people try new things,” Gaylord said.

Chin noted that construction is a high-hazard industry. But that can’t stop innovators.

“Take a risk. The industry needs to change,” Chin said.

DO: Learn from your peers

The world of construction is big — connecting with technologically advanced peers is a great way to understand these processes better.

Belkofer recommended going to conferences, speaking with stakeholders and user groups and sharing notes with other innovators to get a finger on the pulse of the industry and what it’s doing with technology.

“Reach out. There’s a network of folks out there that are probably doing and chasing a lot of the same things that you are,” Belkofer said.

Learning from peers doesn’t necessarily mean other companies, however — it could even be in-house.

And again, don’t look at failures as, well, failures. There’s a lot a builder can glean from a tech that doesn’t pan out the way it’s expected to, Gaylord said. For example, learning why it failed allows an innovation team to pass the lessons on, so the next person doesn’t make those mistakes.

“It’s kind of like investing in the stock market, right? They’re not all going to be winners, but if you can learn from them and share those lessons learned so folks aren’t duplicating those efforts and wasting time, that’s huge,” Gaylord said.