Why construction software is the hard one.
Every few years a generation of software founders rediscovers construction and announces it is finally ready to be disrupted. Then the generation gets tired. The buyers are skeptical. The margins are thin. The data is paper. The decision-makers wear boots. The sales cycle is measured in seasons. Whisenant has been reading from this same script publicly for two years, in the way of a founder who has clearly decided to outlast the cycle rather than outrun it.
His thesis, distilled across dozens of LinkedIn posts and a guest appearance on The ConTech Exec podcast, runs like this. Construction has trillions of dollars of throughput and almost no purpose-built software layer between the drawings and the work. Every project manager keeps a version of the same brain in their head - a giant lookup table of trades, line items, vendors, prices, risks, and the dozen things that always go wrong on a build in their market. That brain does not scale. When the project manager goes home sick, the project goes home sick.
Roger is an attempt to put that brain in software without pretending the software is the project manager. Whisenant is careful about this distinction. The AI does not run the job. It catches the things a tired human misses on page 47 of a 200-page drawing set. It writes the RFQ the PM was going to write anyway. It scores the subcontractor list against the kind of project the PM is actually building. It is, by design, an assistant for the most experienced person in the room, not a replacement for them.
That posture matters in an industry that has seen a parade of contech startups arrive with grand claims and depart with quiet acquihires. The Roger pitch is closer to a hammer than a moonshot. The product helps a PM finish a scope faster. The PM finishes more scopes. The contractor wins more bids. The investor sees retention. Nothing about it requires the buyer to believe in a different future.
Whisenant has been writing about this in a voice rare for a founder - skeptical of his own category, suspicious of contech hype, and clear that most of the people building AI for construction have never had to wear a hard hat for a paycheck. The posts read less like content marketing and more like a public memo to himself. They are the best thing on the internet for understanding why he is doing what he is doing.
The Stanford degree gets him in the room. The Zwick year tells the contractor in the room that the software was built by someone who has been on the other side of the table. The investors get a founder whose first instinct, when asked what's hard about the industry, is to say so directly. The customers get a tool that does the unsexy work without trying to convert them.
Salt Lake City is, quietly, part of the story. Roger is not in San Francisco or New York. It is on E Perry Avenue, a few blocks from the Wasatch foothills, in a city where the cost of being patient is lower than the cost of being loud. Utah's tech scene has gotten comfortable with companies that take a decade to look obvious. Whisenant is building one of those.
The tagline on the front of tryroger.com is a single sentence. "What if every project was like your best project?" It is not a slogan. It is the actual product spec.