Shiva Dhawan is the CEO and cofounder of Attentive.ai, the company behind Beam AI.
The industry’s explanation for project abandonment has always been financing costs, economic uncertainty, labor shortages and material volatility. While those are all true things, if you listen closely to how the industry describes its own problem, a gap appears.
A 2026 Hiring and Business Outlook report from the Associated General Contractors of America (AGC) found that 63% of surveyed contractors reported that an owner postponed, scaled back or canceled a project in the previous six months. Asked why, 37% cited funding uncertainty, and 34% said financing was unavailable or too expensive.
Nearly all of the blame lands on money. The same survey found that contractors were seeing significantly fewer opportunities to bid than a year earlier, a detail that received far less attention than the financing numbers.
However, money alone can’t carry the explanation. Bloomberg reported in April 2026 that roughly half of the U.S. data centers planned for 2026 were likely to be delayed or canceled amid an AI infrastructure boom, with capital lined up behind it. When projects with that much money and conviction behind them still stall, something else is failing.
Owners are reassessing feasibility in a tougher market, but some of these projects may never have gotten a truly competitive bid in the first place.
The market can’t find a price it never produced.
Before an owner can decide whether a project pencils, the market has to produce a reliable price. That price comes from a preconstruction process that relies on estimators, trade partners and suppliers to turn incomplete information into a number a contractor can stand behind, all under a deadline.
That process is running out of people. AGC’s 2026 outlook found 80% of surveyed firms reported difficulty filling salaried openings, the highest share in three years, and estimating desks sit squarely in that category.
I’ve talked to more than 200 construction owners about this over the past few months. Ask them directly, and almost none will say their preconstruction process is broken. Then, they’ll start describing what actually happens: the “inevitable” last-minute scramble before a deadline, the bid they wanted to submit but couldn’t get to in time, the number that got padded because nobody had time to run it down properly.
Rushed bids make viable projects look unviable.
When qualified firms lack the preconstruction capacity to pursue a job, they no-bid. When they do bid without enough time to validate scope or vet supplier pricing, the risk goes somewhere, and it usually goes into the number.
That padding is rational. Construction input prices were up nearly 10% year over year as of May 2026, according to Associated Builders and Contractors’ analysis of federal producer price data. Miss an escalation like that or a gap in the drawings, and the contractor inherits the loss.
Contractors are now pricing conservatively to protect themselves, and the owner is absorbing the cost of that caution: a number that looks too high, or two numbers instead of five, or figures so far apart that nobody trusts any of them. At that point, the project stops looking viable, whether or not it actually is.
More bidders can mean less signal.
Owners rely on competitive tension to surface a project’s market price, but that tension is spreading thin. In the same AGC survey, contractors reported significantly fewer opportunities to bid than a year earlier, which means every project that does go to market draws a bigger crowd, including larger upstream companies and out-of-market entrants chasing work.
That looks like more competition. However, a longer bid list still draws from the same finite pool of estimating capacity, now spread across more pursuits with less time on each. The system produces more bids without producing more accurate ones, and the owner ends up selecting for bandwidth as much as for building ability.
Other industries show where this leads. A 2025 Imagine360 survey of insured Americans found that 38% of respondents had skipped or postponed necessary care or medications because of cost, up from the 27% reported in a separate 2023 study, and 42% of those who put off care said their condition worsened as a result.
These patients needed treatment and carried insurance to pay for it. What failed was the layer in between, a billing system so opaque that they couldn’t get a reliable number before committing, so they walked away rather than risk an unknown one.
Construction’s opacity is a capacity failure. When the price-discovery layer breaks down, buyers and sellers stop finding each other, even when both want to proceed with the transaction.
Fix the bid process before blaming the market.
The industry notices abandonment when a project is formally paused or pulled from a pipeline. The failure usually starts earlier, when a contractor declines to bid because the team is maxed out or when an owner gets one padded number and can’t tell whether the project is genuinely infeasible or just underpriced. Financing costs may be the headline reason projects die, but pricing uncertainty is one of the mechanisms that makes a project look unfinanceable in the first place.
Nobody in this industry can control interest rates, but everyone in it can control how projects get priced. Owners can prequalify fewer firms and give them more runway because three bids with three weeks behind them carry more signal than 12 bids assembled in a scramble. Contractors can track no-bids and rushed-bid win rates as leading indicators since both surface capacity problems months before they show up as lost revenue. Both sides can treat estimating capacity as a strategic investment rather than overhead.
When the bid process works, owners get numbers they can trust, contractors win work they can actually deliver, and viable projects make it to the jobsite instead of dying in a spreadsheet.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?