The process of rebidding on trade categories — and evaluating proposals — has long been a blend of gut instinct, trusted relationships, arbitrary timelines, and laborious manual data gathering. For builders, this traditional approach has limited their ability to make smart, real-time decisions, forcing them to spend excessive time on tedious, imperfect, and often poorly timed pricing analysis rather than focusing on building homes and nurturing supplier partnerships.
The pitfalls of rebidding
While the process has evolved for many builders, some standard practices remain common in rebidding material or trade contracts. Each of these tactics opens the door to missing potential savings — or even inadvertently costing the builder more in the long run.
Relying on instincts and relationships. “This is a people industry,” says Heather Hall, managing director of HomebuilderONE (a product of sa.global, a Microsoft cloud partner serving the production homebuilding sector). “Lived experience and trusted relationships are important and valuable.” However, she cautions against overconfidence in the information obtained. Rebidding based on a gut feeling is unreliable, especially in a time of rampant volatility in supply chains and pricing — as is placing too much value on peer word-of-mouth or the advice of even trusted sales professionals. “You might be sure you have the best price because you asked three friends that all sell this material, but fact check what you hear,” Hall says.
Rebidding on a schedule. It’s common for builders to rebid annually or every two years, regardless of whether the market has fluctuated up or down, but, Hall says, “If you’re doing it on a fixed schedule, you may not be rebidding a material at the right time.” Three potential problems could arise:
- Missing chances to rebid when there’s potential for cost savings
- Wasting time on research and paperwork with minimal impact
- Incurring more expense by rebidding at a time when costs are higher
“You could get lucky with your rebid — but if you harness facts and data, you don’t need to get lucky.” Instead, Hall advises, “Rebid when the data says there might be an opportunity, rather than on an arbitrary cadence.”
Manual data gathering and reporting. Accurate material cost intelligence relies on pulling data from multiple sources, collating to create benchmarks, and analyzing how much (if any) could be saved with a rebid. Doing it manually can take weeks: researching and gathering data, compiling it into spreadsheets, and calculating various scenarios. The right AI-enhanced systems can crunch these vast datasets and generate reports in seconds versus weeks.
Using inadequate data solutions. Many existing digital tools builders use to analyze pricing only allow users to check one material category at a time. “You can’t run a purchasing team at scale that way,” Hall says. “The likelihood of hitting on a money-saving insight at the right time is quite low.” Builders may miss both savings opportunities and margin risks. With HomebuilderONE, she adds, “We stack existing technologies with this newfound ability for AI to synthesize data and streamline what a lot of businesses and humans are currently doing.”
The role and requirements of effective AI assistance
With AI permeating nearly every online function, it’s making headlines not only about its achievements but its potential drawbacks. AI tools are not one-size-fits-all, and not every solution will solve rebidding challenges. The right AI tool to truly assist in pricing intel needs to be:
- Grounded in authoritative, domain-specific data. Unlike general large language models like ChatGPT or Copilot — which draw from broad, public datasets — the ideal AI agent relies on carefully selected, high-quality data sources. “We aggregate material cost benchmarks and actual spend data,” explains Hall, “avoiding the general data lakes that cause hallucinations.”
- Compliant with company data security protocols. One thing AI models struggle with for businesses is security. “It’s an inhibitor for many organizations to implement AI, especially if you’re publicly traded; you can’t risk exposing your entire corporate data set to everyone.” HomebuilderONE’s solution incorporates a security layer allowing users to access AI insights only within their authorized areas, protecting confidential areas and systems. “Many companies have been trying to crack that code, and we’re proud to be one of the few that’s achieved it.”
- Based on mature, established technologies. Effective AI tools build from existing tools rather than trying to reinvent the wheel. “We’ve spent years defining how we can leverage decades of homebuilding technology experience and proven legacy systems,” says Moacyr Galo, chief AI officer at sa.global. “Now we’re adding an agentic AI layer that removes technical complexities and delivers real business value based on the high expectations builders have for AI.”
- Compatible with a range of technologies. Because builders typically use multiple software products, AI should be able to work with a range of technologies, not just a single proprietary solution. “Our solution connects to everything on HomebuilderONE, but it can work with a combination of legacy systems,” Galo says.
- Proactive as well as on-demand. Since one of the main pitfalls of traditional pricing intel is rebidding on fixed schedules or instinct, AI offers the most value when it can proactively surface opportunities to rebid based on market insights. “Our AI pushes alerts to you as conditions change day to day,” Hall says.
- Designed as an assistant with human oversight. “AI helps us make better, more informed decisions, it’s not a human replacement,” Hall says. Homebuilding requires nuanced decision-making by experienced pros. AI augments rather than replaces human intelligence, freeing teams up to make better use of their expertise. It simply makes it easier for users to interrogate data, present options, make decisions and execute.
What AI implementation looks like for builders
Depending on the need, Hall says, AI solutions can be rolled out quickly, in a couple weeks not months and years. For typical pricing intelligence needs, builders can have AI-powered dashboards operational within two weeks. More customized AI solutions — such as new predictive models tailored to unique builder challenges or legacy business applications — may require one to two months to develop.
Adoption can be fast, with minimal training required. The user experience is designed to be intuitive, consisting of web-based interfaces where people ask questions and receive quick, thorough answers.
Where AI is heading next
Looking beyond pricing, sa.global is exploring AI’s potential to optimize option take rates in customizable homes. Builders who offer extensive product choices face the challenge of managing the costs and availability of these options while maximizing profitability. AI agents can analyze which options buyers frequently select, identifying low-margin items that could be repriced and low-uptake options to consider eliminating. By guiding sales teams to promote the right options, builders can streamline inventory, reduce overhead, and increase margins.
Like the pricing intel solution, these AI agents will operate proactively, recommending options based on the buyer’s initial selections. “If you can take a four- to six-hour design meeting down to two and the customer’s happier with what they picked, you’ve just driven customer experience and customer satisfaction to a new level,” Hall says.
The fast track to smarter rebids
For builders ready to leave outdated processes behind, AI promises to free up time and unlock cost savings. Moving from guesswork and instinct to proactive, data-backed insights can reshape the bidding process to get you closer to your goals.
Interested in learning how AI-driven pricing intelligence can transform your building process? Visit our HomebuilderONE toolkit to learn more.
