Construction technology ROI: your questions answered
Are you prepared for what’s next in AECO?
Not many CTOs can say they've spent nearly three decades at the same firm. Atul Khanzode has, and that longevity gives his perspective on construction technology a credibility that's hard to find elsewhere. As CTO at DPR Construction, the 7th largest general contractor in the US, he's responsible for all technology functions across mission critical, life sciences, healthcare, and advanced technology sectors.
In a recent episode of Bridging the Gap: The Conversation — CIO under pressure, he shared how DPR approaches technology investment, rationalization, and AI readiness.
How do you balance expanding your tech stack with controlling licensing costs?
Short answer:
Technology cost growth is normal in a growing firm — what matters is whether value is growing at the same rate. On complex, high-value projects, spend doesn't scale linearly with revenue. Rising costs should always be read alongside margin and outcome data, not in isolation.
Long answer:
Nearly two-thirds of CIOs surveyed in the 2026 Bridging the Gap report reported an increase in software and cloud licensing costs over the past 12 months. At the same time, 41% plan to expand their tech stacks while 39% plan to consolidate. It's a paradox that most technology leaders are living with right now.
Atul's view is that cost growth in technology isn't inherently a problem — it depends on what's driving it.
"A lot of our cost of technology is tied to how our revenue growth has happened. I'm not surprised that it has increased for many corporations, just because of how busy the industry has been."
Atul Khanzode, CTO, DPR Construction
At DPR, the data center construction boom has driven significant revenue growth — the mission critical market has grown from approximately 30% to close to 50% of DPR's revenue in recent years. That growth brings technology cost with it, but the relationship isn't linear. Larger projects require fewer people to deliver than an equivalent volume of smaller projects, which means technology spend doesn't scale at the same rate as revenue.
The practical implication is that technology cost growth needs to be read in context. Rising licensing costs alongside rising revenue and stable or improving margins is a different story to rising costs with flat revenue and declining margins. The question isn't whether costs are going up. It's whether the value is going up at a proportionate rate.
Want to hear it in Atul's own words? Watch the full conversation here.
How do you prove the ROI of construction software?
Short answer:
ROI should be measured against business strategy, not generic metrics. The four buckets (revenue and margin, cost reduction, employee experience, customer experience) provide a consistent framework for every technology investment. Tools that don't connect to the firm's core strategy rarely deliver meaningful ROI regardless of their technical capability.
Long answer:
This is the question that sits behind almost every technology investment decision — and the one that's hardest to answer with confidence when the benefits are diffuse, long-term, or difficult to attribute to a single tool.
Atul's approach at DPR is to evaluate software not against a generic ROI calculation but against how well it advances the firm's core strategy.
"We rationalize our software spend based on how that advances our strategy. Our strategy has been that we should deliver predictable outcomes for our customers through the use of virtual design and construction."
Atul Khanzode, CTO, DPR Construction
Predictable outcomes through VDC, fabrication-level design, prefabrication, and self-performed work. That's the filter through which every technology investment at DPR gets evaluated. If a tool advances that strategy and the results are visible, it earns its place. If it doesn't connect to the strategy, its ROI is irrelevant regardless of what the vendor claims
In practice, this means asking a specific set of questions before any technology investment is made. Atul frames these as four buckets that every tool should be evaluated against:
- Does it improve revenue and margin?
- Does it reduce the cost of doing business?
- Does it improve the employee experience?
- Does it improve the customer experience?
If a tool can move one or more of those dials with evidence, it earns its place. If it can't, the sophistication of the technology is irrelevant.
"Instead of getting enamored by just the technology tools, really having a dialogue internally about what is it that you're trying to accomplish — that will guide your further technology exploration decisions."
Atul Khanzode, CTO, DPR Construction
What is the criteria for retiring a construction technology tool?
Short answer:
Adoption is the most honest signal of whether a tool is delivering value — unused tools retire themselves. A structured innovation team prevents stacks from accumulating by default. Apply the four buckets to retirement decisions as well as acquisition decisions. Tools that can't move any of the four dials are candidates for removal.
Long answer:
Knowing when to stop investing in a tool is as important as knowing when to start. But retirement decisions are rarely straightforward. There are sunk costs, vocal internal advocates, and the inertia of established workflows to contend with.
Atul's approach is pragmatic and deliberately unsentimental.
"We do have a mechanism where we test a bunch of things. Sometimes things work, sometimes they don't. Usually the process sort of takes care of itself, because if it doesn't work, nobody's gonna use it anyway."
Atul Khanzode, CTO, DPR Construction
The conclusion here: unused tools tend to retire themselves. Adoption is the most honest signal of whether a tool is delivering value. If a platform has been available for 12 months and the people it was designed to help have quietly stopped using it, that's a clearer verdict than any ROI calculation.
DPR maintains an innovation team specifically tasked with managing the experimentation and evaluation process. They identify what works, what doesn't, and what's worth scaling. The discipline of that process matters as much as the experimentation itself. Without a structured mechanism for retiring tools that aren't delivering, stacks accumulate by default rather than by design.
The practical criteria for retirement come back to the four buckets. A tool that isn't improving revenue, reducing cost, improving employee experience, or improving customer experience — and where adoption data confirms that the people using it don't value it enough to miss it — is a strong candidate for retirement regardless of how long it's been in the stack.
How do you align technology investment with business strategy?
Short answer:
Start with strategy and work backward to technology, not the other way around. A strategy specific enough to function as a filter makes alignment decisions straightforward. If your technology strategy can't be stated in terms specific enough to evaluate a tool against, that's the problem to solve first.
Long answer:
The gap between technology investment and business strategy is one of the most common (and most expensive) problems in construction technology. Firms buy tools because they're innovative, because competitors are using them, or because a vendor demo was compelling, rather than because the tool advances a specific strategic objective.
Atul's framework starts with the strategy and works backward to the technology rather than the other way around.
"We look at the software stack to figure out how that closely enables our strategy. We should deliver predictable outcomes for our customers through the use of virtual design and construction — really help with the design process so we can complete fabrication-level designs, and then utilize industrialized construction and prefabrication."
Atul Khanzode, CTO, DPR Construction
That strategy — predictable outcomes through VDC and industrialized construction — is specific enough to function as a genuine filter. A tool either advances fabrication-level design, supports self-performed work crews, or improves coordination outcomes, or it doesn't. There's no ambiguity about whether a given investment is strategically aligned.
For firms without that level of strategic clarity, the first step isn't evaluating new tools. It's defining what the technology stack is supposed to help the business achieve, in terms specific enough to make the alignment question answerable.
Platforms like Revizto's Connected Project Intelligence support that alignment by turning coordination data into actionable project insight — giving technology leaders the evidence base they need to demonstrate that VDC investment is delivering measurable outcomes rather than just process compliance.
How do you evaluate new construction technology before committing to it?
Short answer:
Structured experimentation is more reliable than either early or late adoption — the discipline of the evaluation process matters as much as the experimentation itself. Ground experiments in real operational pain points rather than theoretical capability. For AI specifically, data foundations, experimentation infrastructure, and internal training all have to be in place before any tool can deliver value at scale.
Long answer:
The pace of innovation in construction technology — and particularly in AI — means that new tools are arriving faster than most organizations can evaluate them properly. The risk of moving too slowly is missing a genuine competitive advantage. The risk of moving too fast is accumulating tools that don't integrate, don't get adopted, and don't deliver value.
DPR's approach is structured experimentation — a deliberate process of testing, evaluating, and scaling what works rather than either adopting everything or waiting for certainty before adopting anything.
The innovation team plays a central role in that process, providing the organizational infrastructure for experimentation to happen consistently rather than sporadically. Working directly with operational teams and functional work groups to identify pain points and quick wins ensures that experiments are grounded in real problems rather than theoretical capability.
For AI specifically, Atul identifies three components that have to be in place before any AI tool can deliver value at scale: clean data foundations, a mechanism for moving from experimentation to scaled adoption, and internal training infrastructure. Without all three, even the most capable AI tool will underperform.
The Revizto integrations page offers a practical starting point for firms evaluating how their coordination platform connects to the broader technology environment — a key consideration for any firm building the data infrastructure that AI requires.
Where does construction technology go from here?
Atul's closing perspective is both optimistic and grounded. The pace of AI development is genuinely unprecedented — things are possible today that weren't imaginable 18 months ago. But the fundamentals of what makes technology valuable in construction haven't changed.
"I keep going back to really fundamental things for us as a construction company. Instead of getting enamored by just the technology tools, really having a dialogue internally about what is it that you're trying to accomplish."
Atul Khanzode, CTO, DPR Construction
The firms that will get the most from the current wave of innovation are the ones that stay anchored to those fundamentals — strategy first, data foundations second, experimentation third, and education throughout. The technology changes every two months. The discipline required to use it well doesn't.
If you're wondering whether Revizto moves the dial on revenue, cost, employee experience, or customer experience — we'd rather show you than tell you. Talk to us today and we’ll walk you through it.
FAQs
The most reliable approach is to evaluate software against your firm's core business strategy rather than generic ROI metrics. Define what the technology is supposed to help you achieve — predictable outcomes, reduced rework, faster delivery, improved margins — and measure performance against those specific objectives. Atul Khanzode at DPR Construction uses four buckets: does it improve revenue and margin, reduce cost, improve employee experience, or improve customer experience? Tools that can move one or more of those dials with evidence earn their place.
Adoption is the most honest signal. If a tool has been available and the people it was designed to help have stopped using it, that's a clearer verdict than any ROI calculation. Apply the same four-bucket framework to retirement decisions as to acquisition decisions — tools that can't demonstrate impact on revenue, cost, employee experience, or customer experience are strong candidates for removal regardless of sunk cost or internal advocacy.
Start with the strategy and work backward to the technology. Define what your technology stack is supposed to help the business achieve in specific enough terms to function as a filter — then evaluate every tool against that filter. If a tool doesn't advance the strategy, its technical sophistication is irrelevant. Firms that buy technology because competitors are using it, or because a demo was compelling, consistently end up with bloated stacks and low adoption.
The key is distinguishing between cost growth that reflects genuine value creation and cost growth that reflects stack bloat. Rising licensing costs alongside rising revenue, stable margins, and high adoption rates is a different story to rising costs with flat revenue and declining adoption. A structured innovation team and evaluation process — with clear criteria for what gets scaled and what gets retired — prevents stacks from accumulating by default.
Apply the four-bucket framework before committing to any AI tool: does it improve revenue and margin, reduce cost, improve employee experience, or improve customer experience? Also ensure the three foundational components are in place before deploying AI at scale — clean data foundations, a mechanism for moving from pilot to scaled adoption, and internal training infrastructure. AI tools deployed without those foundations consistently underperform regardless of their technical capability.
