This blog content comes from an interview conducted on March 26, 2026. For more information or context, watch the full episode where you get your podcasts.
“No AEC marketer wants to hear that they can do more proposals.”
- Rachelle Ray, Head of AEC Innovation, OpenAsset
AI showed up in marketing and business development before almost anywhere else in AEC, and the early excitement came with plenty of hidden risks. In this segment of the AEC GrowthCast, host Greg Hart talks with Rachelle Ray, an AEC marketer with more than 15 years of experience and now the Head of AEC Innovation at OpenAsset, about where firms are going wrong with AI and what leaders can do to protect their quality, reputation, and bottom line.
1. “Do more, faster” is the wrong pitch, and one bad output can end adoption
“The thing I saw a lot of from AI tools, especially AI proposal tools when they were first starting to roll out, was this messaging around, ‘You can do more faster,’ and no AEC marketer wants to hear that they can do more proposals. And so there was some resistance there, like, ‘Great, you’re handing me this thing when I need something that’s going to make it easier for me to lift the five, 10 proposals I already have. And you’re giving me this thing that’s telling you I can do 20 now.’”
Trust breaks easily, and leaders often miss it:
“There is some hesitation. If it hallucinates, if it goes wrong, there’s instant distrust, and leadership doesn’t always see that if somebody doesn’t use it. We hear it all the time when I go to conferences. Somebody’s like, ‘Yeah, we just bought this thing and it did something dumb. So I just switched back to what I was doing before.’”
Her answer is enablement, not abandonment:
“Unlike a lot of other tools, there is no one right way to use AI. So maybe you found the wrong way, and you do need to do some work with your team, some sharing of lessons learned and best practices, to find the right ways.”
2. Expect a growing divide between firms that adopt AI well and those that don’t
“I think we’re on the cusp of seeing that. I expected to see it a lot sooner, but we’re not quite there yet, because I think we’re still in that adopting phase. We are historically slow to adopt new technology in this industry. It took us like 20 years for Revit to become the standard. And now you can’t imagine life without it.”
She sees the gap becoming impossible to ignore:
“I think in the next five years, that rift will be significant. And you will definitely see that difference between the firms that have adopted these tools and have adopted them well, have implemented that change management, have really woven it into the fabric of their company, and the ones that are either still kind of poking around and playing with it, or the ones that said, ‘This is too risky, this is too dangerous, we’re not going to touch it.’”
3. Unreviewed AI output becomes everyone’s problem
Ray describes a proposal manager at a firm that rolled out an AI proposal automation tool. Its “magic button” auto-answers questions, and the team is supposed to check the results:
“When the AI answers those questions, it doesn’t do yes or no. It does kind of that AI bloat thing. So they’re kind of chunky answers, and they weren’t being edited down by either the technical staff or the marketing staff. And a client reached out to one of the teams that had turned in these proposals and said, ‘Hey, you might want to check your AI.’”
The fallout reached the top:
“Leadership found out at that point, which was bad, and had to send out a memo to the entire company saying, ‘Hey, you have this great tool. It’s amazing. Please review what it’s putting out, because this is a problem.’ So it affects everybody.”
4. AI governance belongs to the whole firm, not just IT or marketing
“This does not belong solely in IT. It doesn’t belong in marketing just because marketing was the low-hanging fruit in the beginning. AI is such a fascinating new technology. It’s unlike anything we’ve seen before. It touches every facet of your business.”
The firms she sees succeeding spread ownership around:
“They are governing it in a 360 way. So they have AI committees or review panels that are made up of representation from all corners of their firms. You get marketing in there, you have leadership in there, you have IT in there. It has to be cross-functional, cross-evaluated. And leadership has to be an active participant in that.”
That means a different role for the C-suite:
“In the past, it was kind of, ‘We trust each department to argue for the need of whatever tools they need,’ and leadership signs off on the checks. But they have to understand now how that tool interacts with their firm data and their staff workflows and how one flows from the other.”
And a clear set of rules:
“Everyone needs to come together to define what those guardrails are. What can the AI access, and who is authorized to use it, and what review processes exist before that AI-generated content goes out the door.”
5. Make change management non-negotiable, and normalize failure
“I think that change management piece is absolutely non-negotiable with AI. You can’t just hand this tool to your team and expect them to figure it out on their own.”
She suggests treating AI like a new teammate:
“If you were thinking about AI, because of how quote-unquote smart it is, as an extension of your team rather than a tool, you almost need to treat it that way as well, and do lessons learned on it, and make sure that you are normalizing, ‘This didn’t work. Let’s figure out how to fix it.’ We used to do that in our postmortems on proposals: what went well, what didn’t.”
Her point is that AI mistakes aren’t always user error:
“Have that conversation about, I didn’t use the tool wrong. Sometimes the tool does things you don’t expect, because it’s not predictable. That’s just the nature of generative AI. And so we have to normalize that conversation around this didn’t work and create a feedback loop so that problems surface early and not after the proposal has been submitted.”
Human judgment stays in the loop:
“Definitely frame it as something that needs human judgment. It’s not a magic button. It’s not a silver bullet. Set those healthy expectations from the start. I see success when you are healthily skeptical about the AI.”
6. The marketing team of the future produces proposals instead of writing them
“Someone the other day said something very profound to me. She said, ‘I no longer write proposals, I produce them.’ And I think that is such an amazing way of thinking about AI, because you’re still involved, but you are directing the creation of the proposal. You’re not directly involved in the crafting of that, but you’re still in it.”
Ray is optimistic about what comes next:
“I’m really, really optimistic that the team of the future is just going to be super empowered with incredible insight, because all of their scattered data is going to come together, not be so burdened by that administrative task of hunting and pecking. We call it dumpster diving, right? Going into the archives to find something. We’ll have that time and space to really think through, ‘What’s a cool thing we can do? What’s an experience we can provide for someone reading our marketing materials or talking to our people?’”
Hear the full conversation
Watch or listen to the complete episode above, and subscribe to the AEC GrowthCast on YouTube, Spotify, or Apple Podcasts so you don’t miss the next one.
Ready to put AI to work at your own firm? Explore PSMJ’s AEC AI Acceleration Workshop.