In this episode
Gina Bhawalkar, Principal Analyst at Forrester, joins Greg Kihlström live from Forrester CX Forum East in Brooklyn to lay out how enterprises can build great customer experiences with an AI-enabled design workflow — without trading quality for speed. Drawing on her research into experience design, design systems, and digital accessibility, Gina makes the case that AI should accelerate human judgment rather than replace it, that a machine-readable design system is the foundation for scaling AI-powered experiences, and that responsible design and hard-earned customer trust are what separate differentiated brands from a wave of forgettable, look-alike experiences.
Key takeaways
- Trust, not speed, is the real risk. Fewer than 20% of consumers trust AI-powered interactions like chatbots and virtual assistants, so shipping fast-but-flawed experiences deepens a trust problem that already exists.
- Customers aren’t asking for AI or for speed — they have a job to be done. Whatever a brand builds has to help customers reach an actual goal, which is why deep experience research is becoming the real differentiator.
- Use AI as an accelerator, not a replacement for human judgment. Discernment, empathy, and ethical rigor are human qualities AI lacks, so brands should keep investing in professional designers, developers, and PMs as stewards of quality.
- “Never delegate understanding.” Borrowed from a chief design officer at National Australia Bank, the principle guards against letting AI hold customer knowledge that should live in people’s heads — the foundation of customer-obsessed decisions.
- A machine-readable design system is the critical foundation for designing with AI. Forrester predicted eight in ten companies would invest more in design systems to mitigate AI risk, and Gina says that prediction is clearly playing out.
- Start with your biggest internal bottlenecks, not with everything. Map how product, design, and development work together, then point AI at the worst friction — one large tech company’s biggest mistake was simply telling designers to “start experimenting.”
- Report efficiency as effort reclaimed, not hours saved. Instead of “we cut analysis time,” frame it as “we can now run 15 more studies a year,” tying reclaimed time to more customer-informed decisions and business growth.
- Human creativity is the differentiator in the AI era. Companies that reinvest saved time into new products and markets win; those that use it as a reason to cut 40% of their designers are playing a short-term game.
- Responsible design needs active guardrails. LLMs don’t produce accessible designs by default, an “impact” dimension belongs alongside desirability/viability/feasibility, and not-yet-ready use cases (like AI research moderation) should stay out of production.
Chapters
- 00:52 — Cold open: getting efficient at forgettable experiences?
- 02:17 — Gina’s background and what she covers at Forrester
- 03:28 — The primary risk of prioritizing speed over quality
- 05:32 — Why customers aren’t asking for speed
- 06:25 — The human-led, intentional decision model
- 07:25 — “Never delegate understanding”
- 10:04 — Making the design system AI-ready
- 12:52 — Finding high-value, low-risk AI use cases
- 16:28 — Measuring success and proving ROI
- 19:01 — Humans as the real differentiator
- 20:20 — Responsible-design blind spots
- 23:59 — The skills design and CX teams will need
- 27:21 — Forrester CX Forum East and staying agile
Why trust — not speed — is the biggest risk of moving fast
Forrester’s data shows trust in AI-powered interactions is already low, with fewer than 20% of consumers trusting experiences like chatbots and virtual assistants. When teams prioritize speed over quality, they ship experiences that aren’t aligned to a real customer need, aren’t accessible, and are riddled with usability problems — all of which reflect poorly on the brand and compound the existing trust gap. Gina flags accessibility compliance as a specific casualty: moving too fast leads brands to skip the testing that keeps them inside accessibility regulations.
What “never delegate understanding” means in practice
The human-led decision model comes down to using AI as an accelerator while keeping human judgment in charge. Gina points to a chief design officer at National Australia Bank whose team adopted the principle “never delegate understanding” — a guardrail against handing AI the planning, running, and analysis of research to the point where no one in the organization actually holds customer knowledge anymore. That knowledge, she argues, is the foundation of customer-obsessed decision-making, so the best teams stand up research squads to vet AI use cases against real customer goals before deploying them.
Why the design system is the foundation for designing with AI
Gina says “design system” is the term she uses more than any other right now, because it’s the foundation enterprises need to design effectively with AI. The work is to make an existing system — style guide, component libraries in design and code — AI-ready and machine-readable, with documentation thorough enough that AI knows which rule or component to use when. Beyond consuming the system, AI can help evolve it: generating documentation, spinning up component variants faster, and, at the most advanced companies, contributing new variants back into the system for others to reuse.
How to find high-value, low-risk AI use cases
Rather than pointing AI at everything, Gina recommends mapping internal journeys — how product, design, and development actually work together — and targeting the biggest bottlenecks. Those friction points are where AI can reduce rework and improve collaboration. She cites a large tech company whose biggest mistake was telling designers to “start experimenting,” which turned into a directionless free-for-all; the fix was refocusing all experimentation on one high-value use case, getting from idea to prototype faster.
How to measure success beyond hours saved
Efficiency is the easy metric, but Gina pushes leaders to reframe it as effort reclaimed and reinvested — for example, running 15 more studies a year rather than reporting hours saved in analysis and synthesis. Fewer rounds of rework is another meaningful outcome, since high-fidelity prototypes can drive teams to alignment with fewer cycles. Ultimately, she says, the UX metrics still matter: whether AI is producing better experiences shows up in conversions, task success, and ease-of-use scores.
The responsible-design blind spots to watch
Gina names three recurring pitfalls: skipping the automated and manual testing needed to catch accessibility violations, because LLMs don’t produce accessible designs and code by default; failing to weigh a fourth “impact” dimension alongside the desirability, viability, and feasibility framework — using tools like “bad headlines” and consequence-scanning workshops to surface potential harm before building; and jumping into use cases that aren’t ready for prime time, such as clunky AI research-moderation tools that undermine trust rather than build it.
FAQ
What’s the biggest risk of prioritizing AI speed over quality in CX? Erosion of consumer trust. Forrester finds fewer than 20% of consumers trust AI-powered interactions, and rushing out experiences that aren’t aligned to customer needs — or that fail accessibility standards — makes that trust problem worse.
What does a “human-led decision model” actually mean? Using AI as an accelerator rather than a replacement for human judgment. It means continuing to invest in designers, developers, and PMs for their discernment, empathy, and ethical rigor, and keeping customer understanding in people’s heads instead of fully delegating it to AI.
Why does Gina emphasize design systems so heavily? A machine-readable, well-documented design system is the foundation for scaling high-quality AI-powered experiences consistently. Forrester predicted eight in ten companies would invest more in design systems to mitigate AI risk, and Gina says that is clearly happening.
How should teams measure the impact of AI on their design workflow? Frame efficiency as effort reclaimed and reinvested — e.g., running more research studies per year — rather than hours or dollars saved. Also track fewer rounds of rework and core UX outcomes like conversions, task success, and ease of use.
What skills will CX and design teams need? A combination of general AI literacy and function-specific skills: “speaking machine” through effective prompting, evaluating AI outputs for bias and quality, grounding in ethical and accessible design, and comfort with ambiguity — with a healthy dose of skepticism.
About Gina Bhawalkar
Gina’s research focuses on digital accessibility and experience design. Gina established and now leads Forrester’s coverage of the digital accessibility space and has a deep background and interest in the topic. She advises organizations on how to establish and scale sustainable accessibility practices and is an expert on the digital accessibility platform (DAP) market. Her research on accessibility has appeared in publications such as The Wall Street Journal and The Financial Brand, and she is a frequent speaker at accessibility events and podcasts. Gina’s other areas of expertise include inclusive and responsible design, design systems, personas, and measuring the impact of digital experience design improvements.
Gina has over 20 years of experience as both a UX/CX practitioner and leader, with eight years in the financial services industry. Prior to joining Forrester, Gina was the director of customer experience research at Bank of the West, a subsidiary of BNP Paribas. There, she led the bank’s digital voice-of-the-customer program and conducted primary research to inform digital product development. Previously, Gina led the user experience and accessibility department at Scottrade, where she built the two disciplines from the ground up. Earlier in her career, Gina was a UX consultant at Perficient, leading design and research projects for clients in the financial services, agribusiness, utilities, insurance, and retail industries. She has also served as an accessibility consultant at both Criterion 508 and the Georgia Tech Research Institute.
Gina holds bachelor’s degrees in psychology and computer science from Trinity University and an MS in human-computer interaction from the Georgia Institute of Technology, where her research focused on evaluating the accessibility of physical and digital products to people with disabilities.
Gina Bhawalkar on LinkedIn: https://www.linkedin.com/in/ginabhawalkar/
Resources
Forrester: https://www.forrester.com
We’re proud to be a media partner for #MAICON26 – Oct. 13-15! Learn how AI can power your marketing and business and help you grow smarter. Use code AGILE150 to save! https://aglbrnd.co/r/7fe458ced0f04658
Reach your customers with Reddit. Spend $500 in ad spend, get $500 back in ad credit! Learn more: https://advertalize.com/r/491818c79fb1873f
The most influential minds in software, AI, and engineering leadership will be at WeAreDevelopers World Congress North America, September 23-25 in San Jose. Learn more: https://aglbrnd.co/r/60a7299222a7bcf1
Enjoyed the show? Tell us more at and give us a rating so others can find the show at: https://aglbrnd.co/r/faaed112fc9887f3
Connect with Greg on LinkedIn: https://www.linkedin.com/in/gregkihlstrom
Don’t miss a thing: get the latest episodes, sign up for our newsletter and more: https://aglbrnd.co/r/35ded3ccfb6716ba
Check out The Agile Brand Guide website with articles, insights, and Martechipedia, the wiki for marketing technology: https://www.agilebrandguide.com
Transcript
[00:00:52] Greg Kihlström: Hi. I’m Greg Kihlström, host of The Agile Brand. And here’s a question for you. As generative and agentic AI promise to accelerate everything we do, are we at risk of becoming incredibly efficient at producing forgettable customer experiences? Today, we’re here in Brooklyn at Forrester CX Forum East, and we’re gonna talk about building better experiences with an AI-enabled design workflow. Specifically, we’re gonna cover balancing the speed of AI with the intentional human-led decisions required to craft exceptional experiences, the role of a robust design system in scaling high-quality, AI-powered experiences consistently across the enterprise, and how to identify high-value AI use cases for your design workflow while maintaining responsible practices that build customer trust. To help me discuss this topic, I’d like to welcome Gina Bhawalkar, principal analyst at Forrester. Gina, welcome to the show.
[00:02:17] Gina Bhawalkar: Thanks for having me. I’m excited to be here.
[00:02:18] Greg Kihlström: Yeah. Looking forward to this, uh, looking forward to the conference, but also, you know, looking forward to talking with you about this. And before we dive in, why don’t you give a little background on yourself and your role at Forrester?
[00:02:28] Gina Bhawalkar: Yeah. Absolutely. So my background is in user experience. I’ve been a user experience designer, UX researcher. Um, before Forrester, I led user experience teams at several companies, predominantly in financial services. And then I got hired eight years ago by Forrester to essentially cover the field that I was working in. So I got hired to do research on how companies build and scale their design practices. Um, and to date, um, today I still cover experience design, um, but the topics have obvi- obviously changed a little bit. So today, I do a lot of research on design systems, on how AI is impacting the profession of experience design, which I know is what we’re here to talk about today. And then I also lead
[00:03:13] Gina Bhawalkar: Forrester’s coverage on digital accessibility compliance. So I tend to do a lot of research on how do we ensure our designs are responsibly designed, inclusive, and that we’re n- not, we’re not leaving anyone out, essentially, in the experiences that we put into market.
[00:03:28] Greg Kihlström: Yeah. I’m talking to the right person about this, this topic here. So let’s, uh, let’s dive in here. So, you know, I, I talk with a lot of leaders on this show and, and elsewhere. So many are under pressure to do AI and, and do it rather quickly, right? From your research and, uh, you know, what’s the primary risk that organizations face when they prioritize the speed part of AI over the quality part of that, not only the, the implementation, but that end user experience?
[00:04:01] Gina Bhawalkar: Absolutely. I mean, there, there’s a lot of risk associated with moving too fast right now. Um, but to your point, that pressure is really there. So to me, the biggest risk, it comes down to trust and this further erosion of consumer trust. We see in our data at Forrester that trust is already pretty low, particularly in AI-powered interactions. So think of things like virtual assistants and chatbots. I mean, we’re talking less than 20% of consumers actually trust those experiences. And so this is what brands are up against. And the problem is when we prioritize speed over quality, that results in experiences that aren’t actually aligned to a customer need, um, that aren’t accessible, that are riddled with usability issues
[00:04:46] Gina Bhawalkar: that make me think, you know, what is this even about?
[00:04:49] Gina Bhawalkar: You know, it reflects poorly on the brand. And so all of these things just add to this trust problem that already exists. And then if I could add one more, ’cause I mentioned I do a lot of research on accessibility compliance, I’m seeing a lot of brands face this risk of putting experiences into market that don’t meet accessibility standards because they’re trying to move so fast that they’re foregoing the important testing they need to do to make sure they’re not falling out of compliance with accessibility regulations. And so that, you know, erosion of trust, accessibility risks, those are the biggest challenges I see right now if organizations don’t pivot back to quality, um, and not sacrifice that due to this need to work more quickly.
[00:05:32] Greg Kihlström: Yeah. ‘Cause I mean, the, the customers, uh, sure, they want their brands to adopt things, you know, that, that they’re using and technologies that they wanna, um, access. But- The customers aren’t really asking for speed, are they?
[00:05:46] Gina Bhawalkar: No, absolutely not.
[00:05:48] Gina Bhawalkar: At the end of the day, customers aren’t asking for AI, they’re not asking-
[00:05:51] Greg Kihlström: Right. (laughs)
[00:05:51] Gina Bhawalkar: … for technology. They have- a job to be done, or they have- a goal. Um, and whatever you’re creating as a brand, it has to service customers being able to accomplish that goal. And so, this is why I think the organizations that have really robust practices around experience research have a leg up at this moment in time, because they’re actually grounding decisions in deep understanding of those customer goals and those customer journeys. And that’s increasingly gonna be the secret sauce to differentiating in this time when it’s very, very easy to create.
[00:06:25] Greg Kihlström: Yeah. Yeah. Well, and so that probably, you know, that brings us to the, this concept of a human-led intentional decision model. Can you maybe unpack that for us? You know, what, what does that mean and what does it look like when it’s done well?
[00:06:39] Gina Bhawalkar: I would boil it down to we need to make sure we’re using AI as an accelerator, but not a replacement for human judgment. And so, what that means in practice is, as a brand, you need to continue to invest in hiring and developing professional designers, professional developers, professional PMs, because these people possess these unique qualities that AI does not. And I’m talking about things like discernment, um, being able to actually look at a design and say, “Yes, this is high quality. This is actually aligned to something customers need.” Um, possessing empathy. Machines are not empathetic. People are empathetic. Making sure that we are designing with deep customer understanding in mind. Ethical rigor.
[00:07:25] Gina Bhawalkar: So, think of the humans on your teams. They’re stewards of quality. And so, that’s why we talk about this human-led decision model. Also part of this is redefining design principles in this age of AI, and I’ll give you just one example, Greg, which is, I interviewed a chief design officer at National Australia Bank, leads a huge design organization, and he told me they have this internal team principle now of never delegate understanding. And he said this is so important as they’re making decisions about the role AI will play in planning, conducting, and analyzing research, because there’s this risk that if you just say, “We’re gonna let AI do all of it,” then all of a sudden the people in your organizations, they don’t hold any of that cu- customer
[00:08:10] Gina Bhawalkar: knowledge in their heads. And that’s not good because that knowledge is the foundation of customer-obsessed decision-making. And so, yes, we absolutely need to create principles like this to ensure we stay grounded, um, in the decisions that, that we’re making. And then the last thing I’ll say is companies that really get this human-led decision model right, they are grounding their decisions in where to deploy AI in the customer experience in discovery research. So for example, this might look like organizations standing up user research squads, where the whole pers- purpose of this squad is to look at those 100 use cases for AI that the company (laughs) has in mind, and to go out and vet them with
[00:08:55] Gina Bhawalkar: customers, to actually do research. Does this align to a customer goal? Would customers trust AI to do this thing that we’re considering? And so, making sure that, you know, at the end of the day, it’s all about grounding things in what your customers actually need and want and value. That’s, that’s the key, so.
[00:09:12] Greg Kihlström: Yeah. Yeah, that-that never delegate understanding, that, that really resonates (laughs) with m- yeah. I think I’ve seen many interfaces designed by AI, or at least I swear they were (laughs) because, uh, you know, the, they make a certain kind of sense, but not the, not a human sense (laughs), right?
[00:09:30] Gina Bhawalkar: And they also y- I mean, you’ve probably had this experience I have, where a lot of experiences are starting to all look the same.
[00:09:37] Gina Bhawalkar: And so, that’s another risk, going back to your earlier question of, we don’t wanna feel just like any other brand out there. We wanna, we don’t wanna become a commodity, right? We wanna create experiences that are differentiated, that feel like our brand, um, that adhere to our brand principles, promises. And that comes from making sure that you’re not just letting AI create and deploy experiences, but making sure that human expertise and judgment is part of the equation.
[00:10:04] Greg Kihlström: Yeah. So then, how do you, how do you do it right? You know, the, there certainly are benefits of scale and, and other things with AI. And so, you know, integrating AI into a design system would make a lot of sen- you know, not just handing over the keys (laughs) to, to your earlier point, but integrating it so that we can benefit from it. You know, what, what does that, what does that look like when it’s done well?
[00:10:28] Gina Bhawalkar: Yeah. So, you mentioned the word design system there, and I probably say this word more than anything else (laughs) these days because I do believe it’s the critical foundation that every organization needs, um, to design effectively with AI. And so, there’s a few things that brands need to do right now. One is, every brand probably has some semblance of a design system. It may be, uh, essentially a style guide. It may be reusable component libraries that are articulated in both design and code. Whatever you have, there’s important work to be done to make sure that design system is AI-ready, that it’s machine-readable, because it’s not just humans using it anymore to assemble experiences. It’s AI as well. So, making sure everything in your design system
[00:11:14] Gina Bhawalkar: is optimized for AEO, for example, that you have thorough documentation so AI knows, “When do I use this rule, or, versus this rule, or this component versus this component?” Um, all of that’s super critical. And then, there’s also a lot of opportunity to use AI to evolve the design system itself, because this is a common challenge I hear from brands, is, “We know we need a mature design system, but we don’t have enough resources to, to do it.” So, I’ve seen organizations use AI to help create documentation for the system, um, to use AI to spin up variants of components more quickly than they could in the past. And then, you know, we’re ultimately gonna be working towards this world where AI also becomes
[00:12:00] Gina Bhawalkar: a contributor back into the design system. You hear a lot about design system contribution models, and those are really centered around people and teams today. Um, but the most advanced companies we’re studying, they’re also enabling AI to make decisions like, “I need a variation of a component that we don’t have in the system, so I’m gonna create it, and then I’m gonna, like, essentially feed it back into the system so others can use and benefit it, from it as well.” Um, so really interesting stuff happening there in the context of AI and design systems. But, um, I think, you know, we predicted this year that eight of 10 companies would invest more in design systems to mitigate AI risks, and we are very much seeing that play out. I think that’s one of those predictions where we’ll be able to confidently say,
[00:12:45] Gina Bhawalkar: “Yes, it came true.” (laughs) Which is not the case with most predictions, so.
[00:12:49] Greg Kihlström: … and so for those, those companies investing, I…
[00:12:52] Greg Kihlström: You know, what’s a … I, I know obviously every company is different, every, you know, every experience is, is different in some ways, but you know, where’s a, where’s a way for, or a framework even, for identifying those high value low risk use cases to, to start this? ‘Cause you know, some may be a little, not quite as far along as others.
[00:13:11] Gina Bhawalkar: Absolutely. So journey mapping is probably something a lot of people listening to this podcast are familiar with. I recommend that teams actually map their internal journeys. How do you work together across product, design and development, and where are the biggest bottlenecks in that process today?
[00:13:29] Gina Bhawalkar: Those are great places to start. How could AI help us collaborate better? How could AI help us reduce the amount of rework that happens at this phase in the journey?
[00:13:39] Gina Bhawalkar: So really picking those places where, you know, things aren’t going so well today, maybe this is an opportunity where AI could help us do better, create better quality outputs, um, work better as cross functional teams.
[00:13:53] Gina Bhawalkar: So I think it’s, start there. Don’t just try to point AI at everything. We, we see-brands making this mistake in their-experiences.
[00:14:01] Gina Bhawalkar: But it applies internally too.
[00:14:03] Gina Bhawalkar: Um, I actually interviewed a large tech company where they said the biggest mistake they made was they told their designers, “Start experimenting with AI.”
[00:14:11] Gina Bhawalkar: And it was like the wild west. People were using these tools for all these different purposes, but there was no focus to it.
[00:14:17] Gina Bhawalkar: So they ended up saying, “Okay, no, we need to really hone this in on where we see the biggest opportunity in the design process today.” And for them, that was getting from idea to prototype faster.
[00:14:30] Gina Bhawalkar: And so they refocused all their experimentation on implementing that particular AI use case. So thinking very intentionally about where you point AI in your workflow is essentially the advice that I would give there.
[00:16:28] Greg Kihlström: And certainly another big thing, you know, I hear all the time is just, you know, I, there certainly is all … plenty of experimentation continuing to go on, but, uh, I feel like the bill is coming due as far as showing ROI and, and you know-kind of proving the, the, the model, no pun intended. How do you advise leaders to measure success of, of things like what you’re, what you’re describing? You know, what, what are some of the key metrics that would go into, into, you know, into this?
[00:16:57] Gina Bhawalkar: Yeah. The most obvious one that people are quite good at measuring is the efficiency metrics, right?
[00:17:03] Greg Kihlström: Right. Right.
[00:17:04] Gina Bhawalkar: But what I tell people there is, “Don’t focus on we saved this amount of money, or we reduced the hours to create concepts from this to this. Talk about it as, how much effort were we able to reclaim and reinvest back into the business.”
[00:17:19] Gina Bhawalkar: And so for example, um, many companies are, you know, cutting the time to do analysis and synthesis in research, um, in some cases from weeks to days. Fantastic. Don’t go and report that as hours saved, saved in analysis and synthesis. Talk about it as, “We can now do 15 more studies per year because we’ve freed up that time. And because we can do 15 more studies per year, we can now ensure that X more decisions at our organization are being infused by that important customer understanding that we gain through research.” And so I would encourage you to, you know, think about efficiency in that way. But then other metrics that I’ve seen, obviously things like fewer
[00:18:04] Gina Bhawalkar: rounds of rework. This is a problem today in product design and development, is lots of, you know, spin, lots of circling that happens. Um, and you know, having to constantly iterate and iterate and iterate on designs because teams can’t make decisions effectively. And so if we can use AI maybe by creating high fidelity prototypes that help us get to alignment faster, and we can measure how many fewer cycles of rework happen, that’s a really powerful outcome that we can talk to. And then obviously you should be looking at your UX metrics. So at the end of the day, um, you know, if you’re a bank and the main experiences that you’re creating are mobile banking and online banking apps, then are product outcomes improving? Um,
[00:18:49] Gina Bhawalkar: is AI actually leading to better experiences as reflected in things like more conversions, higher rates of task success, higher ease of use scores? So those are all critical metrics as, as well.
[00:19:01] Greg Kihlström: Yeah. Yeah, I mean, it, it seems like … I, I think a common thread across that is you …Any major competitor in a market is probably going to be doing similar enough things to get efficiency, right? So, like, what’s the differentiator is what the humans are doing, right? Is that, is that safe to say? I mean, the, the, you know, the human creativity or strategy or just best use of time, right?
[00:19:25] Gina Bhawalkar: Absolutely. I mean, in my, in my talk at this event, I talked about it as, you know, your humans are still your superpower. So to your point, like, the people who have those great thinkers in their organization, the organizations that recognize that by saving time, we can grow the business more. We can enter new markets. We can stand up new products. Those are the organizations that are going to win in this AI era, not the ones who say, “Oh, we’re saving all this time in our product design and development process. We can fire 40% of our designers and developers.” You know, you, you could certainly think that way, right? If you just wanna keep doing the status quo things that your company’s always done, but if you wanna grow your business, that’s not th- … right.
[00:20:04] Greg Kihlström: Yeah, that’s not a long-term play. That’s a short-term-
[00:20:06] Gina Bhawalkar: Not a long-term play.
[00:20:20] Greg Kihlström: And so, you know, you’ve, you’ve touched a little bit on responsible design, but I wanna talk a little bit about, about that as well, ’cause, you know, I think it’s, it’s a pretty broad topic. You know, it’s, it’s not just one thing, you know, that, that it covers, uh, as well, and, and certainly if cons- if consumer trust is not exactly at its, at its peak right now either, certainly a critical piece there. So, you know, what are some of the most common, um, you know, blind spots that you see when organizations implement AI-in-experienced design that could in- inadvertently erode that, that trust even more, I guess? (laughs)
[00:20:47] Gina Bhawalkar: Absolutely. So, I would say three things come to mind. One would be kind of related to the accessibility piece that I mentioned earlier, not putting your AI outputs through the rigorous testing needed to make sure that they aren’t violating ethical principles, such as accessibility. Um, LLMs are not creating accessible designs and code by default. Some organizations make the mistake of thinking they are, but they’re not. So it’s still really critical to make sure that you are running the appropriate automated and manual tests to ensure you’re not violating those standards. So that’s, like, one, I would say, blind spot that needs to be overcome. The second is, in Agile, right, you have, like, the DVF framework, desirability, viability,
[00:21:33] Gina Bhawalkar: feasibility framework. We published a report a couple years ago on responsible design where we proposed that a fourth dimension be added to that, which we would call impact. And this is, how are you considering the end impact that implementing this product or feature is going to have on customers, on society, on the environment? And if it’s determined that it’s more likely this thing would harm customer, society, and environment than help, then we pull back, and we don’t implement that particular thing. So, that’s, you know, the second, I guess, pitfall is not considering that important impact dimension. But the good thing is, there’s really great tools out there to do that. So, my favorite is Bad
[00:22:18] Gina Bhawalkar: Headlines workshops. I don’t know if you’ve ever done one of these, but it’s essentially, okay, we’re considering building this thing for customers. Let’s imagine the worst possible headline that could be published if we don’t get this right. And that can end up surfacing some really great conversation around, whoa, actually there’s so much risk of things going wrong, like, these headlines are really, really bad. We probably shouldn’t even go here. We should be ex- looking at other use cases instead. Um, consequence scanning workshops are also really helpful here. I’ve heard of Black Mirror brainstorms, if you’ve ever watched Black Mirror.
[00:22:53] Greg Kihlström: Yeah, yeah. (laughs)
[00:22:54] Gina Bhawalkar: We don’t wanna be a Black Mirror episode.
[00:22:55] Greg Kihlström: (laughs) Right.
[00:22:56] Gina Bhawalkar: So, the intent of all of these is just to explore potential flaws that could allow for abuse or other negative outcomes-
[00:23:03] Gina Bhawalkar: … on our customers. And then the, the third thing that comes to mind is jumping into use cases that aren’t ready for prime time. For example, I see some research teams exploring AI moderation tools. I think there’s potential in these tools, but the fact of the matter is, that right now, the interfaces are clunky. They’re odd. It’s a very strange participant for the experience. If you’ve ever been in a research study where AI is moderating it, it might ask you the same question three different ways at different points in the interview, and you’re like, “What’s going on here,” right? And it doesn’t really make you trust that brand. And so being very careful not to dive into some of these use cases that, while promising, they’re just not quite ready yet. The technology is not quite there.
[00:23:48] Gina Bhawalkar: So those are a few of the kind of pitfalls that I, that I see and things I would be very careful about to make sure that you’re, you know, designing and deploying these experiences responsibly.
[00:23:59] Greg Kihlström: Yeah. Yeah. So let’s, let’s look ahead a little bit, and, you know, certainly we’ve talked, uh, about quite a few things here from, that require, I would say, a mindset shift if, if nothing else. But let’s … wh- what is the, what do the teams look like that support this? So, you know, what are, what are some of the skillsets or, or other things that design and CX professionals are gonna need to really embrace this a- you know, and, and, and, and make it great?
[00:24:29] Gina Bhawalkar: It’s kind of a combination of general AI literacy and then, you know, function-specific, um, AI skills. So- … general AI literacy, um, you know, anyone at the CX Forum here this week has heard us talk about Forrester’s AIQ, our artificial intelligence quotient, which is a great tool for leaders to understand where, you know, your team is at from an AI perspective, um, and how that compares with your perception of where they’re at as leaders, and identify specific ways to just uplevel their understanding of basic AI concepts. So that’s, like, task number one. And then on the more function-specific skills, so leaders of d- design teams, CX teams, your teams need to get really
[00:25:14] Gina Bhawalkar: good at speaking machine. I’m borrowing, um, that term from John Maeda. He wrote a whole book (laughs)- on that topic. I, I brushed it off and started reading it again recently. We have to know how to speak to these AI tools, right? So effective prompting. If we’re using AI tools to create prototypes, how do we effectively tweak prototypes to get to good outcomes? So that’s really important. Evaluating the quality of AI outputs, so practicing what does it look like to detect bias? What are the types of things we need to be spotting? And then we talked a bit about responsible design. I hope that all product design development teams will brush up on ethical design practices, because AI is not the steward of ethical (laughs) responsible design. The human is. Um, and it’s more important now than ever, because the opportunity for things to go wrong, it’s just happening on a massive scale that we haven’t seen before.
[00:26:10] Gina Bhawalkar: So those are some of the biggest things, and then obviously, like, just getting comfortable with ambiguity, with change. You know, the tools are gonna, you know, keep evolving. We see new tool (laughs) announcements every week, and so I think as, you know, professionals, just getting comfortable with the fact that things are gonna keep changing, and we need to be comfortable experimenting, trying things out, knowing things aren’t always going to work perfectly, but, you know, being curious to explore.
[00:26:36] Greg Kihlström: And maybe even just a, a, a tad, uh, or a little bit of skepticism (laughs), right?
[00:26:41] Greg Kihlström: It’s, I mean, again, I’m, I talk s- very positively about AI and all the stuff most, most of the day, but I, I know not to just take it at its word or the first, you know, result of a prompt, right?
[00:26:55] Gina Bhawalkar: Yes, and I think overindex on the skepticism, right? (laughs)
[00:26:58] Greg Kihlström: (laughs) Yeah.
[00:26:59] Gina Bhawalkar: Like, ask more questions. I think anyone listening, like, be the person in the meeting who’s asking those tough questions, like, “Who could we be leaving out if we launch this experience? What’s the potential bad headline (laughs) that could result if we launch this experience?” These are the things that are gonna prompt those necessary conversations that often aren’t happening because organizations are moving so quickly.
[00:27:21] Greg Kihlström: Well, Gina, thanks so much for joining today. A couple last questions as we wrap up here. First, uh, you know, Forrester CX Forum East is about to kick off here. Uh, what are you most looking forward to?
[00:27:33] Gina Bhawalkar: I’m really excited about the unveiling of our new EX Index. Um, it’s the new component of a brand’s total experience score. This is really exciting to me, because when I was working in the field of UX, you know, I would often be mapping customer journeys, and then I would map employee journeys to show, hey, look, if the employee journey’s broken, customer experience (laughs) is also broken. But it was a hard sell, right? Um, business leaders, they need to see the numbers to really prove to them that employee experience is so important to creating great CX. And so I feel like our new EX Index does that, um, and so I’m super excited.
[00:28:13] Greg Kihlström: Looking forward to it, yeah. And, uh, last question for you, what do you do to stay agile in your role, and how do you find a way to do it consistently?
[00:28:21] Gina Bhawalkar: So I work on the West Coast, um, but I work for a company that’s on the East Coast, so a nice benefit of that is I really have meetings after 2:00 PM Pacific Time. And so one of the things I’ve done is on Fridays from 2:00 to 5:00 PM I focus that time on experimenting with AI tools and how they could help me improve my research workflow, so a lot of the things we talked about today. What are the bottlenecks in my research workflow? Um, where do I feel like I’m not getting to the level, the standard of quality that I hold myself to? And are there ways where I could use Copilot or other tools Forrester has made available to me to be more effective? And so I think just having, like, dedicated time blocks where, you know, I really focus on keeping up to date with all things AI has been super helpful for me.
