Hiring for AI Fluency: Tracy St.Dic on How Zapier Raised the Bar for Every Hire

Show notes

In this episode of the Leaders in Talent podcast, Adriaan Kolff sits down with Tracy St.Dic, Global Head of Talent at Zapier, the fully remote automation company of nearly 1,000 people across 42 countries. Before Zapier she was Senior Vice President of Recruitment at Teach For America, responsible for the largest and most diverse teacher pipeline in the US.

Zapier assesses every hire, for every role, against an AI fluency rubric, now on version two because the internal team upskilled past version one within a year. Tracy breaks down its four components, why she measures the slope and not the snapshot (how fast people learn hard things), and what their AI recruiter found at the top of the funnel: 70 to 80% noise in technical pipelines, hidden gems among the auto-rejected, and candidate opt-in that grew from roughly a third to 81%. Plus where to start if your team is not AI native yet: deterministic automation and protected build time.

Timecodes

01:04 Welcome and Tracy's background

02:17 Zapier in context: 1,000 people, 42 countries, fully remote

04:00 The 2023 "code red" that started AI fluency at Zapier

07:25 Inside the rubric: mindset, strategy, building, accountability

10:38 Four assessment points in the hiring process

11:49 The slope, not the snapshot

15:14 The AI recruiter at the top of the funnel

18:59 What candidates actually think of AI interviews

22:03 The AI surfaces, the recruiter decides

24:15 Rebuilding the hiring process for an AI-enabled world

27:40 What happens to the recruiter role

29:40 Where TA leaders should start: automation first

32:59 Build days: how the team upskills every three weeks

36:00 Where to follow Tracy

___________________________

Connect with us on LinkedIn: https://www.linkedin.com/company/matchr/

Get in touch with us: https://www.matchr.io/who-we-are/contact/

___________________________

Connect with Tracy St.Dic: https://www.linkedin.com/in/tracy-stdic/

Connect with Adriaan Kolff: https://www.linkedin.com/in/adriaankolff/

___________________________

Resources mentioned in this episode:

Zapier's AI fluency rubric (v2), by Tracy: https://zapier.com/blog/raising-ai-fluency-bar-in-hiring/

___________________________

RSS feed: https://media.rss.com/leaders-in-talent/feed.xml

Transcription

[00:01:04] Adriaan: All right, ladies and gentlemen, welcome to another episode of the Leaders in Talent podcast. Today my guest is Tracy St.Dic. She's the Global Head of Talent at Zapier. Tracy, welcome to the podcast.

[00:01:18] Tracy: Thanks so much for having me. I'm excited to be here.

[00:01:21] Adriaan: Tracy, I am personally very interested in our conversation because it's going to be all about AI, and I do feel that Zapier is really at the forefront of what you're doing with AI, especially in the hiring process. The reason why I invited you on the podcast is because I read your article around AI fluency. And what I realized is that this was not your first time launching this framework, it was already updated version two. Something we see across our clients is that people are thinking about this, but you are the first company that at least I have come across that's not on version one, but on version two already. I would love to get to know a little bit more about why you did it, what the successes are, and why you felt the need to already update to the second version. But before we go in, can you give a little bit more context about the company and how big you are, so that our listeners who might not have heard of Zapier understand what you are all about?

[00:02:17] Tracy: Yeah, absolutely. And I'm so excited to be here and excited to represent Zapier. Zapier started as the most connected AI orchestration platform. At our core, we are just a bunch of humans who really think computers should do more work, and that has not changed in this AI era. Zapier is really the governed action layer of how you think about how you do work in your day-to-day. It allows you to do automations, agents, and connect with MCP and SDK, really the whole gamut of how you do work in this new era of AI. I've been at Zapier for a little over four years. I oversee talent acquisition, intelligence, employer branding, partnerships, and internal mobility, and I also lead Zapier's global staffing strategy and AI fluency initiative. So I really try to help shape how the company is building and preparing its workforce for this AI native future. Prior to that, I was the lead and senior vice president of recruitment at Teach For America, which is an education nonprofit in the US responsible for the largest and most diverse teacher pipeline in the country. So very different than tech.

[00:03:23] Adriaan: Very different. Interesting. Hey, and Zapier, you're a fully remote company. How big is the company right now?

[00:03:30] Tracy: We're just under 1,000 people. We're in 42 countries, and we've been fully remote from our founding in 2011, 2012.

[00:03:38] Adriaan: I think that's a topic for another day, but at the same time, it will be interesting when we dive a little bit more into what it's like to be a fully distributed company and the use of AI to help sift through some of the thousands of applicants that you get as a fully remote company.

[00:03:55] Tracy: Yeah, absolutely. I think there's a lot of benefit to being a remote company in this world of AI, which we can get into later.

[00:04:00] Adriaan: Great. So tell me a little bit more about the AI Fluency Framework, why you launched the first version, and why it's important for Zapier that you've introduced this.

[00:04:11] Tracy: Yeah. So let me take you back a little bit in time, to about March 2023. This is about six months after ChatGPT came out, and it became very clear that AI and LLMs were going to really disrupt the industry. Our CEO, Wade Foster, at the time, in that spring 2023, called essentially a code red, and he said, "Hey, this is not only gonna change the way that our product needs to operate in service of our customers, but it also has to change the way we need to operate." And so from there, we started a series of different initiatives inside Zapier to really encourage and weave AI upskilling into our work. It got to the point where, in spring 2025, we started to think about the future of work, the future of Zapier's workforce and hiring, and started to say, "Hey, we really need to put a stake in the ground of what does it mean to be AI enabled? What does it mean to be AI first, and AI fluent even?" I don't think at the time these were words that everyone had a very clear definition of. Maybe you can argue we still don't, but we certainly felt like we needed to put a stake in the ground. So our chief people officer, who is now our AI transformation officer, Brandon Sammut, which is a whole 'nother topic of why I think chief people officers are perfectly positioned to lead an AI transformation, he was really focused on how do we upskill our current workforce to be more AI fluent in order to be able to solve more problems for our customers in different ways. And I said, "Okay, well let me take the front of the funnel. Let me take the hiring, thinking about onboarding. How can we then ensure that everybody we're bringing into Zapier is also meeting our current team where they're at?" And therefore we're meeting in the middle and totally transforming how our workforce looks and how we operate to be a more AI native, AI first company. That was the original reason for the first rubric. So in May, June 2025, we came out with our first version of the AI fluency rubric, and at that point the market and our own usage were in a really different place than we are right now. Using AI was still very early, and at the time being open to AI, being curious, experimenting, that itself put candidates ahead of the curve. But as you can imagine, nine months, almost a year later, that is no longer the case. And so that was part of the impetus of saying, "This is why we need a version two." AI is a critical part of how we get work done at Zapier, and internally, our internal team has just upskilled so much that the bar had already moved internally, and the bar that we were bringing people into Zapier at was way too low. It didn't actually reflect what our teams needed, and we also knew that we wanted to set new hires up for success to come into Zapier ready to go, and be a part of the team and help uplevel us. So there were a lot of reasons why we needed to really rethink and change the bar based on everything we learned over the past year.

[00:07:07] Adriaan: For listeners who might not be aware of what the rubric is or what we're exactly talking about, can you maybe explain in very layman terms what the rubric is and how it works, and maybe off the top of your head, what did version one look like and what changed with version two?

[00:07:25] Tracy: Yes, absolutely. Okay, so there are four components now of our AI fluency rubric, four things that we evaluate throughout the hiring process. Version one had the first three: mindset, strategic acumen, and building skills. And in version two, we added accountability, or accountability and discernment, as the fourth very important pillar. Just to give you a quick rundown of what each of these are: mindset is really focused on your approach and orientation towards AI. Right now in the AI fluency rubric, it's thinking about how are you proactively experimenting with AI, and how are you revisiting what's working as capabilities in AI evolve? Are you leaning in? Are you experimenting? Are you hungry to learn more? And you feel positively that AI is going to change your work. So that is a lot of how people approach AI. Strategy or strategic acumen starts from the outcome and decides deliberately where AI fits, and being able to say what stays human and what are the ways that AI is gonna change my function. So it's really function specific. In hiring, that could look like asking recruiters that I may be hiring, "How do you think AI is gonna change the future of recruiting? Where should you use it and where shouldn't you? How is your job gonna change?" So it's really thinking function specifically about how your role will evolve. Building is actual building skills. That's the more technical component, and that is related to roles specifically, but using tools is by design a very small part of this. So mindset, strategy, and builder skills were the first three components in version one. In version two, adding on accountability just felt more important now, because we all know what AI slop looks like, and we know that we need to avoid it. We don't want people, especially I would say in TA but in any function, to just blindly accept what AI is spitting out at you. We want people who know, again, when you should use it, when you shouldn't, what good looks like, how to iterate and push back on outcomes, and all of that. So those are the four pieces: mindset, strategy, building, and accountability. Those are the four things we test for, and our rubric has a few different levels. We have an unacceptable level, which of course is not meeting our bar. And then we have what we call a capable, an adaptive, and then a transformative level that is very hard to reach. Very briefly, capable is right now being able to demonstrate how you use AI to operate at a meaningfully higher level. Being able to say, "I use it with intention. I use it on a regular basis. It's a part of my workflow, and it has made me more efficient, do higher quality work, or just make my experience as an employee much more enjoyable." Adaptive goes one step further, and those are people who are more orchestrating AI and building systems that elevate how work gets done for themselves or for their team or for the organization, building durable workflows. And then transformative are the people who are really completely redesigning how work happens, reshaping roles, org charts, processes. The work looks fundamentally different than it did six months or a year ago. So that's how it goes together, the four components and the three different levels.

[00:10:38] Adriaan: And then how do you assess that? Because this makes so much sense, and thank you for explaining this so clearly. So how does that look for recruiters? What are some of the questions, or how do you train someone that needs to go through quite a lot of questions in half an hour or 60 minutes?

[00:10:53] Tracy: Great question. So we purposefully do not just assess this at one point in the hiring process. We right now assess it at at least four points. One is the application, where there are questions and we even ask them to share a submission about a workflow they've built or used. Sometimes people will include screenshots or share a video. We wanna see how they actually use AI. So the application is one. The recruiter screen is another one. Then we have the skills assessment, which we have transitioned in version two to being a live, more role specific exercise that's updated to, one, expect AI use of course, but also to see how people iterate and how they actually use it in real time. And then the last one is the executive interview. So we purposefully have an evaluation at the beginning and at the end of our process, because we also care very deeply about the slope and not the snapshot that someone's on.

[00:11:49] Adriaan: Let's zoom in on that word slope. Tell me a little bit more what you mean with slope, because that's a very interesting concept, and I think it's important for you to articulate what you mean and define by slope and how that has changed from version one to version two.

[00:12:02] Tracy: Yeah, absolutely. Before, in version one, I think we asked questions like, "How are you using AI today?" and show us what that looks like. And that made a ton of sense at the time. I think what we've learned is that the technology keeps changing. There is no way, I think, to predict what this technology is gonna enable us to do in six months, let alone a year or two years. And so as we think about hiring for the future, hiring for durable skills that will sustain throughout all of the changes in our dynamic landscape, we're really thinking about, just to boil it down simply, how do people learn hard things fast? I repeat that to my team all the time. And when I am recruiting people or hiring people onto my team, I always wanna know how they learn hard things fast. So specifically with AI, things are changing all the time. I don't wanna just know how you're using AI today, I wanna know how you were using it six months ago, and what have you experimented with? What have you changed? What have you been excited by? What did you try that didn't work? I wanna understand your journey up until this point, and then I also want to understand throughout the hiring process how you may have upleveled or thought about your AI fluency and AI uses even within the hiring process. That gives us a meta signal of what that slope and trajectory looks like. And that, I think, is more important because that helps us see, okay, is this a candidate that's going to keep learning and keep growing and stay really curious as the technology continues to change? That's a different candidate than someone who's used the same three tools for the past three years and never changed.

[00:13:38] Adriaan: It's like a different way of very specifically looking for growth mindset. Is someone willing to learn and iterate and go through that motion. What's interesting is that you said you actually have a skill or a live assessment. Is that run by the recruiter? Is that run by the hiring manager? Who runs that live assessment with AI fluency?

[00:13:58] Tracy: It depends. It depends on the role, honestly, and it depends on what type of role and what type of other skills we're assessing. For example, on our go-to-market roles for sales, it's embedded within the presentation and the technical assessment they do. For our engineers, it's part of a live coding test we do now. And for people team roles, for example, we created a default skills test for non-technical roles. It could be an interviewer who's normally in the loop. In some cases, we have AI automation engineers in a lot of our departments as a role at Zapier, and we have trained them to run this portion of the test, too. So that really depends on the role, but a couple of things are important about it. One, we want to see live iteration as much as possible. This also honestly helps us prevent fraud and cheating. Rather than just say, "Come with a product," we say, "Show us your screen. Use any tools you want. We're gonna walk you through a scenario. Show us how you start building." And then partway through, we give different information and we wanna see how they adapt and change and iterate. So it is both a fraud prevention or cheating prevention technique, but it is also really understanding how they think and operate in real time, 'cause that gives us an understanding of how they're gonna do that here at Zapier.

[00:15:14] Adriaan: Hey, now, we're a fully remote company, so the pros are the world is our oyster. The downside is the minute we put a job out, we get hundreds and sometimes even thousands of applicants, which overwhelms our recruitment team, and especially a company as well known as Zapier has to deal with that. And I know you've implemented an AI recruiter at the beginning of the process. Tell me a little bit more about what your experience is, what that looks like, what's the feedback from candidates?

[00:15:48] Tracy: Yeah, absolutely. So first I'll start with the problem we are trying to solve. Exactly to your point, we're a global remote company. We are a target for fraud, but we also just have generally a lot of noise in our funnel. It's not all fraud with a capital F. Sometimes there's fraud with a lowercase f, which is just embellishment, a lot of AI auto-appliers where we're getting pumped hundreds and thousands of applications that are not serious candidates who definitely wanna work for Zapier. They're just trying to apply to as many roles as possible, which I empathize with in this market. So there's a broad spectrum, but we were just seeing incredible noise at the top of our funnel, specifically for our technical roles, where in our latest audit we're seeing 70 to 80% straight noise, and a lot of that is fraud. All sorts of fraud. So that's the problem we were trying to solve. And it was taking up a ton of time for our recruiters and taking their time away from legitimate, genuine candidates who wanted to work here and who thought we could really be a good fit. So when we started this experiment, and we partner with Ezra AI Labs, which has been an awesome company to work with, they, I believe, just got acquired by Greenhouse, an awesome company to design partner with, we started working with them because we first thought that it was going to be just a volume issue. We needed a way to do more screens, get better information before we move people through the funnel. What we saw, though, what was fascinating, was that, one, we were able to push more people into getting this first touchpoint with an AI screening tool, which allowed us to get more information from people, yes. But it also allowed us to actually get more information from the people we normally would not have been able to meet with just due to sheer capacity. So we really tracked this. We tracked, okay, these are the 5% of people we would've met with anyway 'cause their resumes are great, they look awesome on paper. Here's the maybe 50% of other people we would not have met with just due to capacity alone, we would've rejected right after the application. But we found in our initial experiments, up to 30% of those people actually had amazing qualities that made us wanna put them into the hiring process. So I call these the hidden gems, and so it became less just an efficiency move and also more of a quality move. So that was really important. How we're using AI interviewers now, how we're starting to even push it further in the funnel, is that we're thinking about the AI interview screen as a part of the application. So you apply for a role, fill out some basic kind of knockout questions around eligibility and things like that, and then the very next thing you do is you actually meet with an AI interviewer. It allows us to get a lot more information. They probe really great questions. And so when the recruiters are actually looking at their entire talent pipeline and determining who they meet with and who moves on to the very first stage of the process, they are looking at so much more deep and rich information that we were able to get. And it also works as a fraud deterrent. It also has things that are built in to help detect and flag potential cheating. So it's really helping to quiet a lot of that noise, as well as get us richer information. So we are continuing to experiment with this. But you asked also about the candidate experience.

[00:18:59] Adriaan: Yeah. Tell me, I have so many questions here.

[00:19:01] Tracy: Feel free to cut me off and redirect, because I could talk about this forever. But the candidate experience has been one of the most pleasantly surprising parts. I think when we started this experiment, the reason why we went with Ezra versus the five or six other ones that we tested is because I truly thought it was the best, most sophisticated tool in terms of low voice latency. It felt very natural. The follow-up questions were great. And no AI is perfect, but I thought this was the best in the market. And so we really saw, pleasantly, that candidates really responded to this. In fact, in some cases, candidates preferred to talk to AI before a human because it's a little bit less nerve-wracking. You don't have to worry about facial expressions and someone looking bored on their 10th interview of the day. And we also saw, in terms of data, our candidates were saying, I think it was a 4.5 out of five stars that they rated the experience. Even senior candidates up to senior director level have said that this was a really pleasant experience, and they were surprised at how seamless it was. So I think sentiment is changing. I think at the very beginning we were worried a lot of people wouldn't wanna do this, but we've seen upwards of 80% opt-in rates to our recruiter screen, and we see something like a 97% completion rate. So once people do it, they get into it. And yeah, it's been an incredible tool for us, especially when we hire globally in places like India, for example. There's no way we could meet with all of those people without a tool like this.

[00:20:29] Adriaan: So just for me to wrap my head around, do candidates have the option to choose between an AI interview and an actual recruiter, or does everyone go first through the AI interviewer, and then if positive, speaks to the recruiter?

[00:20:45] Tracy: So we always give an option to opt out, and that's true of any of our AI, that's a legal compliance issue. So we give them an alternate path. If you don't wanna speak to an AI interviewer, which we think is to the candidate's benefit, then we just assess you through our normal pathway. That may mean that you get to speak with a live recruiter. It may not. We will just assess your information, what we have. So the screen really allows people to go above and beyond and give more color to their application or their candidacy. If they don't wanna do that, totally fine. We might ask for some written responses, but we're going to assess with the information we have.

[00:21:25] Adriaan: And you're saying that 80% of the applicants say, "Hey, I'm fine to speak to an AI interviewer"? That's what you've seen?

[00:21:33] Tracy: And that has grown from about 31 to 35% when we first started testing this, to upwards of, I think, 81%. Different roles and levels are a bit different, but we've changed a lot of our messaging to be super transparent on why this is in the benefit of the candidate, how we use this information, why it helps us and helps them. And so I think that, in addition to the fact that sentiment is changing across the industry anyway, people are a lot more comfortable with this.

[00:22:03] Adriaan: And then once a candidate goes through it, is the AI making a ranking or a judgment call, like this candidate gave the right answers, I think a recruiter should speak to this person? Or is it actually the recruiter going through all the transcripts and being able to do that faster to see who gets scheduled?

[00:22:22] Tracy: It's more the latter. The judgment always lies with the recruiter, and that's true in anything we use that is AI-enabled. But the AI recruiter screen is able to surface the insights much more quickly and help summarize and categorize and all of that. And we do a really great job, I think, of ensuring that we're training the AI to use the same consistent questions and to be able to compare answers to that. That's why it's also so good at follow-ups, 'cause it knows exactly what information we're trying to get. And then the recruiters go in and they review all that information. They can click into the video if they want. They can go back and read parts of the transcript. So it depends on how they're working through that workflow, but the recruiter is always the one who is responsible for the decision. As we always say at Zapier, you can delegate the work, but you can't delegate the accountability. And so it's always the human that is responsible.

[00:23:12] Adriaan: What's the feedback from your recruitment team?

[00:23:15] Tracy: They love it. They honestly love it. There's some people on our team that are like, "I could not hire for this role without an AI recruiter." And again, specifically in places where we just wouldn't have the capacity for all of the volume. Hiring in India, where we have an entity, is a really good example of this. We also had a role that a hiring manager wanted to open three days before Christmas, and we were like, "Why do you wanna do that?" And they're like, "Well, I just wanna get this started." And we're like, "Okay, you can do it with the AI interviewer screen." And they set it up. They did it. Over the holiday break, the AI interviewed tons of candidates for them. So when they came back two weeks later, they weren't starting from scratch. They were starting with an incredibly rich pool, and they were able then to move forward towards hires in two or three weeks. So drastically shorter than what we normally would've seen. So overall, I think our team is really embracing it, and I think we're still experimenting with what are the best places to use it, how do we frame it, and what are better ways that we can increase our efficiency and quality in the workflow.

[00:24:15] Adriaan: If you look from your vantage point of view, if you look at the overall recruitment process within Zapier, do you have a vision of where this could go, or some of the things that you're now experimenting with, or where you would like to take this?

[00:24:35] Tracy: This is such a big question, and it's honestly the one I'm obsessed with right now, because I think that the hiring process in general is absolutely due for an overhaul. If you really think about hiring, I think it comes down to three core things. It is talent identification, it is talent evaluation, both actual skills and then fit with the company, and it's the candidate's evaluation of the company, because it's a two-way street. It is only those three things, in my opinion. There's nothing that means you have to have a resume or an application or even an interview. All of those are things that we've created over time to have different filters. It's a big question. I think that the future of recruiting needs to completely change. We think we need to rebuild it in this AI-enabled world, thinking about, not shying away from, the fact that candidates are just as AI-enabled as us, and what does that mean in terms of the signal that we need? I just actually think we as an industry have to figure out this challenge and figure out what the new world is gonna be. It's not just about making everything faster for recruiters, faster resume review, faster application review. We wanna think about how do we make our decisions higher signal, better quality, so we're making stronger, more confident decisions related to quality of hire. And how are we improving the experience for ourselves as people professionals, but also for our candidates and our hiring managers and everybody involved? I think it has to be all three of those things. I see a lot of companies right now focused on how can we cut costs and make TA more efficient. I think that's absolutely the wrong question. I really think that Zapier, to its credit, really has an abundance mindset, where it's like, "How can we actually leverage this technology to do even better human things? To do more of what humans are really good at, and also what they really love." And so I think that's an important mindset that we keep front and center at Zapier.

[00:26:34] Adriaan: And do you think that the role of the recruiter will be significantly smaller and more focused on the high profile roles where there's a lot more human connection needed? Because AI will be so good that the interview will even be, because there's no bias, always available, based on all the data that you can pull in from all the interviews that you've recorded, you create this really smart AI that will soon feel very natural, like how you and I are having this conversation. It will feel, most likely in the next one to two, three years, that natural. And then skip the recruitment assessment. Compliance-wise we're not there yet, legally-wise, but let's hypothetically say that's not blocking us, to then directly speak to the hiring manager. Or do you feel, no, there's always that need for the recruiter to have that conversation, and recruiters will just be more and more enabled, similar to where we are now, with more tools at his or her disposal?

[00:27:40] Tracy: That's such a great question. My take is that we are still going to have recruiters in some form in the future, for sure. I think there is a science and an art to finding people and evaluating them and compelling them to join your company. I think that those who keep that process as human as possible for the experience of the candidate are going to have a real competitive advantage. And so I think that's actually going to become a differentiator. What I do think, though, is that the role of a recruiter as we see it today is drastically going to change. I don't even know if they'll be called recruiters anymore. But I don't necessarily see a narrowing of their responsibilities, though I'm sure you could say they're focusing on more of the high touch, white glove, human to human relationship building. That could be seen as narrowing. I also think it's expanding the role of recruiters. I think they could have a lot more insights to think about talent intelligence and market intelligence, to think about talent branding, to focus on and work with our people business partners on workforce planning. Being more of that talent advisor to hiring managers, understanding the business inside and out, being able to take all of the context of everything that they're looking at and really help guide the process to make better hiring decisions. I still think that hiring is going to be one of the most important things that companies do. Talent strategy is the biggest line item in most cases for most companies. And so I don't think the importance of this is something we would want to delegate to AI. But I do think the actual day to day responsibilities of how recruiters go through their workflow are going to change 100%. And I think we need to break out of this recruiting versus sourcing versus coordinating versus other people roles. I think a lot of that is going to start to blend together, and we're going to be thinking less about specific roles and more about what are the types of work that we need to do in order to bring someone into the company and feel confident they're going to thrive. And I think a lot of that is going to just come together and reshape in a different way.

[00:29:40] Adriaan: For a head of TA or senior people leader that's maybe not as AI native as Zapier is, what would your advice be? Why is this important? Where to start?

[00:29:57] Tracy: I think that's a great question, and I also think there are so many TA leaders out there that for whatever reason, including potentially constraints within their own company, are not able to really dive in to being focused on AI fluency. I would say an AI native team comes down to two things. I think it is the strategic vision of the TA leader, and it is the AI fluency of your own team. And I'll break those down really succinctly, hopefully. As a TA leader, even if you can't use or you don't currently use as much AI in your processes today, where I find that honestly most people need to start is with simple automation. Automation is deterministic. It is more accurate than using agentic workflows. It is something that I think most IT and legal teams can get on board with. So start there. Start to reap the benefits of how simple automation and workflows can actually free up the capacity for your team to do more of the human things. Whenever I hear about a TA team that's moving data from one place to another or doing the same menial task over and over again, it really breaks my heart, because that is not the best use of human ability. So as a TA leader, think about: what do I want my hiring process to look like? What is my vision for what is ideal for my company and for the candidates that we recruit? And then start trying to figure out how to build in simple automations. Zapier is a great tool. There are plenty of other good tools out there. And work with your tech teams to figure that out. That way you're actually able to start solving some of your own problems yourself instead of waiting for a specific tool or a vendor to do that for you. So that's one. TA leaders have to really hold the pen here. The second thing that is related to that is you have to build time for your team to explore the technology. Again, that could be as simple as automation, or most teams I think have, whether it's Microsoft Copilot or Claude or ChatGPT or something, they're starting to have these tools available to them. So how are you upskilling your team on AI fluency? Simply by giving them the time and space to play around with it. I think you will be surprised at how quickly your team can grow in AI fluency and how it will empower them to truly solve their own problems, but also, since they know their work best, figure out the places where they can make their work even better by speeding some things up or making it higher quality. So I think TA leaders have to really think about: where do I see the hiring process at my company going? What's my strategic vision? And then you have to think about how do I start building skills within my team. And if you're not doing that right now, we say at Zapier, that is at this point honestly management malpractice. Because this is the future of work, and if you are not building in the time, or refusing to let your team play with these tools or use these tools, it's really gonna be to their detriment as professionals. Because I can't see a world where recruiting teams will hire a non-AI fluent recruiter. You are then just paying all the salary with half of the output. It's not gonna make sense from a business perspective. And so I really think this is where the future of work is going.

[00:32:59] Adriaan: Before we wrap up, one last question. In your team, how do you motivate people to share what they're doing? How do you motivate people to play around with it?

[00:33:11] Tracy: So much of this is the Zapier culture, honestly. We have a really big culture of experimentation and curiosity and leaning in. But I would say a couple of concrete things for my team. One is that we have really focused at the very beginning on psychological safety. "Hey, it's okay if your experiments don't work out. We're all just learning this together. It's okay if you feel like you're behind or ahead, or whatever it is." Really creating the space to say experimentation in itself is valuable. And over time that will turn into "let's look at the real outcome of this", but experimentation as a starting place is really valuable. The second thing is just building in time. When I see people who are resistant to AI, it's usually not because they're philosophically opposed, it is because they haven't had clear guidelines, they don't know how they can use it, they don't know what tools they can use, they don't know what good looks like, and they don't have the time to experiment. Not only do we wanna make those guidelines really clear, which I think is a good starting point, but we have carved out time every three weeks for my team to have build time, and that looks different depending on which week it is. But it could include just open several hours in a day where you get to experiment, try building some workflows, create automations, et cetera. It could be working in small groups or in partners to do that. So we put our money where our mouth is, we put it on the calendar, we prioritize it, and we give our team guidance on what to do. And then in terms of accountability, at the end of each of those days, they have to come back and share what they've built, and we have to be able to share that and learn from each other, including the things that didn't go well in our build. So again, that helps with psychological safety, it helps with people understanding what they need to do to learn from peers and also push themselves, and it makes it a normal part of our TA operating rhythm.

[00:34:57] Adriaan: Anything cool that has come out of these sessions?

[00:34:59] Tracy: Oh my gosh, so many cool things. I will say, my team is on fire. I did a quick little check-in with people on how they liked the build days, and they love them. They're like, "I would not have this concentrated time if we didn't have the build days." These are really fun. So tons of stuff. We have had people building sourcing tools. We have had people building Chrome extensions to help make things on LinkedIn faster and easier. We have had people building fraud detectors. We have little groups and tiger teams that have worked on things like our AI-powered hiring plan and our AI-powered interview prep and scheduling. The list goes on. So much of this is coming from my team and not top-down, and I think that's the really exciting part.

[00:35:41] Adriaan: Tracy, this is so good. Thank you so much for sharing your wisdom. This is so fascinating, and I love that you're open so that all boats rise with you sharing the knowledge, on this podcast but also through your blog. Is there any particular platform where people can follow you?

[00:36:00] Tracy: Just the normal ones. People can find me on LinkedIn, and yes, we try to share as much as possible on the things that we're doing too, because we get a lot of great learning and feedback from people who are trying different things as well. So happy to connect there.

[00:36:14] Adriaan: Awesome. Tracy, thank you so much.

[00:36:17] Tracy: Thank you. Thanks for having me.