Organizations face a critical AI inflection point: the pressure to implement quickly conflicts with uncertainty about the right approach. While some companies are still cautiously experimenting with basic AI tools, others have fundamentally reimagined how work gets done. What separates the companies achieving remarkable results from those stuck in endless pilots? Often, it’s not the technology itself but the culture and approach to adoption.
So how exactly are companies at the cutting edge getting to the cutting edge?
My guest this week is Brandon Sammut, Chief People Officer at Zapier. Ever since Zapier’s CEO issued a ‘Code Red’ on AI back in March 2023, Zapier has been undergoing an AI transformation across every part of its business. In our conversation, Brandon shares how their AI strategy is boosting both productivity and employee engagement, why and how every new hire must demonstrate AI fluency, and why it is critical to maintain the human touch in talent acquisition and throughout the employee journey.
In the interview, we discuss:
• Competing in an AI-first world
• AI is a tool, not an outcome in itself.
• AI in People Operations
• Increasing personalization and connection in talent acquisition
• Understanding the moments that matter
• Using the human touch as a competitive advantage
• The critical importance of psychological safety and experimentation
• Why culture beats skills
• The four levels of fluency in Zapier’s AI Fluency Framework
• Bias & ethics
• Where is AI transformation heading, and what will the future look like
Recruiting Future helps Talent Acquisition teams drive measurable impact by developing strategic capability in Foresight, Influence, Talent, and Teyou’regy.
If you’re interested in finding out how your TA function measures up in these four critical areas, I’ve created the free FITT for the Future Assessment. It’ll give you personalised insights to help you build strategic clarity and drive greater impact immediately.
Just head to mattalder.me/podcast to complete the assessment—it only takes a few minutes.
Follow this podcast on Apple Podcasts.
Follow this podcast on Spotify.
00:00
Matt Alder
AI isn’t just a technology upgrade, it’s reshaping how work gets done. While many companies are still struggling with pilots and proof of concept, others are already seeing dramatic results. So how do you move from cautious experimentation to real transformation? Just keep listening to find out. Support for this podcast comes from smart recruiters. Are you looking to supercharge your hiring? Meet Winston Smart Recruiter’s AI Powered Companion. I’ve had a demo of Winston. The capabilities are extremely powerful and it’s been crafted to elevate hiring to a whole new level. This AI sidekick goes beyond the usual assistant handling all the time consuming admin work so you can focus on connecting with top talent and making better hiring decisions. From screening candidates to scheduling interviews, Winston manages it all with AI precision, keeping the hiring process fast, smart and effective.
01:05
Matt Alder
Head over to smartrecruiters.com and see how Winston can deliver superhuman results. This is Matt Alder. Welcome to episode 721 of Recruiting Future, the podcast that helps talent acquisition teams drive measurable impact by developing their strategic capability in foresight, influence, talent and technology. If you’re interested in finding out how your TA function measures up in these four critical areas, I’ve created the free Fit for Future assessment. It’ll give you personalized insights to help you build strategic clarity and drive greater impact immediately. Just head over to Mattalder.Me/podcast to complete the assessment. It only takes a few minutes. In this episode, we’re going to be exploring foresight and technology Organizations face a critical AI inflection point. The pressure to implement quickly conflicts with uncertainty about the right approach.
02:28
Matt Alder
While some companies are still cautiously experimenting with basic AI tools, others have fundamentally reimagined how their work gets done. What separates the companies achieving remarkable results from those stuck in endless pilots? Often it’s not the technology itself, but the culture and the approach to adoption. So how exactly are companies at the cutting edge Getting to the cutting edge? My guest this week is Brandon Sammut, Chief People Officer at Zapier. Ever since Zapier’s CEO issued a code red on AI back in March 2023, Zapier has been undergoing an AI transformation in every part of its business. In our conversation, Brandon shares how their AI strategy is boosting both productivity and employee engagement, why and how every new hire must demonstrate AI fluency and why it’s critical to keep the human touch in talent acquisition and throughout the key moments of the employee journey.
03:36
Matt Alder
Hi, Brandon, and welcome back to the podcast.
03:38
Brandon Sammut
Good to be with you again, Matt.
03:40
Matt Alder
Well, it’s a pleasure to have you back on the show. Could we just start? Could you just introduce yourself and tell everyone what you do?
03:49
Brandon Sammut
Of course. I’m Brandon Sammut. I lead the people team at Zapier as its Chief People officer. Zapier is an AI orchestration platform that makes it easy for teams to use any combination of LLM and automation tools that they like to solve important problems and help people do their best work.
04:07
Matt Alder
Fantastic. And unsurprisingly, you’re doing some pretty kind of amazing stuff when it comes to AI within the sort of the operations of the business, which very keen to dig into. But I know that also you’re doing a lot of speaking, a lot of talking to people in terms of sort of AI in general in the space. So where are we at the moment? What does the landscape look like? Are companies implementing AI for transformation particularly well, or are they still experimenting or are they still ignoring it? Where do you think we are?
04:38
Brandon Sammut
I think the snapshot is early days. Early days would be the header. A lot of optimism and also still a lot of questions. So I would say we’re in the first chapter of a story yet to unfold, but when I talk with other Chief people, officers, CEOs and so on, I mean, the. The interest in puzzling through this is effectively like 100% right. I can’t remember the last time I talked with an organization that’s not figuring out what this could mean for them.
05:07
Matt Alder
And tell us what’s happening at your organization, because I know it’s something that you started quite some time ago and you’re sort of perhaps further along than some other organizations with this.
05:15
Brandon Sammut
Well, you know, to the extent we may be further along, it came from a place of total necessity. So it was almost two and a half years ago, maybe three, four months after ChatGPT launched, that it became really clear to all of us at Zapier that AI was going to meaningfully influence how we work to achieve our mission with our customers. Zapier’s mission is to make automation work for everyone. And any of us that have even dabbled with AI can see that AI provides some really fresh, interesting ways to automate tedium so that we can do the things that humans uniquely do best. That’s really what Zapier has always been about. Now we have this whole new way to think about achieving that mission. It was also very important for us to get on Top of it. So it was like opportunity.
06:01
Brandon Sammut
But were very clear, try to be really intellectually honest with ourselves as we got into the spring of 2023, that not only was there a meaningful opportunity for us, but if we didn’t start learning and building with AI, that opportunity can turn into a big headwind for an organization like Zapier. And so In March of 2023, Wade issued Zapier’s first ever actually literal Code Red. That was the title of the blog that Wade published to the company. Code Red. Competing in an AI First World was the name of the call to action. And that’s exactly what it was. It was a call to action. It did not have answers. So for some of the team, it would just be perfectly open with you. Matt, that was not a satisfying post for many members of the Zapier team because it’s.
06:47
Brandon Sammut
It called us into a big challenge and opportunity, but with very little detail on how, like what the answers were, what the path right to greater success might be. When I look back on that two and a quarter years later, I think we did the right thing at the right time. You know, if we had waited for that call to action to be kind of like almost universally popular within the organization, we would have waited way too long. And importantly, I don’t know if we ever would have gotten there, you know, to broad based kind of understanding and support for that.
07:16
Brandon Sammut
Because part of what we’ve seen at Zapier, part of what I’ve certainly seen or heard secondhand from other practitioners and other organizations is part of the way that you build confidence, psychological safety and also start delivering value with AI, is to start learning and building ideally before you have broad support. I think that’s hard to come by without starting to put it into practice.
07:38
Matt Alder
Yeah, I think that’s really interesting because obviously this year’s seen sort of a number of high profile CEO announcements about AI and are companies adopting AI and do this and do that, and that kind of ramp up time in that experimentation is obviously an important part of this. How long did that take? Where have you got to? Where have you got to now? How has that sort of process developed over those two and a half years?
08:03
Brandon Sammut
Well, you know, when we first got started In March of 2023, there were a couple of like foundation components that we snapped into place that still help us today. So one of those, as kind of unsexy as it may sound, AI use guidelines, and you look at that and like, wait, was that really a big ingredient for scaling AI adoption at Zapier you know, it really was. And why a great set of AI use guidelines isn’t just a list of like don’t do’s, although it can certainly include that it effectively creates the boundaries of the sandbox that the team is able to build in. And I think a lot of us have read several really landmark studies that to maximize a team’s creativity, those guardrails are actually really helpful. And in fact the opposite is also true.
08:50
Brandon Sammut
That in the absence of any guardrails, like a kind of a free for all, you know, that can sound like build whatever, but build whatever. Like people intuit that, like, I don’t think that’s actually true. I think there are probably limits to that. And in the absence of knowing what those are, usually what we’ll do is we’ll actually constrain ourselves even further than is needed. And that can actually limit our creativity. And so literally the first thing we did after that Code Reddit March of 2023 is developed our AI use guidelines. And then I’d say the second thing that has really helped the last two years is thinking about AI as a tool, not an outcome unto itself. And this is true of technology in general, right? Like technologies are tools, they’re mediums, they’re ways of getting things done.
09:31
Brandon Sammut
And when we think about AI that way as well, then when we think about like how to build AI fluency in the organization or how to get value through thoughtful use of AI. For the last two years we’ve been embedding, learning about AI, kind of sharing use cases, celebrating success and things we’ve learned through failure into as many of the company’s existing rituals as possible. So that includes things like weaving AI into our quarterly hack weeks, weaving AI fluency into how we onboard folks in the company, and more recently you may have seen into what we look for and every person we bring into the company, which we could talk about later if that’s interesting, and so on. And in doing so, I think it reflects what we want to be true, which is again, AI is a means it is not the end.
10:19
Brandon Sammut
And to minimize the amount of whiplash or change fatigue within the organization, let’s weave AI skill development, use case sharing and so on into so much of how the company already runs.
10:32
Matt Alder
What’s it doing for the business? I mean, what does it look like practically? Where have you got to, I mean, give us examples of some of the use cases or the shifts that have taken place.
10:44
Brandon Sammut
Absolutely. Well, let’s do a walk across the business. So let’s start in Our customer support team, we have 80 customer support folks waking up every day serving millions of customers. Over the last two years, that team has been able to reduce the number of minutes that it takes to successfully respond to a customer support ticket by 50%. And they were doing pretty well before they started on that endeavor. Now they did a variety of things to do that on the kind of staffing side and time zone coverages and handoffs. So there’s some operational things that they tightened up.
11:16
Brandon Sammut
But the single biggest thing that moved the needle was using more AI and automation as particularly to summarize these otherwise really long tickets into something really succinct that helps the customer support rep get to the root of the issue quickly and then solve the customer’s problem faster. So incredible example there. Now that was ambitious. You know, our head of customer support gets in front of the team about 18 months ago, says we’re going to reduce this metric, it’s called average ticket handle time, by at least 50%. We’re going to do it over the next 12 months and then that will become our performance expectation. And importantly, she did not know exactly how the team was going to do that, but she had seen enough of what the team was experimenting with to basically have the proof of concept.
12:00
Brandon Sammut
So I don’t know the precise shape of this, but I understand the general shape of how we can do this. And I feel good therefore about calling us into not like trying to do a thing, but like we will do the thing. And sure enough, not only did Lauren, our VP of customer support and her team achieve that in the 12 months that followed, but some of the other things they did, you know, those interesting things, like those jobs got better, those jobs got more interesting and less tedious. And so in that same 12 month period, not only did they reduce average ticket handle time, but they also increased employee engagement by 20 or 30 points on most of the topics that we measure, which is, you know, for anyone in the, you know, employee engagement, you know, paying attention business is very rare.
12:42
Brandon Sammut
And so it goes to show that some of these really ambitious and challenging kind of missions around using AI to move the needle on like a core company productivity metric can also be done in a way that not only like maintains engagement, but can actually increase it as folks feel more powerful like it was. I think it’s very clear folks were learning very relevant future facing skills and of course their jobs were changing in a way that was just, I think simply more interesting. So customer support is a great case study in our sales and Marketing teams. They’ve built a variety of agents and AI powered workflows to do everything from, you know, helping write and scale social media copy to lead routing. This is a company that are a problem that most companies have. You have a lot of go to market tools.
13:29
Brandon Sammut
It can be hard to take data from all those different places and intelligently route leads to exactly where they need to go. And so the team uses that and has been able to increase effectively the speed to which we can get back with the lead and also in a way that’s like basically purpose built for what that customer is trying to do.
13:48
Matt Alder
Amazing. And what about things like people operations recruiting, the kind of the sort of the talent side of the business.
13:54
Brandon Sammut
In people operations we’ve used AI and automation to automate almost all of our candidate communications but in a way that’s then personalized for the role and what have you. And so zapier and again all credit to our talent acquisition team routinely gets accolades for candidate experience. And you might wonder was that just like they’re spending more hours with each candidate or they just have an unusually large recruiting team for the amount of hiring you’re doing. And it’s effectively none of that. It’s just thoughtful use of automation and AI. And what the work of the talent acquisition team shows us is that using AI and automation thoughtfully can actually increase personalization, sense of belonging, sense of connection.
14:32
Brandon Sammut
In part because sometimes right the alternative to using AI thoughtfully to personalize, you know, follow up from an interview or how to prepare for your next, you know, candidate conversation as a hiring manager. The alternative is either is not necessarily like a really intense amount of personal, human driven personalization. The alternatives are often a template that is not personalized at all, sent by a human. So it still takes human time, but it’s not personalized. It’s like the worst of all worlds or not doing it at all. So you’re just missing steps. I often demo a very simple zapier use case for new hire welcome. So I have an automation personal automation for myself where every time someone’s new hire start date is triggered in our hris I automatically send the new hire a welcome message in Slack.
15:26
Brandon Sammut
And you know, and I, you know, you might look at the Brandon, why are you giving that away? Because then everyone that joins the company is going to know that’s just, that’s just zapier, right? That’s not you. It’s like well it is me. I set that up with the intention of providing like a warm welcome to everyone in the company. And the alternative to me doing it this way is probably not doing it at all. It’s exactly the type of thing like I wish I had time for, but typically never have time for. And now it happens. And it happens routinely, including when I’m on vacation, you know, so I don’t miss anyone. And, you know, the truth is, for most people that join the company, I haven’t really met them yet.
15:57
Brandon Sammut
There’s not a whole lot to personalize at that stage in the employee experience.
16:01
Matt Alder
Yeah, that kind of makes sense. And I guess that if people kind of reply, then you kind of get involved as a human at that point.
16:08
Brandon Sammut
That’s exactly right. That’s exactly right. So that’s as simple as it gets. But one of the things that we’ve learned over the last couple of years is that the complexity of the build is not always correlated with the level of impact, which means that in some cases there are fairly straightforward, low complexity things you can build with automation and AI that actually can have a lot of impact on personal productivity or employee experience. A couple other quick examples, if it’s interesting in terms of how the people team is using AI, we have a variety of AI coaches set up to help folks with things that are typically pretty sticky and that can cause anxiety. So things that are sticky and cause anxiety. So two examples. One, preparing to give constructive feedback. Gosh, I can’t tell you how many times I’ve had to do it.
16:56
Brandon Sammut
Still not comfortable. It’s never supposed to be comfortable. I think that’s kind of the whole point. But I do want to be skillful at it. And when we talked with some of our teammates, where do you go for help when you’re preparing to give constructive feedback? Some of the answers you would get, and they’re very human answers. Or, you know, if it’s really sensitive, Like, I don’t necessarily want to bring someone else in to talk about it because maybe the infinite context is. Is pretty sensitive. I want to keep it to myself or, you know, I’m not ready to tell, you know, my people business partner yet. I want to see if I can handle it on my own. And I’m like, okay, well, how do you think about your alternatives? Like, you’re not going to talk to a human about.
17:31
Brandon Sammut
They’re like, well, I don’t know, I’m just not going to. I will prepare for it myself the best that I can. Today at Zapier, we have an AI feedback coach that helps folks get ready for that hopefully productive but potentially uncomfortable feedback conversation. It’s trained on some very specific examples of what we think extraordinary constructive feedback look like. And it’s also trained on our core feedback framework that we train folks on all the way, starting with the new hire experience. So that’s one. The second is around writing performance reviews. This is another one where it’s like, it’s a moment that matters. Like you want to do it really well, whether you’re doing your self evaluation or you manager writing one for one of your team members.
18:10
Brandon Sammut
You want to be like really strong about how you collect evidence and you want to write it in a way that’s like clear, actionable and to the point. And there too we have an AI coach where you can drop in the performance evidence that you want accounted for and it will use again, very concrete examples of excellent reviews, plus our impact behaviors which are effectively our performance standards that are differentiated by job level. And all you do is provide the name and job level of the person you’re writing about, including yourself. If you’re doing a selfie val, you put in all the performance evidence and it will draft for you something that you know.
18:43
Brandon Sammut
As we’ve talked with folks and done some of the measurement around this, folks feel like it gets them 70 to 80% of the way there in moments in a way that again, those pieces of evidence, you know, they’re not tilted right by the coach in a way that maybe if like something happened more recently, I might have recency bias and anchor really hard on that thing in my review. Well, the AI coach is trained to balance things independent of when they happen and so on. So those are two concrete examples. On the people team.
19:13
Matt Alder
Is your TA function fit for the future? Over the last 10 years, I’ve done podcast interviews with hundreds of the world’s most successful TA leaders and discovered exactly what they all have in common. Four critical capabilities that drive their strategic impact, foresight, influence, talent and technology. I’ve created a free assessment specifically designed for busy TA leaders like you to benchmark your capabilities against these four essential areas. You’ll receive clear insights into your current strengths and opportunities with practical guidance on enhancing your strategic impact. The assessment is completely free, takes five minutes to complete, and you get your results instantly. Benchmark your talent acquisition capability right now and start proactively shaping your future. Visit Mattalder.Me/podcast. That’s Mattalder.Me/podcast.
20:21
Matt Alder
What’s interesting there is, I suppose, what you’ve learned about that kind of balance between humans and machines. We’re constantly sort of hearing about this unpersonalized experiences Or AI is taking everyone’s job or all of this kind of stuff. And I think that you make that really interesting point about actually if AI didn’t do this, then it might not happen at all. What does that balance look like? Are there sort of particular things where you found that, you know, having humans very much in the loop is critical. What sort of learnings have you had on that human robot split, as it were?
20:52
Brandon Sammut
Yeah, there are two areas where we feel really good about keeping humans in the loop. And the first is moments in the employee experience where that human touch is just simply valued by the individual. And I’m intentionally saying valued by the individual not necessarily produces a better outcome. We’ll get to that in a minute. But in some cases, when it comes to employee experience and people feeling connected to their company, having trust in management, for example, the human touch does produce a tangible benefit. It can inspire trust in ways that it’s a bodied format does not always. And so candidate experience when we’re hiring is an example. We have not pulled out all the human touches throughout.
21:38
Brandon Sammut
We believe that for teams that can field a recruiting organization that can still provide some of these really human, compelling touch points along the way will be a distinguisher. Right. There are definitely going to be organizations who either for economic reasons or whatever the case, different strategy may AI away most of those touch points in the recruiting experience. And in some cases that may be needed and it may actually be the right thing to do. On balance, for the type of hiring that we do, which tends to be like very selective, very bespoke, we’re trying to attract talent that has inevitably many other choices. We think that human in the loop is a differentiator for us. Now the second vein where we keep humans in the loop around employment decisions. So we’re not turning over employment decisions of any kind to AI.
22:26
Brandon Sammut
That includes who we hire, who we promote, who gets paid what, who gets what performance rating, and so on. They’re definitely. We’re going to see organizations trying to use AI to do exactly those things. We’re not there or there yet at zapier. But what we have seen is that AI can help humans make more informed decisions and also check for various forms of bias. So AI is a helper for employment decisions. Sure. Like that’s an interesting opportunity. But making those decisions, we’re not there. Yeah.
22:58
Matt Alder
Again, kind of really interesting. And I think with that kind of first set that sort of moments that matter, that things that are kind of really human, it’s almost like particularly in the recruitment process. The feeling that you have a recruiter kind of advocating for you in the process, even if they’re not actually doing anything. But there’s a human there who is kind of in your corner, I suppose, is a big thing for people. A bit like not wanting to get on a plane that’s completely flown by autopilot, even if it could be.
23:23
Brandon Sammut
I think that’s a good way to think about it.
23:26
Matt Alder
Just in terms of the employee journey.
23:28
Matt Alder
Through all of this. One of the things that has struck me a lot of other companies that talk about AI is this kind of almost demand that you will be AI literate. And this is what we’re going to do without anything to kind of back that up. And this can be really complicated stuff. So you sort of touched on this already. But, you know, what sort of support are you giving employees to make sure that they have the information they need and they know how to use these tools in the right way?
23:54
Brandon Sammut
You know, I’ll share something that may sound counterintuitive or even provocative, which is because sometimes folks ask, like, what’s your curriculum for training folks on how to use AI at Zapier? And we do not have one. We do not have one in part because the technology and kind of like the best practice for putting it to work for you is shifting so fast that the idea is, like, by the time we, like, write, like a traditional curriculum, like, it is already outdated. And so we’ve taken a different tact. I mean, we’re still very interested in helping folks feel fluent, confident, and also folks have the opportunity to ask questions. This is the type of learning where how much we know at any given point in time is much less important than our rate of learning.
24:35
Brandon Sammut
And then you think about, okay, well, what maximizes the rate of learning within an organization? And then you start thinking about things like sense of psychological safety, a culture that incentivizes experimentation, a culture that celebrates sharing what you’re learning so that others can learn with you. So there’s a multiplier effect within there. You have really good feedback loops, by the way, particularly when something doesn’t work the way you thought it would. It’s some of the best learnings, but you think about those are organizational culture ingredients that are not present or at least not uniformly present in all organizations. And, wow, do they really matter if you’re trying to build genuine AI fluency? Genuine AI fluency. And so we’ve spent, I think, as much time on making sure those cultural components are in place as we have on more traditional Forms of skilling.
25:30
Brandon Sammut
However, a couple of the specific rituals that are working for us in terms of helping folks learn together. One, just really. And again, this is a place where clarity helps. So we try to be super clear about where do you ask a question about AI. And we have a couple Slack channels, just two. They each have a complimentary purpose. And importantly, we have a few folks at Zapier who are some of our AI experts and they effectively sit on top of those channels and make sure that when you ask a question about, should I use this model or I ran into something I was building through this error and I tried to troubleshoot, can’t figure it out. Can you help me? That folks get quick answers to those questions.
26:12
Brandon Sammut
And similarly, we have a place where folks share things that they’ve built, what kind of value they were trying to unlock, how they approached it, the results they got, and what they learned along the way.
26:21
Matt Alder
That’s cool. And I suppose kind of referencing back to those guardrails, Obviously, things are changing all the time with AI. We tend to have a week of positive headlines and then a week of negative headlines about everything from bias to hallucinations or things like that. How do you kind of make sure that what’s being done is current, but also isn’t introducing bias or things that shouldn’t be happening around us?
26:47
Brandon Sammut
Yeah, isn’t that interesting? Because it’s one of those things where several things can be true at the same time. And so I’ll read in a single feed, for example, AI makes you dumb. And then you scroll for a minute and it’s like, AI supercharges human intelligence. And then you read the articles and you’re like, well, it appears that both of those things may be true. And what it depends on is how you’re using it and what you’re using it for, which makes a lot of sense. Just like a lot of tools, right? It’s not an outcome, it’s a tool. And so, you know, I can use a. I can use a band saw to do certain jobs and get a brilliant result.
27:23
Brandon Sammut
I can use the same band saw to do a different job or use it in a different way and get a terrible result. And then on the bias side, incredibly important topic. Maggie, one of our leaders here on the People team is doing a bunch of research with us because there too, what we’re seeing pretty clearly, it’s all in the details, right? There are ways or jobs you can use AI for which could introduce bias and lead to worse outcomes or worse decisions. You can use AI in ways that help check for bias and help us make better decisions. And so the thing we’re using a very like. Well, how do we.
28:00
Brandon Sammut
What is a framework or set of principles, by the way, are also diagnostics on the things we’re building that we can run to make sure that they’re maximizing the potential benefits of putting AI in the loop and minimizing some of the risks. I think for many of things, these are both opportunities and risks. Do we understand what those two buckets are? And then how are we testing and training? Similar to what we would do for an employee. If we give a big new job to someone who’s never done it before, we’d probably do a whole period of monitoring and testing before we set them loose at scale.
28:33
Matt Alder
And I know from. I suppose from a hiring perspective and a skills perspective, you’ve been doing work to look at the. Almost like the levels of AI fluency people need within. Within the organization for like, different tasks and different roles and things like that. Talk us through that a little bit.
28:49
Brandon Sammut
You bet. You know, the interesting thing is this AI fluency framework that Zapier now uses was built. The first use case for it was actually external. In evaluating candidates, we roughly 90 days ago said, you know, every single person we hire from this point forward must be fluent with AI to some degree. And the question was like, well, what are the degrees? And so for that, you know, to answer to that, we created Zapier’s AI fluency framework. And the framework has effectively four levels, the first of which is unacceptable. So you can get hired at Zapier at the first level. I think. I think that level is. I don’t think. I know. That level is literally called unacceptable. And that would include, you know, someone who is either actively resistant to the use of AI for whatever the reason may be.
29:31
Brandon Sammut
I can respect that. Depending on the reasons. That’s not a great fit for what we’re trying to do at our company. And we try to be really upfront about that. Yeah. Or who just hasn’t done any kind of legwork themselves to kind of get their heads around what the opportunity might be. They do not have to be, by the way, the next level. So after unacceptable is capable, and capable being sick. You have a working sense of prompting or what you could use AI for, but you haven’t necessarily repeatedly actually used it to increase productivity in your role as it exists today. So that’s capable. So you’ve got all the promise. A lot of the foundational ingredients but haven’t yet started producing impact. The third level is called adoptive.
30:12
Brandon Sammut
So you have in fact adopted AI to produce increased productivity or quality in the work you do as the job exists today. And then the fourth and kind of the mountaintop, which I think in very few cases would we say we are there at the individual or team level, but that’s why we have it, because it calls us to something greater, is called transformative and that means exactly what it says. So the difference between adoptive and transformative is that with adoptive you are producing value, measurable value with AI, but kind of within the way that the job gets done today. You haven’t redesigned it at the transformative level. You have effectively reconceived of how the work should get done from first principles with the technology that’s now available.
30:53
Brandon Sammut
And that could lead to a pretty divergent way of thinking about doing the work from the ground up.
30:59
Matt Alder
Do you think that any organization can do this? Because obviously your business automation AI is very much in the DNA of the company, of the business and what the business does. Do you think we’re at the stage though where any company could kind of adopt AI for transformation in the same kind of way?
31:18
Brandon Sammut
Yes, but I think there are two things. One, like we talked about earlier Matt, the cultural ingredients for strong AI adoption and creating real value that lasts over time as you hire new people. Those are just as important as all the skills. And so for organizations I would recommend you do a health check. Do we have strong sense of psychological safety, strong trusted management, strong culture of experimentation? Do you think there’s another ingredient that matters culturally? Baseline it and make sure it’s fairly good or that you have a path to making it quite good because those will really be wind at your back as you do the actual puzzling through like how to use AI, building it, testing it and then scaling it.
31:59
Brandon Sammut
The second piece, and I think we’re about to say you can see proved out by what other companies are doing, which is that most organizations will not be able to puzzle through this on their own, which is why you see every major consulting company like huge revenue line for them, Even the last 12 months, just explosive growth in demand for advisory services. And you even see now some of the LLM make themselves starting to off build out these consulting teams. So OpenAI, most notably over the last couple of weeks they’re hiring a ton of, you know, what they call forward deploy engineers. Well what’s a forward deployed engineer?
32:35
Brandon Sammut
It’s basically a consultant that also has the builder know how they can Help you figure out where to get value and then they can actually train your kind of teach your team how to fish on how to actually build and maintain the things that you develop.
32:46
Matt Alder
Final question for you. Where’s this taking us? We’re talking about, you know, we talked about experimentation, we talked about adoption, you know, we talked about transformation. What is that transformation? What are the bigger opportunities? You know, what do you think, for example, people operations or talent acquisition might look like in a few years time, if indeed they exist in a few years time.
33:06
Brandon Sammut
I will tell you, Matt, in all humility, I can make no claims about the macroeconomic impact of AI over the next three years. I think it’s really hard to forecast. I think there will be some material changes. I couldn’t say what they are, but it does feel like some major things will transpire. But if I zoom in on an organization like Zapier, which still plans to have plenty of humans involved in the work of talent acquisition, talent development and so on, what I would like to be able to say is true is the following.
33:37
Brandon Sammut
One, where we have traditionally had limitations like the number of people we can provide coaches to, or the number of people can receive this really intensive form of development, or the number of candidates we can give really specific feedback to when we don’t hire them, or the level of engagement, ongoing connection we can have to our alumni, or how effectively we can identify, you know, the future executives of this company, I would like to say we are doing that not just like 2 to 10% better, but like 2 to 10 times better than we are today. And that’s kind of the difference between kind of like marginal improvements and multiple level improvements. So I think that opportunity is on the table for some of these use cases over the coming years.
34:28
Brandon Sammut
The other thing I’d like to be able to say though is that for all the humans on Zapier’s people team that continue to do that work, that those are meaningfully more interesting jobs. And of course, hopefully by definition they’ll also be meaningfully more impactful because what a single person doing that work can now do. So I hope the jobs are more engaging, more interesting, and more impactful in ways that feel actually like that much more human, like that much more life giving than they are today. And then the other thing, if I were to kind of pull that thread a little bit further, is that it’ll be really interesting to see what management teams do as teams start delivering on this materially greater productivity. So for example, on Zapier’s customer support team coming all the way back to that example from earlier.
35:13
Brandon Sammut
You know, that’s a team of elite performers at this point. They are producing outcomes that are fairly rare in the world of customer support and they are paid in kind. So customer support person at Zapier is paid top 10 percentile of the market and so there’s a nice alignment there. I would, I would like to. I would. I hope, I hope that management teams will think about how to share in the share the wealth of the incremental productivity driven by teams rowing really hard to think creatively and redesign work through AI.
35:46
Matt Alder
Brandon, thank you very much for talking to me.
35:49
Brandon Sammut
Hey, it’s a pleasure, Matt. Good to be with you again.
35:52
Matt Alder
My thanks to Brandon. If you haven’t already, you can benchmark your talent acquisition capability quickly and easily by completing the free Fit for the Future assessment. Just head over toMattalder.Me/podcast. It only takes a few minutes and you’ll receive valuable insights right away. You can follow this podcast on Apple Podcasts, on Spotify or wherever you listen to your podcasts. You can also search through all the past episodes at recruitingfuture.com where you can also subscribe to our weekly newsletter, Recruiting Future Feast and get the inside track on everything that’s coming up on the show. Thanks very much for listening. I’ll be back next time and I hope you’ll join me.
