Subscribe
Share
Share
Embed
The WorkOps Podcast is your weekly conversation with HR leaders and People Ops practitioners doing the real work.
In every episode we dig into one story. A process that went sideways, a system that just didn't work, and what someone actually did about it. Packed with practical lessons you'll want to bring back to your team. Whether you're supporting 500 employees or 5,000, this is how the best People leaders are building for what comes next.
Jeet (00:01.152)
All right. Hey everybody. Today I am joined by Tony Castellanos, who's the EVP of People at Nextdoor. Tony, thanks so much for joining us today. Before we dive into things, can you tell us a little bit more about yourself and how did you choose HR?
Tony Castellanos (00:18.158)
Yeah, well first off, thanks so much for having me. It's great to be here. Excited to get into this today. How we got into HR. I started in recruiting and in my lifetime I've only met two intentional recruiters. I was not one of them. Coming out of college, you know, I interviewed in a lot of places and one of the things that defined my upbringing was really participation in team sports. So I've always enjoyed how do get a group of people around together on a common goal in service of trying to achieve something greater than themselves, right? And one of the things that I liked about one of the places I
was it was a recruiting agency and this concept of how do we get talented people together to solve business problems was one that I found really interesting. You know, it helped that they were all 22 just like me and everybody was in the work hard play hard mentality. But you working there...
You really got much greater insight into filling a job wasn't just about like pattern matching on skill sets and the problems, right? There was this whole host of variables and dynamics that all these leaders are considering that are much greater than just that individual role, right? Cause all these people are on growth journeys together. How do I think about bringing this group together in service of this thing? But,
make sure that they're also equipped for the long haul and they're gonna drive to success over time. And so that, it was just always really intriguing. And the longer that I've been in the people function, the more layers of the onion I've found to peel back and people are endlessly entertaining, interesting, frustrating, all the things, you name it. So it's been a fun ride.
Jeet (01:40.728)
Thank
I like that. It sounds like the core of that is people and then curiosity and maybe onions, which I hope you're into. Nice. Well, we're going to talk about a dysfunctional process or story that you've encountered in the past.
Tony Castellanos (01:56.142)
Absolutely.
Jeet (02:08.908)
Are you able to share the name of the company? If not, let's talk about the type of company that this story is from.
Tony Castellanos (02:16.172)
Yeah, well, I think it's a story about for me, it stems from working at larger organizations and it's really rooted in the evolution of HR and people practices. let me start with, know, HR in its conception really began as a protect the employer and be compliant function. Right. And so there was a lot that was built around policy, regulation, rules, all these things that are incredibly important to especially, I mean, we are a public company.
Jeet (02:38.116)
No.
Tony Castellanos (02:46.166)
and so you want to be considerate of all the things that you need to comply with. But those design principles really embraced control over everything else.
And so, you know, as I've thought about it, one of the things that we've seen as people functions have evolved, right, and particularly over last two decades, as you've seen sort of the war for talent and what is a strategic people function look like and how do they elevate the capabilities of the company. It shifted from this policy management function to one that is all about elevating the company and thinking about how do we amass a talent pool that's going to increasingly improve our service to our customers.
And so the lenses that we apply and the skills that we build in people teams look very different from its origin. But this notion of how do we work that is control and rules based hasn't kept in step with that. And so the stories that I have around dysfunctional people processes are the ones that are really rooted in legacy thinking because
I think there's this real opportunity and we can talk about how AI plays into this, but to embrace this notion of context over control in this function. Now you still need some of the controls for sure, but one of the things that I've found is that.
When you think about your own information architecture and how much you as a people team hold and how siloed that can be in an organization, it really prevents you and your partners from being maximally effective. And so I want to focus on how do we break some of that down so that we can go faster through shared context. And the place that I experienced this most, and I think every recruiter has this story, right, is around offer processes. How do we put together an offer? How do we approve an offer? How do the exceptions work? Like, when does it make sense?
Jeet (04:32.194)
Yeah.
Tony Castellanos (04:36.899)
all those things, so that's the one I think I would get into.
Jeet (04:40.255)
Nice, I love it. I want to spend a little bit of time on the fact that you mentioned is the people team and you mentioned HR and people. Do you see a difference in the wording between HR and people or is that just a natural flip for you?
Tony Castellanos (04:56.801)
No, it's quite intentional.
This is something I credit a lot to Brian Power, who I've worked with for many years. This notion that humans as resources is sort of like humans as bits and bytes that we throw around and we use to accomplish these tasks. Even though human isn't the title, it loses all sense of humanity. And so people is really important because people are so much more than the tasks that they accomplish at work or the things that they move forward. Everybody has a very complicated, intricate life
Jeet (05:16.183)
Yeah.
Tony Castellanos (05:28.441)
with their own and I think denying that really is a disservice to every employer. This is a place where you want to be able to embrace all of the qualities that people bring. Now there are some topics that of course that are not work appropriate that you want to keep at the door but
Jeet (05:33.763)
Hmm.
Jeet (05:42.945)
Yeah.
Tony Castellanos (05:43.49)
To pretend as though all that matters is the work that somebody's doing it well, to pretend as though all that matters is the work that somebody is doing in the hours that they're dedicated to your job is naive, right? And I think you embrace all that people bring.
Jeet (05:57.548)
Yeah.
I think that's, yeah, it's such an important shift. It also means though, that the people function as remit has just grown and grown, right? So now you're having to manage a lot more than previously. How do you make time for that when you still have the same hours in the day, if not less, to be able to achieve what you do? Because with that simple word change, suddenly the responsibility and remit has grown and it's continuing to grow.
Tony Castellanos (06:30.369)
More layers of the onion, know, a lot of fun.
Jeet (06:31.882)
Yeah.
Tony Castellanos (06:33.685)
No, it's true. Like I think that you end up having to squeeze more and more in. And this is also one that leads into getting to better context over control, avoiding bottlenecks is key. And you have to hire great people. You have to hire people who you said curiosity earlier. And I think that's a great word to use. People who want to understand how their work flows into not just what is the people set of goals, but what are our partners goals and what are the organizational objectives that we're trying to achieve. Right. When you have that shared mentality and perspective,
Jeet (06:44.419)
you
Tony Castellanos (07:03.599)
you end up scaling your people capability much faster and more deeply than you ever could. So you do have to add some additional resources, I'll even use that word, but I think that one of the benefits that you get is that the work is more engaging and fulfilling. And so it doesn't feel like more tasks, it's more than you're accomplishing.
Jeet (07:12.483)
Thank you.
Jeet (07:21.347)
you
Jeet (07:25.547)
Yeah, I like that. So let's talk about the process, the offer process itself. Can you give me an example where it didn't quite work out?
Tony Castellanos (07:35.458)
Yeah, well, you know.
I think most companies have a very similar offer process, right? You've got the recruiter is going to figure out what the candidate wants or what it's going to take to close them. They propose a package that flows then to your compensation team who's looking at the market data and the internal equity stuff and saying, you know, does this or does this not make sense? And then there's some escalation path when it's out of band that is to the VP or CFO or CHRO. And then that flows back over to the business leader or the business executive who's actually hiring the person who wants to object. If you said no, that flows back.
to your CHRO. And so the decision tree looks very similar in many companies. And I think that through an academic lens, right, when you look at that diagram, you're this makes sense. There are checks and balances in the system that keep everybody on the same page and aligned to the same goals. In practice, though, it's actually pretty chaotic, right? You have
A recruiter saying, we're going to lose this person if we don't go up. You have the comp person saying, hey, we have all these internal equity problems if we do this. Plus, this is way out of band. Makes no sense within the market norms. You've got a finance leader saying, I don't want to spend that much money. Although there are a few who are understanding that it's a worthwhile investment from time to time. You have your CHRO who's getting pinged from every direction on all these things, including the business leader who's just saying, just make it happen. Right? So you're...
Jeet (08:49.987)
Yeah.
Jeet (08:57.237)
Yeah.
Tony Castellanos (08:59.233)
finely printed blueprint doesn't actually result in a cohesive process. It's actually one where you have all these independent lenses being applied and you have these pieces of information that sit in different parties and nobody's taking a collective look at that. And so, it's not really a compensation process. It's conflict resolution, but it doesn't get resolved in a fulfilling way.
Jeet (09:17.014)
Yeah.
Jeet (09:27.703)
Yeah.
Tony Castellanos (09:28.621)
And there's a lot that we've been trained that is like this tension is really healthy, right? It's because we make sure that all of the opinions flow in, but we do that serially rather than collectively. the fact that people, that there's not a single entity or this team that is responsible for delivering this doesn't have the cohesive picture, I think is a real myth.
Jeet (09:40.194)
Hmm.
Jeet (09:52.324)
That's interesting because I think what you're leaning into here is this is not a tooling problem. It is, it's something else. So I guess I'd be kind of interested to hear how have you solved this previously, or is this one of those things that's just really, really hard to solve and something that you're working on?
Tony Castellanos (10:16.567)
think it's really hard to solve, I think we've started down a good path. And in your point, tooling is not going to solve this. This is a human and operations problem. the things that we have done is we've kept the ranges, the approval thresholds, the checks on internal equity, all that stuff, right? But we've...
Jeet (10:28.739)
Mm-hmm.
Tony Castellanos (10:37.259)
we've changed the way that all that information flows into the decision making system. So instead of having this be, you know, TA to comp, da da da da da following the chain, what we want to do is bring all of that into the very beginning of the process. So, you know, the recruiter is bringing our market realities, the candidate realities, their context to the table. Our compensation team is bringing what are the internal equity benchmarks? How do we think about the external market data? What is the long-term cost of this hire?
You have the hiring executive or business leader who's bringing the perspective of what is the complexity level of the problem that we're solving? What is the appropriate investment that we want to make in moving this forward? And so by having that conversation early on, instead of having these isolated questions like, can we break the band for this person or is this person going to be okay to be paid more than Jill or Jim? What you have now is really the experts in the room on
Is this an investment that we want to make in this person given all the reverberating effects that it will have? Whether that's a new precedent that we're setting in compensation, whether there are new internal pay dynamics that we're going to have to manage, because ultimately, everybody in that room is responsible for some portion of this and is going to have to deal with the consequences. So when you have that full picture, people can make much more informed decisions and it's really nice.
Jeet (11:55.853)
Yeah.
Tony Castellanos (12:04.725)
to see people feel empowered and to adopt points of view that they wouldn't naturally have, right?
Jeet (12:05.312)
Ha ha.
Jeet (12:13.451)
Yeah. And I think that's interesting because I think what you're saying is almost, it becomes naturally scalable because the more upfront information and context that we have, it means that the next time a similar scenario comes up, that context is already there and that you have to ask those questions and there's familiarity and alignment upfront. Do I have that right? Or are you seeing that you're having to do these things over and over again at the same depth and at the same level of detail for roles that come up?
Tony Castellanos (12:43.245)
Well, I think the practice and the repetition helps. But one of the things that I've seen take shape, right, and one of the things I've experienced in my professional life is I vividly recall as a recruiter, you go to a candidate, they give you their comp expectations, you go to the compensation team, and I remember the fight. I remember threatening, like, I'm gonna mark this as a decline for compensation, which is gonna mean that it becomes your problem in our reviews and everything if you don't give me this much money.
Jeet (12:57.411)
Thank
Jeet (13:01.719)
haha
Jeet (13:08.386)
Yeah.
Tony Castellanos (13:11.839)
And it's crazy, right? You're kind of like fighting with one another versus, know, one of the things, and these are the moments that I'm most proud of our team, right? Is when our recruiters, our talent acquisition team, they stop fighting for every single dollar to close because closing at all costs actually has very negative consequences. And they see that, right? And the compensation team,
They stop saying no all the time because they understand more around the business context. And then our leadership team gets to own the trade-offs between those things, right? Because they're ultimately the ones that have to live with the decisions. And I'll say, the person, we have a leader of our talent acquisition function, his name is Ed Ventura. He really embraces this. It's something where he wants consideration of what does it mean to hire this person?
holistically, right? And it's not just what are they going to do for the team or what is their growth trajectory, but ultimately like every piece that you add to a team and all the comp dynamics and everything else, their leveling, their authority, their span of control, whatever it is, it affects everybody else. And it...
For me, it's always the best sign that you've built a healthy partnership and healthy perspective when not just your head of TA who sees all this, but when your recruiters start to really embrace this concept of, know what, this hire is not worth it for whatever reason, right? And I love it when that extends beyond comp dynamics because like it's easier to say yes or no on the money because it's binary or it's fixed, right? It's quantifiable is really what I should say versus
Jeet (14:36.278)
Yeah.
Jeet (14:51.107)
Yeah.
Tony Castellanos (14:51.757)
You know, it's also, it's great to see it when they say no to hire because of how that person or the character will impact the team. And the reason I love that so much is because it's the person moving, it's the recruiter, mean, moving beyond their near term goals. Cause everybody's incentivized to close and to hire. That's literally the core of their job. They're putting on the perspective or putting on the hat and adopting the perspective of my real job is to create long-term value in this organization.
Jeet (15:10.147)
Yeah.
Tony Castellanos (15:20.301)
and this piece maybe doesn't fit that bill. And so you from HR to people.
Jeet (15:20.515)
Hmm.
That is.
Jeet (15:27.899)
that is beautifully put. I like that a lot. And some of the language you're using is almost product management language where you're doing discovery upfront and it's not a waterfall anymore and you're building consensus and then you're allowing the right people to make trade-offs that then feed into the decision-making process. that seems like a really good way to advance this.
And it's likely that you probably didn't even need too much tech. This was more about context sharing upfront that you were saying. I'm interested to hear what happens when there is a deadlock, like who makes that final decision of we're going to go ahead or we're not going to go ahead with this particular candidate.
Tony Castellanos (16:07.809)
Yeah, where we've tilted is really toward allowing the executive to run the team as they see fit, because ultimately they're the ones that are on the hook and they have to deal with this day to day. And I think that as soon as you become the people team that is just the blanket no, right, then all of a sudden you are disempowering the people who are supposed to be leading your organization. Now,
I would say that it's very rare that we get there because the times that I or our people team would actually say no, it's for a very good reason and that's usually something that we all agree on. there it is. I actually can't think of any recent examples where we've reached that deadlock. And maybe that's more of a cultural thing where I think the disagree and commit is a real part of how we operate in this realm. But it's very rare that we disagree when it's a highly consequential decision.
Jeet (16:46.797)
Yeah.
Jeet (16:59.457)
Nice. So it sounds like you're not having to threaten the compensation team anymore and market down to compensation issues.
Tony Castellanos (17:05.363)
I mean it just goes back to when everybody gets it, when everybody understands the point of view, there's not a decision you're being handed, you're being informed, it feels very different.
Jeet (17:13.869)
Hmm. Yeah. Yeah. And it also makes it a big difference for the, for the candidate themselves, right? And when they're on the other side and they join as an employee and maybe they're part of the process, they get to kind of see what's happening behind the scenes. And it's also likely pretty empowering. And I bet your candidate experience scores are also above average because of this very reason, because of the way you guys work.
Tony Castellanos (17:38.094)
I think the alignment definitely helps, It makes it feel like you're coming into a team in a community rather than just you were hired. And then on the employee side, the thing that I've found that it helps most is trust. People trust that they're being treated fairly and equitably because that's how we do this. And it's on all things on the people end. mean, think comp is the one that hits closest to home for people, but...
We really try to embrace this concept with all the benefits that we offer, with the programs that we run, with the reviews that we conduct, all this kind of thing. It fosters a much greater sense of belief in our team when they know that there is shared context and it's not just rules-based decision-making with exceptions as we see fit. That doesn't serve anyone.
Jeet (18:27.011)
Yeah, nice. And in this, in this new age of AI agents, I got to ask the question, are you using AI in kind of this process? Cause you mentioned context and
AI's large language models are really good at holding context and sharing that context and making sure that we almost have an SOP that people can access and they can query and speak with. We've also heard other HR leaders who have created a digital twin of themselves and they've gone on maternity or paternity leave and they've come back and they're like, hey guys, you aren't asking me any more questions, why not? It's like, well, we have the spot and I know how you're going to answer these things. So I'm wondering, like, is there a space for...
Tony Castellanos (19:06.477)
You
Jeet (19:11.651)
accelerating or scaling this process or bringing alignment even further with AI? Or do you think this is one of those areas where actually it's not necessary?
Tony Castellanos (19:20.705)
I think there's a ton of opportunity because of what you said. think housing in the context is really important. I don't want to have to bring humans together on this all the time when the underlying information set doesn't change all that often.
You know, there's a lot that goes into this, right? And actually we have a head of PeopleOps who is, her name is Shandy Ortiz, who is amazing at systems thinking and designing our information architecture. And ultimately, like we are investing in those foundations right now. You know, I think we have built small AI services on top that surface, know, how, sorry, let me start again. We have built small AI services on top that better contextualize some of the compensation recommendations and what have you. But I'd say we're earlier in that game.
precisely because we want to make sure that our infrastructure is rock solid. You know, this is one where
It's actually fundamentally changed how we work, right? We're no longer spending our time trying to surface this. We're trying to make sure that the underlying database, dataset information cohort that you have is really well structured and well designed in a way that AI can parse and then deliver in a really effective way. I think AI can expedite and the on-demand nature of it to the point around the virtual twin there, right? If you don't have to wait, that's always an improvement.
Jeet (20:16.227)
Hmm.
Jeet (20:29.699)
Yeah.
Jeet (20:36.387)
you
Thank
Tony Castellanos (20:40.079)
But ultimately, think when there is any level of tension, disagreement, you're going to want humans to interpret, right? You don't want it to do all the interpretation for you.
Jeet (20:48.151)
Yes.
Totally. the investing in foundations, what we hear sometimes from people teams is that, I don't even have time to invest in the foundations. Like we're being so reactive, there's so much to do, we're already under-resourced. Any advice for those teams on how you are carving out that time to build those foundations while still moving the business forward?
Tony Castellanos (21:13.773)
I count myself as extremely fortunate that I get to work with Shandy on this. She will not allow anything else to happen. But I think she embraces very long-term thinking. I do agree. There's a lot of pressure for what can you deliver right now? What gains? What efficiency can you show? And I think over last two years, everybody has done the same route of we generate the job description. Like any content that you're creating is with AI and that's going faster. But if you want to get to
Jeet (21:16.995)
Ha ha ha ha ha ha ha
Tony Castellanos (21:42.496)
AI first rather than a by enabled right if you want to get to AI first or you want to get to something that is more core AI native these contracts constructs are essential there's no getting around the fact that your underlying information has to be rock solid has to be maintained and so more and more I think that leaders across companies are recognizing that the value proposition in these investments now of course taking the time is not something that everybody wants to do but you also
The focus time on either, they're not mutually exclusive. You can invest in your information architecture and the systems around it while you're also, I everybody should be experimenting at a local level with what can I do with these new capabilities because it seems like every two weeks there's some new major unlock that has turned the world on its head that we should be trying out, right?
Jeet (22:15.779)
Hmm.
Jeet (22:26.883)
Thank
Jeet (22:35.297)
Yeah.
Tony Castellanos (22:36.183)
you're building many applications for yourself or you're thinking about integrating the service into Slack to make it more efficient. There are all sorts of things that aren't dependent on the underlying sources that you can do yourself.
Jeet (22:45.739)
Yeah. Yeah. And it comes back to that curiosity piece and taking on that ownership and accountability to say, hey, let's, let's try something. And let's see if we can solve it in a different way than having to repeat those processes manually. And I'd be curious to hear. Go ahead.
Tony Castellanos (23:00.489)
I will say that I think that one of the most important things for leaders though is to create dedicated space for this. It is too much to say to somebody, yo, in that half hour between meetings, you should really be experimenting. That doesn't work. The context switching is too much. And so we have invested in, we did an offsite that was really a two day build with AI.
Jeet (23:16.747)
No.
Tony Castellanos (23:25.235)
And given the velocity and rate of change that we're seeing on all this stuff, we're now instituting a day a month where it's focused on how can we be rethinking the way we work, whether it's one type of work, one workflow, something you do in your day to day to be more AI first, AI forward.
Jeet (23:45.323)
Nice. What are some of those examples, I guess, that you're thinking about prioritizing or some of those low hanging fruits that is causing the most trouble that you want to automate away?
Tony Castellanos (23:56.814)
I mean, there are so many, right now, I think the hosting infrastructure for agents is lacking, right? Like for most of these tools, when you're building your agents, you are then committed to their ecosystem, right? Cause there aren't many that are allowing you as a non-technical person, right? To port your build into some like more generally available infrastructure. And so that's a big challenge. But you know, the thing that is top of mind for me most, well, I
Jeet (24:04.323)
you
Jeet (24:19.747)
you
Tony Castellanos (24:25.997)
something that's top of mind for me most. Some of the things that the team came up with in our experiments, we've seen a lot more candidate fraud, which has been true for every people leader that I've spoken with over the last six months. There's just been a massive uptick. so their products have hit the market now that allow for more fraud detection. And they go deeper and deeper on the application review and the auditing of is this a person or not. We've had
mixed results with some of those. And so we wanted to see like, is this something we could do on our own? What are the signals that we would need to collect? What are the inputs that that makes sense for us? You know, and in a day and a half, you're not going to sort of out maneuver fully built teams. But, but it's interesting to play with the concepts because then when you do get into, we want to outsource this? Are we going to select a vendor? You're much more informed. You can go much deeper than, than just,
Jeet (24:56.568)
Mm-mm.
Jeet (25:15.235)
.
Jeet (25:19.341)
Totally.
Tony Castellanos (25:21.207)
good or bad, is, I've tried this. How do you get around this problem that I encountered? So that was one. We've done performance coaching. We've done analytics agents. There are a bunch of things that, again, synthesize content for us to digest and then leverage in our day to
Jeet (25:26.667)
Exactly.
Jeet (25:39.332)
I love that. And I think it's, it's actually even better for, for vendors, because it means that you've tried things and you've identified gap and then it's much easier to talk around that gap as opposed to, Hey, I your agents to do all of this stuff.
And we always have more interesting conversations with teams when they have experiments themselves, because then it's more of like a level set of here's where we are, here's where want to get to, and here are the limits that we're hitting. And often it's not build or buy anymore, but it's often build and buy. And you're kind of being a much more strategic and surgical about what are the gaps you're trying to fill. So, that's been, that's been really good to hear in terms of your approach, Tony. We are almost...
Tony Castellanos (26:16.834)
And I think that's how you lent me a sympathetic ear as I complained about all this stuff. So thank you.
Jeet (26:20.931)
No, it was an awesome conversation at Transform, which yet we will continue. And we're almost out of time here, Tony. So thanks for joining us. I guess before we wrap up, any final thoughts that you want to leave someone with who's in the middle of process optimization in this new world of AI?
Tony Castellanos (26:45.454)
I
You know, one of the things that I like to hire for is people who are courageously creative, and I would use that here. You know, this is a moment where you can't wait for the polished training to be delivered to you because by the time somebody's had the space to create the training, something new has come on the market.
Jeet (26:55.479)
Nice.
Tony Castellanos (27:06.51)
So this is a moment to embrace the inner maker, the inner builder in you and go experiment. You know, it feels like it is adjacent to your job or it's on top of your job, but it's ultimately, you're gonna discover these unlocks that make you better, faster, more effective that people are gonna benefit from. And so making the space to create on your own is the biggest thing. Don't wait. And I know that this is something that is out there, like everybody is.
This is not revolutionary, but I can't underscore it enough and the way that we have sort of our forcing function for getting this into the system is just by actually on the calendar blocking off a day a month and saying like this is what we are all going to do and so we have our accountability buddies, we list out what we're going to focus on, we share out the results and so it is actually a very fun and engaging time for us.
Jeet (27:50.691)
Hmm.
Jeet (27:58.637)
I like that a lot, Tony. Thanks so much again for joining us on the Workups Podcast. And where can people find you? that LinkedIn?
Tony Castellanos (28:07.702)
LinkedIn, yes, definitely. Please check me out. I'm happy to engage. This stuff is fascinating and fun, so I'd love to continue the conversations with anybody who's interested.
Jeet (28:09.826)
Okay, awesome.
Jeet (28:19.083)
All righty, we're going to share your LinkedIn URL in the call channel and they can check you out there.
Tony Castellanos (28:24.982)
My picture is very old. It's about, you know, here.
Jeet (28:27.063)
Yeah, that's a good call out so they don't get confused of who's this person when they click on your URL. We do have the right URL for Tony indeed. So Tony, thanks again and for everybody listening. Thank you and we'll catch you on the next one.
Tony Castellanos (28:32.686)
You
Tony Castellanos (28:41.496)
My pleasure. Thanks so much, Jeet.