Latest / Range Podcast / The vesting cliff your joiners hit in year five
Transcript
- Giac Soliman: Welcome to another episode of the Range Podcast. Today I am with Colette. She is head of Toll Rewards at Chainlink Labs, where she transformed the long-term incentive program and built an in-house equity platform supporting more than 1200 users. With 15 plus years across high-growth technology and consulting, including Uber and Deloitte, she is a strategic operator. Focused on solving complex company problems. Colette, welcome to the podcast. Colette: Thank you for having me. Really happy to be here and just to share a bit more about the trenches I get to be in in the last few years. Giac Soliman: We can't wait to hear more about that. But before we go into AI and w all what that's happening in your world, you told me you did a Ruby on Rails boot camp over ten years ago. And you didn't pivot from that. And then at some point, you know, you resurrected your own GitHub account from the dead. Why don't you take me back to that moment first? Colette: Well that was I think twenty fifteen. if you remember twenty fifteen, that was when coding boot camp was flourishing everywhere. I had to jump on that wagon. back in the days I would say Ruby on Rails and Python were like on par. But clearly we know who the winner is now. Python clearly took the lead and just kept running. So at that time we I got to join this coding bootcamp. I thought it would be a really good timing for me to learn how to build because my background is in finance. I did tons of consulting. But there's a voice in me that always feels like there had been a urgency, like a a desire to build and be creative. So that's why I took on that challenge of like during my gap period to learn coding. But very quickly I realized that. Coding at that time in 2015 was about going into Stack Overflow to search for a solution that already exists and and eating tons of pizza. So very quickly I learned that is not really the lifestyle I had in mind. but what that career gap and time to learn helped me is that I actually do like technology a lot. I really like the how we Can transform different use cases and just the kind of the thinking on how to build a good product was embedded in me through this process. So from that experience, I then started working for a a company that deployed SAP Success Factor, did a lot of sales there, and did led quite a few technology deployments. So I think it's all a little bit connected. That was the period of early days of coding bootcamp. Software engineer boom and also SaaS and it's just jumped on the wagon. Giac Soliman: You said Python won the war, the coding war. I would also ask, is learning Python still relevant? Colette: It's a good question. I honestly not. Giac Soliman: I guess I nowadays, like if there's one thing, you know, there's people are living the hype with AI and people are comp like they're very skeptical, right? But there's if there's one thing both groups can agree on, is that AI, if there's one job it's good at, it's coding, right? And it is one of the reasons why people like you and me now are able to get things done and build tools, not on Excel anymore, but on you know, on the web or on different platforms without actual coding language. So what it's different like compared to ten years ago when you were doing bootcamp, what it's different from how you you were building then and how you're building today? Colette: I think it a lot of the building is focused on the developments back then, but because the the building today is now focused a lot about having actual smart requirements and also testing. So pre previously I would say like maybe out of the whole development like the product, you know, from concession to deployment, maybe seventy percent is just you developing things. But now it's like because that now shrinks to like let's say ten percent, if not less, of of what that is. Your focus now is having really smart requirements. but let's just let's just think about it. Like I would say one of the one of the things I learned throughout this process is like probably my requirements are stupid. I just don't know how stupid it is yet. So having smart requirements is really the starting point. And then how do we test things? Yeah, having a really good, smart testing strategy to make sure the thing is working. And then the last part is like really focusing on the fine tuning and making it perfect, 'cause that still takes like that more longer iteration. Giac Soliman: And if you were to apply this in your current job, you know, this this softer developer thinking that you inherited from the boot camp, can you tell us one opportunity you identified in your current job that you said, right, I'm gonna take a stab at this because is that one thing I complain to my husband every day when I have a bad day at least? You know, and then perhaps you say, Right, maybe I can look at how AI can solve this to help me enable me to solve it. what's that one use case that you started from building with AI in your total rewards work? Colette: I think like a lot of people, the starting point was chatbots. And you know, as silly as it is, I think chatbots still has a lot of value. the first thing I did was create a working manual of my boss. Based on what based on everything I know Giac Soliman: Excuse me. Colette: about my manager and our conversation, I try to un use AI to articulate how best to operate and work with her. And then that becomes its own little working tool for me. and that is amazing. It helps me with meeting preps, it helps me with just like navigating my own career and also communication styles with certain with certain individuals. So so building that brain, it was the was probably the first use case I used. It's not specific to compensation, but it's very, very handy for everybody. Giac Soliman: I mean that is brilliant. If you think about like work is is done by people. Yes, agents as well, but if you think about it, building a ch yes, true point chatbots, we're all familiar with it. So a perfect starting point. And I think what you build also shows how deep and engaged you are in building strong relationships at work. Right. So I I think it's what is a brilliant idea. Does your manager know that you build this? Colette: Yes. and to be fair, like Giac Soliman: All right, cool. She's not gonna learn from it from the podcast at least. Whew. Colette: I think sh I think she also did the same for her boss. So I also encourage Giac Soliman: It's a chain. Colette: I also cr encourage my team members to do the same. I even send them the manual of how to work with me. we work in a I don't know if I've had the chance to share this, but I work at Chaining Labs and it's a fully distributed remote company. So a lot of my Giac Soliman: Awesome. Colette: life exists on Slack and Being good communicator is like survival here because of how much Giac Soliman: Absolutely. Colette: can get lost in translation, just in general. So having clear expectations on how I would like to be communicated and how others want me to communicate with them is is vital to my work. Giac Soliman: Yeah. And this is like next level. We're not talking about polishing up Slack DMs or literally we're talking about adapting and customizing your communication approach to fit with the preferences of your internal stakeholders. Colette: Yeah. Giac Soliman: and you've started a chain. I mean, no wonders you work for a company called Chainlink, right? Your your manager has the same with her manager, has the same with her manager and also downstream. So it's like you created this domino effect, which is so interesting. Okay, so that was a start. Now, why don't you tell us about something that you've been working recently? we know how dreadful sometimes equity comp communications can be because people never understand it no matter how much you communicate. why don't you why don't we move on to that? Tell me where the need stem from and essentially what you've what you've built. Colette: it's vesting schedule and LTI in general is a passion topic for me because of how because of how much work and how how how specific the business problem is that we're trying to solve. So there's many layers to this where I we I think a lot of startup and different companies anchor on the traditional four-year equal vesting. But as companies mature, and especially in my current company where it's crypto can be extremely volatile and our long term incentive is anchored on this vehicle, it's extra important for for for us to design programs that helps buffer for some of the volatility or at least be able to help with explaining some of this volatility. So vesting schedule is is helpful because if you imagine A lot of your compensation is tied to one price point in the beginning. And subsequently, all of your subsequent values that you gain from this company is anchored on that one price point. and that could be you could be a winner on this formula or you could be a loser in this formula, depending on who you join. Yeah, Giac Soliman: Yeah. Winfall gains or losses both ways. Colette: correct. So similar to high growth company like Nvidia. And I think there was a few examples of like Robin Hood and just these ones where they also recognize this problem. So there is a trend of exploring either shorter vestings or front-loaded vestings, which can help with just creating more leverage in the latter years. So part of my goal through this tool is to be able to model out all the different scenarios of designing a customized, custom design vesting schedule for the company. We want an equitable program that's easy to explain. And when you have people joining in different times of the year, how do we design a vesting schedule so that people don't fall off the cliff suddenly? You know, say if you move things to say everyone has a one-year system, but then now there has a gap between their their joining dates and like the year end of the year. How do we make sure that that their vesting experience is enjoyable to say the least? So the tool that I created was just something very simple that I I initially just told with the specific requirements. It's otherwise something that I would build in a Google Sheet and have to iterate and like do multiple scenarios. What AI has unlocked for me is to create like a very simple web app that can help visualize. joiners in different quarters, how their total comp would look like in in LTI in the different quarters in year one, in year two, and and help me like even design things like, hey, if I change my refresh cycle from December to January, what would that look like? If I change my vesting schedule instead of 2025, 25, 25 to like, you know, 40, 30, 2010, how would that look like? so that's really helped unlock like just a way for me to communicate like first from a designer perspective, fig find out the optimal solution, but also to land the decision with leadership because now they can see this. It's very tangible for them to go into the system to also make tweaks themselves and understand like the rationales of why we would do things a certain way. Giac Soliman: I'm so curious what kind of feedback you got. Like when I not every rewards team is building so actively with AI, right? And when you went to your exec team and showed them, hey guys, we we built this, right? what was the initial reaction? Colette: I think frankly the the goal of this is to make better business decisions. So I think their reaction was focus on the outcome itself. Chainlink Labs is a very much AI forward company. So it w everybody is I would say building tool is the baseline expectation now. And it's very, very easy. you just talk to Claude. Like there's not not nothing that is like magical about this. So I would say they were impressed by the outcome. Of just how tangible the the simulation is. But it doesn't take away from the need of the total rewards theme, me helping them articulate the business impacts of different areas. So it's is simply just a tool within my tool belt. But the layer that that I need to add in to help them understand the cultural impacts of things or like the financial impacts of certain things is is still is still that incremental thing that they look for. So the tool doesn't doesn't change that. Giac Soliman: someone still needs to explain to them what a tool is doing and what to do with the data. Like, you know, Colette: Exactly. Giac Soliman: you see the outcome, but if you don't deal with this stuff on a day to day basis, you need someone from that domain with that expertise to to guide you in making more informed decisions. And I think that's where this is going. It if I recall correctly, you didn't just build a tool that is able to like analyze scenarios and display vesting schedules in a more, let's say, graphic visual way for employees. You built you w you went a step further, is that right? Colette: in well I I can actually maybe show how it looks like, but essentially this was the vesting simulator. Giac Soliman: So just for anyone that's listening in, we're now showing you know a dummy version with dummy data of the tool of the vesting s of the vesting simulator that Collette has built. I can see lots of nice colors. I really like colorful tools. I think aesthetics obviously Colette: Yeah. Giac Soliman: are important. Why don't you walk us through just briefly you know the the structure and perhaps Tell us a little bit more of why you structure this way. So there's an input pan on the left, right, where you can change inputs and see the impact on the scenarios. Presumably Colette: Yeah. Giac Soliman: you're measuring things like are you are you measuring dilution by any chance? Are you looking at that as well, or are you looking at the vesting granting frequency only? Colette: Yeah, s simply on the granting side. So this is essentially looking at how an a a new joiner would experience their vesting in the first one and two years. So what we have here is a like how what typical new hire grants with four year equal vesting would look like, plus refresh on top. So essentially the problem you see here is regardless of when you join. What you'll experience is first you will have a four-year equal vesting, and then subsequently you will have refreshes on top, which are these red and yellow Giac Soliman: Mm-hmm. Colette: green elements here. And then you have this drop-off at year five. if we were to assume that the blue line is already the market competitive rates, essentially we have year two, year three, year four are like. Pay that is on top of like a very competitive rate already. And that creates like a expect heightened expectations of compensation in in different years. So with this in mind, there's a need to rethink how we should articulate and design compensation so that the vesting is in a market competitive level and also enable more pay at risk. So Hypothetically, like this would allow me to simulate other options here, such as, hey, what if the new hire grant if if the annual vesting is this much, what would the new hire grant could be like if it was a you know 30, 30, 20, 10? Like just a like or just a different way so that we are able to kind of show show folks like what different options of the vesting so that it's lines up differently. So hyper so essentially what you would get is like a a version of of of pay that is that is looks that that's maybe a smoother path going forward versus previously. Giac Soliman: Every time you change the numbers I almost like get hypnotized and by seeing things moving, really it gives like a dopamine boost for a compensation person. Colette: Yeah. Giac Soliman: what would you say are the key pitfalls to avoid for anyone in your shoes? So think about another reward leader that wants to build something similar or comparable to this. Colette: Look, I think I would say one of the y there's always a appetite to build things to solve problems and that's doesn't go away and that's still like something that people should explore more. But when you're designing things for end user, you the the pitfall here is you fall in love with the solution and not the problem itself. So Giac Soliman: yes. Back to one one developer thinking here, Colette: Yeah. It's it it's so easy to get in the trap of like, hey, this is the product and everything becomes like a product led topic and it delineates from like what the business problem or issue we're trying to solve in the first place. So there's this meme of just like, Hey, we're you're falling in love with the solution and not the problem itself and that creates that that doesn't move things forward as you'd like. So my one thing to think about in building is just like Things that w helps me do my job, I I'm happy to just keep iterating, be creative, and now the the the barrier to entry is so low that it's just something you can do on the side. But when it comes to things that solves a specific business problem, we we gotta be much more we gotta be much more in tune with the so problem itself, with what the customer needs are, and avoid the proliferation of just throwing new products to people because there's gonna be a product Giac Soliman: Mm-hmm. Colette: fatigue. It's like, there's another tool that's being thrown at us, another link, another issue. And that's not gonna be the end goal of that's not the end state of where I think the the future of this of this is where this is going. we're not gonna have like a tens of dozens of tools that just every team gets throws in the in the in the like in the in the user's view too. For them to absorb. Giac Soliman: so much resonate with what you said there. 'Cause it's not just about building cool things and just because you can, you're gonna waste your weekends. I speak from experience. You're gonna waste your weekends building stuff that no one's gonna use. That's why it's so important to be deliberate about what you want to build and then chase the outcome, narrow it down to a few sentences so it's understandable before you actually start wasting your afternoons building something. a lot of people stress about this. Like we always hear it from consultants as well. Get your data foundations in place before you build. have you experienced that thought or concern or obstacle in any way when building something at work? Be it your vesting schem simulator or the you know equity LTI tool slash platform you built, which we haven't mentioned yet, but I thought of Weaving it in. Colette: Yeah. Look, it's Data Foundation, it's you know it it's I've worked on, I think in the last 10 months, that's been really the focus. And the reason why this is extra close to my heart is because when I first joined the current company two years ago, we were operating long-term incentive in Google Sheets. So we got to really figure out if we go out to market through a Carter, Morgan Stanley, or Computer Share or what alternatives there are. And because we're a crypto company and with a very complex like LTI programs, there was four when I joined that I had to harmonize, we decided to go built in-house. And that was probably one of the crazier decisions that we had, but in a way had taught us so much about why the data is important because we built the data from scratch, from Google Sheets. So we Giac Soliman: You gotta Colette: designed how we called vesting schedule. We designed IDs. We designed how we want the reports to look so that the data can in the future integrate. So the for the first year and a half year and a half maybe maybe a year my focus was working with this external developer and me as the business owner like giving them requirements of how to build this tool and lesson number one of like how you build tool is your requirement is probably stupid and it's true and it's probably too long and it's true. It's kind of like AI prompting like once you give the developer like too much of a context and you keep adding changes change requests, the code quality just shatters. So we learn all of this the hard way in how to really have good requirements that developers can understand, they can build, they can articulate understand how this impacts other use cases. And with this And with this like development, we also learn how to develop things in the right environments, a testing environment, a developer environment, and a production environment, and how to really partner with our internal stakeholders with security and IT so that they can help with making sure like we have the guardrails in place to hold all of the private PII data in a in a secure way. So lots of great learnings there. But that leads and and that leads to like the data foundation topic because it was a night and day before this was deployed and after this system of record was deployed. We now have a system that has a thousand two hundred users. We have grants new grants being granted every single month and we have new we have settlements every single month so hundreds and thousands of payments is going out through this tool that we've now built in this live for a year. With all of this data, we got to really design the data architecture of this. You know, of course we can always export something and have this sit in another Google Sheet and also keep staying on Google Sheets to like that's where we live and breathe. but Google Sheet has a limit. And we've hit that limit really, really fast because vesting schedule data and just effective dates and all of those things will the I think the limits qu I'm don't quote me on this one, but like after like hundred thousand lines like it breaks. So we were forced onto figuring out a long-term solution to data to re-data engineer what we had into something sustainable. And that path was big query. So my focus for the last eight to 10 months had been to transition our all the data that had been living on the sheet into a system and subsequently through that those data in that system, how to bring those along with our workday data. Into Google BigQuery, into a place where we can do all that data transformation, do all those validations so that we have a real-time source of truth that can be used for like enabling future agentic use cases and other things that we have in mind. But but that's for another podcast, probably. Giac Soliman: Yes, there's you're you're definitely a treasure of content, Colette. We could dig deeper and deeper, but yes, just wrapping this up, what's that one check not that you run before you trust what the AI or whatever you build hands you? Based on your experience, going back and forth to get the data foundations right. What's that one good requirement? Colette: The it probably sounds like something on the internet already, but my one quick requirement is that just acknowledge that your requirement is probably stupid and it's probably too like it probably can be simplified. the reason I say this is it's it's easy to it's easy to want to have a lot of different options and use cases and cover all the things. But without doing the hard thing of just like cutting the requirement in half and thinking from a first principle way of what is really important, you will start just automating stupid things. And that had, you know, in in just thinking about how we build beta foundation, have we been putting too much in there that ultimately was just maybe information more noise than real information that draws insight? Probably yes. So my challenge to myself had always been like acknowledging that probably when we have a starting point of what I want, probably half of that, like really challenging myself on like can we eliminate half of this? And then think from a first principle way to rebuild. Giac Soliman: That's deep. Now, last one for you, and I want you to dream a bit here, if you could wave a magic AI wand and fix something in your professional or personal life, you can choose. No matter how impossible, what's that headache you'd make disappear with this AI magic wand? Colette: Yeah, imagine. Okay. look, I actually I'm not 100% sure if AI can fix this right away, but I do think that we as a people function need to fix this. I would call it quantifying the difficult to quantify things because AI feeds on data, and as much as we can build the data foundation to like channel everything, help people make good decisions. If that if there's too much noise in the performance signals or like we don't really know how to measure impact, me as a total rewards person can't really better advise a leader on where they should invest dollars. Because end of the day, total rewards is about where you put your dollars, where you're investing and where you're betting. And and I would love for a better way to quantify impact and performance so that I can confidently advise leaders and help them invest the dollars better.