Latest / ThinkData Podcast / S4 | E9 | The Hidden Cost of Time: AI, Productivity, and Measuring Real ROI with Laurel CTO - Jason Li

Transcript

S4 | E9 | The Hidden Cost of Time: AI, Productivity, and Measuring Real ROI with Laurel CTO - Jason Li · ThinkData Podcast

• 0:03
  Alex Hutchings: Welcome to the Think Data podcast brought to you in partnership with Mydataworks. If you want to stay up to date with the latest breakthroughs and trends in the world of data and artificial intelligence, and if you're curious about some of the strategies that companies and founders use to launch data and AI products, then you're in the right place. Our aim is to bring together a diverse lineup of fantastic guests from the founders, through to accomplished leaders and product owners at some of the most fascinating data and AI companies worldwide. They will each offer you their own unique insight into what it takes to launch and scale a great data business. Thanks for tuning in and I hope you enjoy the episode. Welcome to the Think Data podcast and today I'm really excited to welcome the CTO of Laurel. Jason Lee to the show. Jason has a really impressive background. He previously served as the CTO of Ironclad. He spent nearly 10 years there. They were building contracting software. And prior to that, he worked for the likes of Salesforce IQ and Grovo. Laurel are a really interesting Series C AI powered time intelligence platform. We are going to get into the intricacies of what it works and how it works in practice. But basically, they are four professional services firms. And the way it works, it automatically captures how professionals spend their time and uses that data and turns it into actionable insights to really help those organizations improve efficiencies and most importantly, profitability. We're going to dig on AI in the workplace, engineering in a very AI first world. And probably one of the most important topics is why measuring? real ROI is still such a hard challenge for most firms and also in this new AI world. One, I'm going to kind of look forward to unpicking with you, Jason. So yeah, thanks so much for coming on. And I know you've been involved with Ironclad for a while, then you left and obviously joined Laurel as CTO. But for people who haven't come across you, who is Jason and what led you to Laurel?
• 2:17
  Jason Lee: Yeah, yeah, yeah. So thank you for the intro and thank you for having me on. So I've been working in this kind of enterprise software B2B space for a while. I mentioned Salesforce and Ironclad, et cetera, et cetera. And I've really enjoyed this space for a few reasons, right? Like I think when you work at startups, you are lots of, you know this, it's really dynamic. You have new problems all the time. Things are shifting, which is really exciting. Enterprise software I really enjoy, as I can expect very clearly, Laurel is building CRM software, building time sheet software, et cetera. It doesn't sound the most exciting thing in the world. However, what I find really interesting about it is you end up working on the tools that people use every day in their jobs and whatever. And I think the exciting bit is you allow people to do what they like about the job. If you think about, in our example at Laurel, nobody goes to become a professional services person, be a lawyer, be an accountant, a track time. That's not what's exciting about the job. You have to do it, but that's not why you do it. And there's so many cases of this across software, right? For like software engineers, like nobody, like version control is not the most exciting thing. But if you didn't have it, you would spend a lot of time worrying about that problem. And so I think that's the exciting thing about the space and what we're doing. And then, of course, the law itself is like, OK, hey, there have been some problems that have not been easily addressed with software before the advent of AI. There are things that are not the easiest to do. Timekeeping would be a great example of where people spend a lot of time and effort trying to keep track of the time so they get paid, their firm gets paid. you soccer hasn't done a great job you know it helps a little bit but just a little bit and so what appealed to me was this opportunity to go and change uh how people work to make it much easier so they could focus on their actual job right be it you know providing legal counsel could be doing accounting work doing whatever um you'd be focusing on that rather than doing that and also focusing on tracking all those things yeah which is that's really interesting in
• 4:14
  Alex Hutchings: It's fascinating because before I knew I had you on, actually, one of my friends runs a consulting firm. They're a small firm. And one of the big challenges he had about a year back was so many different consultants working on different mandates was actually just tracking time accurately. And back then it was still quite a, I'm talking about a year ago, very kind of labor intensive process. So how did companies and individuals previously track time? And at what point and how did you, or not necessarily you, but the founders think there's an opportunity here?
• 4:43
  Jason Lee: Yeah, yeah, yeah. So previously, and honestly, still now, a lot of lawyers and accountants, they effectively would do. They might write down a note. They a lot of them have a notepad where they just like from this time to this time, I did X, right? And so the context here for especially lawyers is they bill hours, but they not only have to bill hours, they generally have to explain what they did, right? So a lot of lawyers, they bill by increments of the 10th of an hour. So, right, they effectively could, and you know, the smallest increment is a 10th of an hour, but they'll bill time that way. And when they send a bill out, they'll like describe, you know, I did this for six minutes or an hour or whatever, right? And so a lot of them have developed pretty sophisticated personal systems for that, right? Where they do that. Someone might have a few software, they hit a start timer and start, stop the timer. But at the end of the day, it is generally been an exceedingly manual process, right? Like where somebody, like as they do their work, they also hit a timer, right? And sometimes people buy physical timer clocks that they use, all kinds of stuff like that. It's pretty wild. And I think if you, as myself, not being a lawyer, not being a accountant, not having done it, it's like when you start to see like, wow, it's a pretty wild kind of difference than people who don't have to do it. to build time and don't have to track it like that. And so I think a lot of the founders and what the folks at Laurel saw was like, that is just like so painful that it's crazy, right? And if you talk to the lawyers who go from working at a firm to go in-house, right? So they go, you know, not working in a firm, but somewhere, you know, a company in an in-house counsel role. It may not be the most important thing they care about, but often it is in the top three or five. It's like, I'm happy I don't have to track time anymore. And so...
• 6:13
  Alex Hutchings: Yeah, I bet it's done.
• 6:15
  Jason Lee: Exactly, exactly. And so like, that's effectively what drove it, right? It's like, you know, you can very obviously see it's like, hey, this, this is like a very important, but not doesn't really, it's not the highest leverage thing you can do, right? It's like very far from the highest leverage thing, but it's a must do. And so it was a clear pain point. And it's also one of those things where it's been done for so long. You know, it's like one of a lot of these things where people I think in it kind of get blinded to it, right? It's just like, hey, this is just a part of our life, right? And so Sometimes you kind of have to have that moment of like, why is it part of our life? Can we change it? And especially now with the advent of AI, can you actually do it in a different way? Like both, not the AI side, but can we collect the data, but then can we then use AI to make sense of that work? And that part had always been really hard previously. It's like, okay, hey, even if you could automatically figure out what you were doing, but if you don't really know what that means, you might be on a call together, but if you have no sense about what that call is, then it's not that helpful for you. you something yeah
• 7:14
  Alex Hutchings: It's really interesting because obviously the use cases there, the problem statement, there's evidently on your point, really valid point, you know, why are you making do with the way you're doing it now? Why don't you kind of adapt and consider a new technology like Laurel? But that's all well and good because a lot of these companies and organizations I'm guessing you're working with, say Laurel as an example, some of them older firms, very established, very legacy tech. very legacy processes. People are stuck in their way. So what was your biggest challenge as CTO to kind of get Laurel working properly and reliably in what can only be described as some companies very complex and legacy platforms?
• 7:55
  Jason Lee: Yeah. Yeah. Yeah. So, I mean, I think I would say that there's a bucket of two thematic areas, right? Like there's technology and there's actually culture. On the technology side, as you said, a lot of these firms are, have... a lot of legacy setups, right? They're not the most progressive when it comes to technology. They're not, you know, they're pretty conservative a lot of these things, right? Think about law firms, accounting firms. They're pretty, risk management is a big part of their job, right? So they're pretty risk averse. And so even when they're like, okay, hey, hey, this is a real opportunity with AI. We could do something here. I think one of the things we can run into are everything from, let's call policies and process stuff, right? Like, okay, hey, everyone's really excited. Yeah, we see the value. Let's go do it. And then you run into procurement and conversations like, oh, hey, we have these 15 different data policy rules, et cetera, that aren't yet updated fully for a more AI-forward world and even sometimes public cloud world. And so sometimes there's just more work to be done of like, okay, hey, if we do these things, what you want from a risk mitigation standpoint is to dissolve. If we have our SOC 2 things, we have our bring your own key encryption, et cetera. You don't actually have to have it running on your own device, on your own servers. We'll solve the same things for you. And so there's some of that kind of conversation of how do we get people to move forward? Even if they have a desire for AI, there's the other things that have been put in place that kind of sit there. And the other one is just simple things like IT hardware, right?
• 9:23
  Alex Hutchings: Yeah.
• 9:25
  Jason Lee: I don't know what device you use, but a lot of folks in startup and tech use these latest and greatest MacBooks that do all these great things, very powerful. A lot of these firms we work with aren't that, right? Yeah. They use more. integrated devices. They have some pretty heavy IT stacks for whatever reason, for whatever mix of things they have. And so, you know, part of what our service provides is that we have a client that sits on your computer to help track what you do across the different programs. And so a lot of our work and a lot of actually the challenge for us is like, how do we get it to work across a diverse set of hardware? Some of the hardware, not the most advanced, not the fastest hardware, but how do we make sure it doesn't get in the way of your work, doesn't crash. And then also like a lot of these systems have pretty lockdown IT setups, as you might guess, right? they're managing risk again. And so like, okay, hey, given all that, given the various tools you use, most firms use what's called classic outlook, right? And which is actually about to be sunset, but they're pretty much everyone is still using it, right? And I think that thematically kind of matches here, right? It's like there's the desire to jump forward with using AI, but so many of the other things are quite antiquated. And so like, okay, how do we kind of make sure to meet them? Or we are, they are, like our ass is not like you got to go change everything. but that does mean extra work for us of like okay hey given that how do we make it work how do we not break your other things how do we make it as seamless as as possible and that ends up being a fair amount of the work um and then the other interesting thing is there's a really interesting cultural element to this where it's like less technical and more more cultural some of this is the data policy stuff but some of this is just the individual actions right i mentioned before is like people have developed systems where they write down time and do certain things And sometimes it's just difficult to get people actually to change from those behaviors. Right.
• 11:13
  Alex Hutchings: Yeah. It's an intrinsic behavior they might have had done for years. And you just thought you're going to come in and tell a 25 year partner, actually, you're going to be using this. Yeah. It's a tricky way. It's an education piece, right?
• 11:26
  Jason Lee: Exactly. Exactly. And like a great example of that is so, you know, I mentioned, you know, we track what people do and then we help kind of people figure that out, you know, describe that. Right. So oftentimes the lawyers, you know, you track your time, you describe it. Firms generally have the rules of how they want you to write about that time that you're billing for various organizational reasons. And so we have a feature that helps do that for you. AI will be like, this is what we think you did. Follow the rules, how you describe them and do that. And then we had feedback from one customer once who was like, hey, the lawyers are not really liking what we're generating and they're not using it. And we're like, great, we should let's go spend time with them. Let's understand what are we doing? What can we do differently? And we just spent some of that time and we found out that that lawyer. didn't like what we're generating, rewrote it to be something of their own. But if you, if you showed their, their compliance folks, they're like, no, no, the AI generated one's better. I don't know. They just like they wanted to do something different because they're just so used to like, I want to do it this way. And so like, there's that, that's just like an education plus, you know, Hey, where do you do a change? And like, it's a funny thing there. Cause like it was costing them more time. Right. They were like changing things that didn't need to be changed and a result that the firm actually didn't prefer either. And so there's that interesting cultural piece. Some of it is like the global. data policy stuff but it's something that's just the individual i'm used to doing things a certain way where can i trust ai actually just help me out yeah and that's you know that there's there's sometimes different steps along the way i will say that the very cool thing about what we do is that because the value is so strong for both the individual user and the firm at large it helps push through a lot of this stuff of course they see the value exactly exactly right the end user is like if we can get best some of these things. It's like, oh, actually, I save a lot of time. And like, we've talked to lawyers at firms where they're like, hey, I have my Saturday mornings back because of you. Because like, you know, you're able to like, okay, fine, do all the things I would have done for half a day much quicker. It's awesome. And then, of course, the firms appreciate it because you're tracking it more easily and you're just finding more time and so you build more hours as well. And so I think that helps push through some of these both like technology and policy and cultural things.
• 13:29
  Alex Hutchings: That's interesting. I know you can't necessarily name names, but... Has there been an example or a client where it's just not been that easy to implement because they've been stuck in their ways or maybe the technology isn't ready? Is there a point where, and I appreciate you're not going to admit that you can't work for everyone, but is there a point where actually you go, they're not ready for this? Is it more around them being in a position to adopt this or is this really a platform and service that anyone can have?
• 13:59
  Jason Lee: I mean, I think this is actually where it's... this is nothing unique about AI as it comes to enterprise software. I think there's always been, and there's still true now of like, Hey, you can have a piece of software that actually can be great if people are willing to go do the thing, right? If you're not willing to go to do the thing, then, then it's, there's always obviously going to be limits. And I think, you know, we don't see that very often because of that, that would be, that would be a high value thing that varies, but you know, you can see that. And I think, I think what we're seeing, and I think this, this might be true with applications at large. If you think about traditionally with a lot of software, especially enterprise software, especially like workflow type software, whatever, some things that have complex change management, you only get the value if everyone buys it. It's like, if you go... roll out a new CRM, a new Jira instance or whatever. You really got to make everyone do it. Otherwise, you're not getting the value. A lot of the things with AI, oftentimes you can get value. Like whoever is the one who's willing to go do it will get value. And the other people will be fine, right? Example for us is like this example, we're talking about some of this AI stuff that we do. We still have like regular timers. You can just go and click and manage your time that way. And that's fine. But that doesn't prevent the people who are going to be all in to get value of the value of that. Right. And I think that's actually been one of the unlocks for us, which I think apply to a lot of these tools is like because it isn't this like really quite as rigid of a software workflow for a lot of those stuff. We can provide a lot of that value, honestly, in a for people where they are. If you want to do it, great. If you want to do this part, it's great. And you get value. If you don't, it's OK. The rest of your firm will. And that's OK. And like then the firm can decide how much they want to push, et cetera. It's not a, the whole thing won't have value unless you're all in. And I think that's been really nice for adoption. It's been really nice for us to be like, hey, we can prove it out in a small group and go from there. And I think that's been a really nice way to do it.
• 15:53
  Alex Hutchings: Yeah. And also you're getting that validation now. That lawyer who said, obviously you've got the Saturday mornings back. And actually that's the value it's generating for them personally. But equally, we're in this world of AI where there's so many tools and technologies. saying there's many if any similar to laurel but you know you can almost get drowned out by all the latest shiny new ai features and products you should use but organizations now i think we've hit this inflection point where actually organizations already struggle to justify and measure real roi you've got the value which obviously saving time helping the front line staff in their work but what challenges do you see companies overcoming and mistakes they're making when it comes to actually being able to measure. real ROI because I guess that's that's the big question isn't it you can go in with the best product in the world but equally they need to justify to their board and the CFO this is delivering ROI yeah yeah I mean I think the beautiful thing for Laura is like we we sit pretty nice place to justify it but I think the
• 16:56
  Jason Lee: I think what we've seen across the market I've both seen and honestly even as an AI consumer as a sort of as a buyer of vendors providing I think it's... the the biggest miss is people just don't actually measure roi it's at the end of the day right like um and like they because martin honestly a lot of times because it's very difficult right um but that everyone is excited about the promise of the value of it and people can see where that could be but i think a lot of folks are having a lot of difficulty actually measuring it right and so you know i i've heard and seen like people like all right let's do a survey well survey people on this tool like is it adding value as we know people aren't the best actually no i want to find that in any meaningful way right and or or they just don't do it or they or they have to do some jumps across a bunch of different things and actually we've talked to different companies like some of these big ai you know agent providers where they're like hey yeah that's actually part of our problems like how do we show and talk about roi for people and so i think that's actually the huge part of it is like just needing to figure out as a company me. How do you go and measure that value? Which I think this is actually, once again, where it comes back to a lot of just the things that are just normal about doing business, not about AI even, right? It's like, because measuring AI ROI comes down to, can you measure the work product, right? If you measure the work product, you should hopefully be able to measure, how was it before and after you implemented some AI to one? Can you see that, right? And so I think more and more trying to figure out. how can you do that where can you see that um and then go from there and i think this is why that some of the ones the ai tools have had the most success with this are the ones that are doing like cost-free savings right you're like replacing support like okay cool can we do the same as this 14 but with this thing like it's easier to show that our life not the most exciting right it's not the like the upside thing that people want to do but it's the easy one and so i think trying to find the ones that are kind of working through those harder problems of like hey how do i find these things that aren't about as much about cost savings but about like hey this is the work we created and because of this tool we can do it better and i think for us like the way i see laurels like we're we're at the beginning stages of it where we very nicely can do it when we quantify by amount of billable hours but i think the nice thing is we're getting a lot of the data of like hey what is that actual work product that's happening and what you know how can we define that better and better of like hey not only is it the hours but this is actually what we produce and this is why it's so exciting
• 19:28
  Alex Hutchings: Yeah, and on the point of that reference you got back about the Saturday mornings back, it's... You know, that's harder to measure in terms of a tangible ROI, but equally, if they're during the week able to do more work, more case studies or case work, because ultimately they're saving time from all the write-ups, then that's quite a more straightforward kind of justifiable ROI use case there, I'm guessing.
• 19:54
  Jason Lee: Yeah, yeah, exactly, exactly. When you can, I think for us, it's so nice because like it is super measurable. At the end of the day, the business is measured on that. I think as we go forward, I think it'll be really exciting because like, all right, you'll save time on doing these other things. Then you can actually probably do more high value work, not just you'll have more time for work. We can do even more high value work as well, especially as you use other tools as well. And so I think the opportunity to continue to both quantify in our case, like the amount of hours you put a bill and the amount of time you're working, but then to continue to up level that work that you're doing to be the highest level work over and over again. And like that, that. to me feels like a lot of what we're seeing with this rapid induction introduction of ai is like hey how can we get the people that we have at this company whenever the company is doing the highest value work every time right um and and a lot of it is like hey what tools can we use to like offload this other work so we can really focus on this one that we really need that judgment that expertise that etc yeah
• 20:54
  Alex Hutchings: that makes sense and i'm keen to touch on your journey at laurel because obviously you've... You're at Ironclad for nearly 10 years. You work for obviously a pocket of Salesforce as well. And you come into Laurel, this kind of hyper growth series. What's been the biggest challenge for you personally in terms of joining an organization in such a key role as obviously CTI? Yeah,
• 21:13
  Jason Lee: yeah. I mean, I think a big thing that I'm seeing is like it feels like Laurel is going through all the growing that you might have done, like at Ironclad and similar like companies. very quickly, right? Like the pace of change that's happening right now across the industry is very rapid, right? So like the company, just in terms of employee base is growing a bunch over the last year, the customer base is growing quickly over this last year, the sophistication of the customer base is growing quickly as well. And so I think that's one of the things I'm seeing is like, okay, hey, all these things that I've seen before and are things that we need to do, but just like, oh, hey, we just need to do them faster and change faster in lots of ways, right? in the In the sense that like, you know, there are even things like internal process stuff or internal tool stuff. Like, oh, I totally get why this was, this is the way it was even three months ago or even definitely six months ago. But things are different now because XYZ has changed. And so we just have to go and do it differently, which was always true at startups. But like that time, that's effectively time compression, right? Of like a startup. So you might be like every year, like we got to reinvent ourselves because XYZ changed. And that's three months now or six months now or whatever. And I think that's it is like that, that change. And like... A lot of that actually, I think, comes down to culturally where not being precious about how we do things, right? The way I kind of always described it is like nobody really cares about process and organization, those actual artifacts. They care about how it feels, right? Like how is my work? Like what does it feel like? And like all those things are in service of that. It's like, okay, hey, can I move quickly? Can I have ownership? Can I do these things? Or can we build in this way, et cetera? And so because of this, Rapids... state of change in the market and for laura specifically it means that you know we have to reinvent ourselves we have to go build and you know update how we do tooling because of this so we have to you know change how we work together because there's more people doing this these things right and so there's just that and getting to me that's one of the big challenges is like okay how do we get everyone in that mindset and then just keep doing that right just being comfortable continuously doing that um that to just kind of match what we're we're seeing in And the marketing, that's what we, you know.
• 23:24
  Alex Hutchings: the realities of the company and yeah do it and then and just make people used to it is such a big part of it so yeah i think it's really a good point it's a the rate at which innovation and technology teams are having to adapt now I've been a staff member for 23 years in tech. I've never seen the rate of new tooling, new ways of working, obviously the impact of AI on engineering teams and workflows. And it's just, how do you personally keep on top of that? Because obviously you're the one that's obviously sitting with the board, the founders, the investors and saying, we need to commit to XYZ tool or these processes. How hard is that to do?
• 24:03
  Jason Lee: It is both tremendously hard and also easy. Tremendously hard in the sense that, like... there is so much and like it is it is more than anybody can fully keep up with right like if we're honest like it's it's there's so much there um but it's also easy because like it is super fun and exciting right like was i think i saw something about like sergey going back to google like he was i think it was him i was like you can't sit this out this is not like this is not the moment to sit this out if you work in software right and i think i think i think a lot of people the work in software feel this way right it's like it's such a moment of an exciting moment of change that it is easy to get excited and to try things and and do that and so that's where it's you know a lot of people in tech and software like it's a job but it's also a hobby like a lot of people got into because they really have passion for the thing right it's your spare time right yeah yeah yeah exactly and honestly for me you mentioned i was at ironclad for a long time so i was there for a long time and then i actually i took a break working full-time before All right. can be laurel and then part of it was like great like i don't have an org to run i don't have all that stuff i can go build software so i actually did that i was like all right let me go build some things you know i wrote some game with some biome coding i did some some stuff with uh you know 3d printing and building technology that i don't normally work with and like that um that like that hobby aspect is is there and i think for a lot of us that helps make it a lot easier and so like even in the last few months i think everyone knows like december january there's been And this feels like a sea change in hell. development is happening and so like a lot of that stuff is in a lot of ways easy because it's not like we have to go force yourself to go read some you know industry mag that we don't don't care about it's just like oh this is a fun new thing let's go and try and play um and so i think a lot of it is that and i've always been a big believer of like a lot of that is learning by doing right you can read what you can read and that's great but you just got to try the things right and you got to try the thing and get the feel for it um which is an amazing fit for this like a lot of it is actually not really i mean there's a tiny bit there's some reading but a lot of it is is the trying and and kind of uh going through that iterative learning process yourself um is such a big one and so that's it for me right and i think and you see that i think with lots of folks who are the same of like hey i'm really interested in this and like either doing with their work or like i have a side thing that i tried it out myself and now i want to bring it into work this this idea that tried um and how i built things and i think that's that's worked quite well Both for myself and folks I've seen as well.
• 26:35
  Alex Hutchings: And what's next for Laurel? I know obviously you guys are still growing at a rapid pace, but from a product standpoint, what can people start to see and hear about you guys in the market?
• 26:46
  Jason Lee: Yeah, yeah. So we just announced a new product called Signal. It's based on the same core technology, right? So our core technology ends up being when we track what people do for work, right? We have cloud integrations. We have clients that sit on people's machines to get us its people. We had this cloud integration. We had this desktop client. to track what people work. And our current product is creating these timesheets more easily for billable hours. Signal is effective at taking that core technology and really opening up to more of the market, right? And the way I think about this is everybody doesn't really know where they spend their time working, right? If you ask people, they will guess, but they have no idea. And so the Signal product is really to help companies to get a better sense for that. It's like, hey, where are people spending their time, you know, in, you know, working with which clients or which projects? in which programs doing what. And so then the question is then hopefully we're be able to surface all, we will be able to surface insights that people can then make strategic decisions of like, okay, we're spending a lot of time doing X and with these high value people. And we really shouldn't be doing that. Like, can we provide tooling or support, et cetera. And so it's really that, um, taking that core, like understanding where you're spending your time, um, and now servicing a new product, which is, you know, the aggregate insights for companies so they can.
• 28:00
  Alex Hutchings: they can make those calls exciting times now it's been super interesting as i said but when i knew you're coming on and i then bizarrely had this conversation my friend three four months ago i was like is it can make sense i know no surprise you guys have seen the growth you have the investment but equally the successes and it sounds like it really is kind of a hyper growth business and i'm yeah really good to have you on jason you were kind of really uh Really clear. And as a true passionate technologist, it's kind of nice to kind of have the other side of the growth journey than kind of the sales and go to market piece. So thanks so much for coming on. And I wish you and kind of Laurel all the best. I'll be sure to keep an eye on your success.
• 28:39
  Jason Lee: Thank you for having me on, Alex. I really appreciate it.
• 28:40
  Alex Hutchings: Appreciate it. Thanks, Jason.