How to lead a product team through AI adoption - Tiama Hanson-Drury (CPTO, Opus 2)

Tiama Hanson-Drury is CPTO at Opus 2, the legal technology company behind the systems that run some of the world's largest and most high-stakes litigation. She started her career in sales, moved into product in 2010 and has spent the last 16 years split roughly evenly between individual contributor and leadership roles. Her verdict on AI in product organisations is neither evangelism nor scepticism: the tooling has removed software as a bottleneck, but the markers of quality that mattered before AI have not changed, and teams that forget them end up shipping slop and burning out. In this episode, we discuss:

1. Why a sales background is better preparation for product than an engineering one, and how to lead engineers without pretending to be an expert
2. How the promise of doing more with AI is producing overload and burnout, and what leaders owe their teams in response
3. Why high AI adoption scores mean nothing if pull request sizes are ballooning and someone else has to review the output
4. The three things Tiama looks for in her teams now: outcome focus over craft, curiosity, and a culture of sharing failure as readily as success
5. Why the right move is not to hand three of your seven steps to AI, but to start again from the outcome you want
6. How Opus 2 went from shipping every six to 12 weeks to shipping daily, and why that made documentation and planning more important, not less
7. The case for betting on humans in the loop when your customers explicitly prefer it
8. Why Opus 2 is still hiring juniors, and how to build commercial judgment when the groundwork can be skipped
9. The one situation where Tiama tells people to let AI write the first draft rather than the second

Chapters

• (01:31) From sales to product
• (03:24) Why the move is less obvious than it sounds
• (04:34) Leading engineering without an engineering background
• (06:28) How outsider questions sharpen engineering teams
• (07:36) What AI has changed for a CPTO
• (08:16) Overload, burnout and the pressure to token max
• (10:32) Adoption scores versus quality signals
• (11:39) What AI adoption actually looks like inside law firms
• (12:42) Outcomes over output
• (14:07) How Opus 2 built its AI strategy
• (16:07) Removing friction from an AI rollout
• (16:53) Partnering with AI instead of splitting up the steps
• (18:08) A message from Mike Belsito
• (19:40) Three things AI asks of a team
• (21:47) More agency, and why planning matters more
• (23:51) Scaling a team on conflicting evidence
• (26:09) Betting on humans in the loop
• (27:36) Why Opus 2 keeps hiring juniors
• (29:44) Building judgment when the work moves faster
• (32:36) Sharing your own AI operating system
• (33:22) AI slop, CVs and the hiring algorithm
• (36:17) Top tips for a product career in the age of AI
• (38:36) Wrap-up

Referenced

• Opus 2: https://www.opus2.com
• Opus 2 announces addition of Tiama Hanson-Drury as chief product and technology officer: https://www.opus2.com/news/opus-2-announces-cpto-tiama-hanson-drury/
• Minna Technologies, where Tiama was previously chief product officer: https://minnatechnologies.com
• DX, the developer intelligence platform used to track squad performance: https://getdx.com
• DORA metrics: https://dora.dev
• Vote for Mike Belsito's South by Southwest 2027 session: https://tinyurl.com/belsitosxsw

Where to find Tiama Hanson-Drury

• LinkedIn: https://www.linkedin.com/in/tiamahansondrury

Our Hosts
Lily Smith enjoys working as a consultant product manager with early-stage and growing startups and as a mentor to other product managers. She’s currently Chief Product Officer at BBC Maestro, and has spent 13 years in the tech industry working with startups in the SaaS and mobile space. She’s worked on a diverse range of products – leading the product teams through discovery, prototyping, testing and delivery. Lily also founded ProductTank Bristol and runs ProductCamp in Bristol and Bath.

Randy Silver is a Leadership & Product Coach and Consultant. He gets teams unstuck, helping you to supercharge your results. Randy's held interim CPO and Leadership roles at scale-ups and SMEs, advised start-ups, and been Head of Product at HSBC and Sainsbury’s. He participated in Silicon Valley Product Group’s Coaching the Coaches forum, and speaks frequently at conferences and events. You can join one of communities he runs for CPOs (CPO Circles), Product Managers (Product In the {A}ether) and Product Coaches. He’s the author of What Do We Do Now? A Product Manager’s Guide to Strategy in the Time of COVID-19. A recovering music journalist and editor, Randy also launched Amazon’s music stores in the US & UK.

Transcript

Copy

SPEAKER_02 0:00

Anybody who works with AI recognizes AI slot. AI helps you do so much more. And the general feeling is that, oh my gosh, I'm getting really overloaded. They're just starting to feel a lot of burnout. Do things because you have a strategy and they're on strategy. Don't do them because everybody's talking about them. The last thing I want is just because we can build something, we build it. How does AI change how you think about scaling your team? It's not so that we can automate and take people out of the loop. Like we want those same people who our customers love and trust to keep servicing them. 77% of jobs are cheaper to be done with human. 40% of corporate agent projects will fail in the next few years. It's great that you're using technology to move fast. But I also want your thoughts. I want your opinion. I hired you as an individual for a reason. Do you want firsthand usage? It will make you so much more confident. Don't lose your core of who you are, those unique things that drive you, and then keep your hands dirty.

SPEAKER_01 0:58

Tiana, welcome to the product experience. How are you doing? I'm well, how are you? Yeah, very good. Thank you. And we're gonna have a chat today about kind of AI in product careers and um and how things have changed for people as they look at you know developing their career in product and what AI is doing to that. But before we get stuck into it, um, it would be great if you could give us a quick intro to you and your own kind of career in product and how you've got to where you are today. Sure. Um everybody who's uh watching.

SPEAKER_02 1:34

And yeah, I started my career in sales actually. Um before getting into the working world, I was raised by two clinical psychologists. I'm the only person in my family who doesn't have a PhD and work in medicine or science. But like the things that drove them actually were the things that drove me getting into a business, which was a desire to help people achieve things. Um and I started in sales because I'm pretty conversational and solutions focused, but quickly realized that what was making me successful in sales was listening to customers' problems and kind of creating repeatable, scalable solutions for it by partnering with tech or data or ops. And after a while of doing that, I realized that was a discipline that was becoming known as product. And so I moved into product um in 2010, I think, roughly. Um, and I've now been working in it for 15, 16 years. Um, I spent about half my career in product manager roles and then about half of it in leadership roles. And I'm currently the chief product and technology officer at Opus 2, which is a legal tech company, and we help provide basically like the engine room for high-profile litigation cases. So everything from when you decide to take on a case and you need to pull in all those, you know, hundreds of thousands of documents and you need to put together your case strategy, you need to assign all the kind of work to do across the life cycle of a case. Um, we also do the technology that sits in the UK hearings room. So anyone who's in court presenting evidence, we do the speech to text transcription so that we can apply AI on that for them, all the way through to hopefully settling for the outcomes that are good for the customer. Um and yeah, it's been a really fun journey.

SPEAKER_01 3:23

Hearing you talk about it, it sounds like a really obvious step from sales into product, but I think you might be the first person I've met who's made that move.

SPEAKER_02 3:32

It's not, yeah, it's not the obvious one. But it, you know, I think it comes down to the things that make a product person um effective or and and also a technologist, I think, is like curiosity around like what is that user problem that's happening, how do you find a scalable way to solve them? Um, and how do you do that in a way that drives good margins or good financial performance? So it was I it was I remember my CFO pulled me aside and was like, Tamma, what are you selling? And I was like, what do you mean? And he was like, well, your margin on these projects are so much higher. And I realized I was like, you know, I started to learn this whole concept of gross margin and um why having 70% was a lot better than you know, 50%, which was the average project that we had. And so yeah, I mean, I think in some ways it is obvious, but I also think that, you know, in this was true before AI, there was a perception that you needed to grow out of an engineering background in order to get into product, and I've always believed that's a that's a fallacy.

SPEAKER_01 4:34

And we talk about kind of product being this intersection between sort of the commercial side of things and then the customer side of things and the technology side of things. Um, I guess with sales, it's very kind of customer and commercial. How have you found like that gap, I guess, on the technology side of things in your career? Like, how have you filled that?

SPEAKER_02 4:56

I think it's come down to being able to just not try to be an expert in things I don't know, right? Um, I think so. This is my second time with the CPTO title. Um, last time I didn't take the title because our um founder and technical, like technically the CTO was still with the business, but I was he didn't want to act in the CTO role, if you will. It just didn't give him a lot of energy. And so I took it on and it was a nice learning ground because I could ask him safely, like, well, why are we doing this? Or that doesn't make sense to me, or how does that fit with the strategy, or um, are we certain that this is gonna help us, you know, in 10 years versus, you know, 10 months from now, right? Um, so it's a good place to learn, and I think that prepared me to just feel comfortable treating my engineering counterparts as sources of expertise and subject matter, you know, kind of um experts in the business. And so then we just partner on like, well, what is it that we need to do? Yeah, so I you know, I think in many ways, not being an expert in that space allows me to give them more freedom to be creative. Um, but me and the value I bring is just bringing it back to the larger picture of what we're doing, the business model, the KPIs, the KRs that we've said, how are we gonna drive that customer value? So um it works, but you have to be humble on both sides, I think. And you need to be willing to say that you don't know the answer and to ask, you know, five whys until you finally get to the, oh okay, now I get it. That's why we're making this decision.

SPEAKER_01 6:28

Yeah. I always think it's interesting as well, coming from a non-engineering background as a product person, that you know, the the questions that you end up asking the engineers and and the way that you end up working with engineers, it it almost gets them to think differently or to kind of go back to first principles and things like that and be able to explain things in a non-technical way, which then helps get to the to a better solution.

SPEAKER_02 6:54

Totally. Yeah, I mean, you know, every engineering org that I've uh led officially or even you know previous to the CPTO roles was looking after and squads, right? Um they always find it funny when I like drop into their, you know, rituals, whatever they're doing, and make them go through the flywheel with me or relate something back to the strategy. Um, but in the end, they generally appreciate it because they see the connections that aren't necessarily there. And you think about what engineers are generally asking for from product, they want more context on the why. They want more context on what we're doing. So I agree with you. I think going through that conversation helps connect the dots for them and for myself and therefore for the execs as well. So yeah, it works well. Awesome.

SPEAKER_01 7:38

So, how has AI changed things for you as a CPTO?

SPEAKER_02 7:43

It's just, I mean, I think this is everybody's topic, whether it's the CPTO or any role that you have or how you serve your customer, um, how does it change things? So I think it's made a lot of things possible to go faster, I would say. Um and I think if you care about the metrics that told you you were doing the right stuff before AI, it can also help you go faster and have good quality. Um I think that one of the topics that I see discussed a lot with it, whether it's our customers or whether it's my teams or whether it's my fellow executives or my girlfriends, you know. Um there is a perception that AI helps you do so much more. And the general feeling is that, oh my gosh, I'm getting really overloaded. Um, so you see a lot of teams talking about mental health challenges and the fact that they're just starting to feel a lot of burnout. Um and I don't, I don't actually think, and I don't, I try not to lead my teams to think that they need to be, you know, token maxing or even working in the same ways that they were before. I really try to model for them, but also hold them accountable for taking a beat. So, you know, if if somebody is doing, you know, loop engineering and they've got something working on, you know, a prototype, I don't want them to like feel like they have to sit there in front of the machine and then also have five other agents that they are like guarding at the same time if they're gonna do that for eight hours a day. I think that isn't a good use of their time. I'm totally supportive of them, you know, realizing that that is helping them build things like, you know, so much more quickly. And as long as, again, we use our guardrails, both human and process, to make sure that the quality is there, I'm totally fine with us adjusting and I'm encouraging of us adjusting our ways of working to that. But I also find myself saying, like, have you gone outside? Like, have you gone and played with your kid? Have you, you know, gone and done some research about what's happening in the market? Have you listened to a podcast? Um, because I think our brains are, they need, they need that different type of stimulation. And if you're just sitting in front of a screen all the time and managing, you know, a set of agents who are managing teams of agents in the background, you may be progressing a lot further through your work packets that you need to do. But in theory, I think you would want to be freeing up some time to use that for the things that only you can do as a human. Yeah, and in order for you as a human to perform well, I don't know about our listeners or you, but like me being in front of a screen 12 hours a day doesn't lead to the best outcomes for me.

SPEAKER_03 10:29

Yeah.

SPEAKER_02 10:29

So I mean, I think it's changed a lot. I have intentionally, the team was saying that we affect we effectiveness max. And I don't to be honest, I don't think we're quite effectiveness maxing yet. Um, I can look at our DX scores and I can see that we are definitely above average in terms of adopting AI, time savings for AI, but I also look at things like the you know PR size and they're they're massive, right? And that means that you're it doesn't, it's not the only thing that this means, but it can mean that we have slop that's going in. And so then that just means somebody else has to kind of look at that. So we're we're using tools like DX within our teams to have the squads look at their performance and say, okay, you're doing a great job adopting AI, but are we actually driving our Dora scores up? Are our DX scores and the snapshot that they do every month? Does it actually show that the quality of work is good too? And I think that's what I'm gonna be working with the teams on is how do we make sure that we adopt AI, but to help us do what good used to look like, which is to deliver, you know, small, well planned, you know, user value that drives customer impact. Um, and I think that same kind of level of focus is the same thing that our execs, our customers are talking about. Like our clients are legal tech is one of the hottest spaces for AI. And if you were to go to a lot of the legal tech conferences right now, every law firm is you know hiring fewer associates next year, and every law firm is adopting and making adoption the biggest thing for AI. But then you speak to the lawyers themselves and they're they're figuring it out. They don't necessarily know how to do this, and their jobs can't just be, you know, pulled into 32 repeatable workflows, you know, and it's I'm not an AI skeptic, far to the far to the other side of that. Like AI is a huge part of our strategy. It's a huge part of both internally strategy, customer focused strategy. But I just think that we would all find it more refreshing to not focus on just a new world of AI working, but also to remember the things that were quality markers in the past and use those as guides because those things haven't really changed.

SPEAKER_01 12:42

Yeah. Right. Um, so and by that you mean, like you say, the effectiveness maxing, I think you called it. But like they called it both sides. Very cute. I like their idea of it, yes. So by by which I assume you mean sort of focusing on outcomes rather than the volume of output. Yeah, I think it's the outcomes.

SPEAKER_02 13:00

That's it. It's a simple, clean way to put it. But I mean, I just think in a world where you can ship, like as people are saying all the time, like software is no longer building software is no longer the barrier. But for us, we're a PE-owned company. Like our customers use us for the world's biggest and most high stakes litigation matters. We're not just going to be shipping stuff into production for them. Like, we need to make sure that what we're putting in adds to their experience. Um, the designer at Lotus once said, like, if you add something, add lightness. And I tend to think about that in software a lot a lot. Like, we're trying to make it less cumbersome to use uh legal software. And so the last thing I want is just because we can build something, we build it. And so I think it's more about just making sure that we can clearly explain what we're shipping, why that drives the strategy, how it's measured in terms of customer value, and ideally the code isn't bloated on the back end, and we're not creating a bunch of like work for other engineers who have to peer review that stuff before it goes into production.

SPEAKER_01 14:06

And it sounds like in terms of Opus2's kind of approach to AI or AI strategy, as you as you mentioned, you're quite far down the road with like embracing it within your culture. What like how did that happen? What does what does that look like within the teams? Like, how did people have to adapt?

SPEAKER_02 14:27

I think we're still adapting. I and I think everybody who feels like they're still adapting should take a beat and just like no, it's not as sunshine and roses as I think a lot of people make it sound. Um, there are companies who are fully AI native from the get-go, and that's just a totally different paradigm. But you know, Opus has been around for almost 20 years. Our heritage is in technology, right? And building our product with our customers. And so AI has been a natural why I joined. I don't come from legal tech at all, but the reason why I thought this would be an interesting next um adventure to do was because legal was so complementary with LLMs in many ways. Um, and so for us, I mean, it's been a twofold strategy. Uh, and credit goes to the CEO before I even got here. You know, he had said we've got to be looking at reinventing the way our customers get value from us. If AI should be supporting them, let's do it. So they had shipped their first AI set of modules before I even got there. We've gone a lot further since then. And I think the reason why we've been able to adapt AI pretty well internally is because we work all day with our customers on how to use AI effectively. So if you're an associate within a law firm and you used to, you know, go in and manually build a chronology, if we can help you pull out, you know, events that might be significant or relevant for that, that's great. Uh, if you're putting together a description of who Willie is because she is a witness in a case, well, AI is great at summarization. So, you know, you can pull these things out. And I think the team learning how to sell AI to our customers, help them understand some of the pretty now tried and true use cases for LLMs. Um, so I think we had that foundation. And then I think I've drawn I've driven the AI rollout at two companies now. And you learn a lot through doing that. You basically learn that you want to remove it's like product, you want to remove as much friction as possible for people. So we're working on the rollout of Enterprise Cloud now. We had it in pro everywhere, and we were like, let's let's make the jump to actually doing enterprise. And the gentleman who's doing the setup is working, you know, to make sure that everything comes configured out of the box. So we don't have to have the teams figure out how to link it into Pendo. We don't have to have the teams figure out how to link it into Grafana, into Linear, into Launch Darkly. Like, let's connect all the systems for them so that they have the ability to get right in and start um reinventing what they do. And from then, it's from that point forward, I'm just trying to model to them that the in my experience so far, the best way to adopt AI is to not look at what I used to do and say, okay, I did a seven-step process. So I'm gonna I'm gonna assign AI to three of those, and then I'm gonna do the other four. It's actually what's the outcome that I want to achieve, and I'm gonna partner with AI and how can I do that? And the the difference that makes that a positive outcome and a quality outcome is also using your judgment about knowing like what are the things that maybe they weren't represented as steps in the past in those seven steps, but they were judgment and they were decision hurdles across that. And that's what I bring forward to okay, I've partnered with AI to reinvent the way I'm gonna achieve this outcome. But my judgment and my experience tells me that this this response has come out of it isn't good enough, and I need to refine my skill file to reflect that, right? So I don't know if that answers the question, but I think we had a we had an advantage that we were selling it to our customers. We had a heritage in building with our customers, and so those two things allowed us to start to reinvent the way we worked with AI internally.

SPEAKER_00 18:02

Um yeah, hey everybody, Mike Belcito at Mind the Product here, and I just wanted to pop in with a very exciting announcement and a favor to ask of you. So last year, right around this time, I got asked by Wiley to write a book. And in fact, I just turned in the manuscript of that book on August 1st. The book comes out next February, and it's all about the timeless skills that help product people evolve during any major shift. I'm talking about things like product sense, taste, judgment, which in this AI everything world that we're in right now, these things become even more important. And the book's about not only those things, but how we can actually improve at these things. Anyway, next March is South by Southwest in Austin, Texas. And I would love to launch the book at South by Southwest. And I have a session that's up for a vote. So I would love your help in voting for this session so we can make that happen. If you go to tinyurl.com slash belcito at sxsw, that's a tinyurl.com slash delcito at sxsw. You can read all about the session. And if it's something that interests you, just click on the heart. That's the vote button there. And the more interest there is, the more of a chance we'll have for me to launch as a book at South by Southwest. And I appreciate your help in making that happen.

SPEAKER_01 19:40

What did it require from your teams in order to make that change?

SPEAKER_02 19:46

I think we're learning still. Um, but I think for us it's really three, it's three things. It is willingness to focus entirely on outcome versus craft. So that's the most important thing. I'm looking for people in my teams who are showing and signaling to me that they care about the outcome and not the way that they arrive at it. And like I said, it's not about throwing away all the lessons. That's not it. But if you are so focused on using that same example of there's seven steps I did before I would push into production, uh, you're not gonna get the benefit of um AI that you can leverage because so much of that, those steps don't need to happen anymore because the models have gotten so good. So they don't have to account or they don't have to account for the way you achieve that outcome. So the first is are you outcome focused? I found that a lot of people will self-identify as I don't, I actually don't like this new way of working. I really like the craft of doing things the old way. And I I tell this people that's totally fine. Like you should life is too short to work in the wrong company. If that's what you want to do, there's plenty of companies right now where you can go and work and you can still kind of take that sequential process. We can't do that because we have so much opportunity and I want us to eat it as quickly as possible. But I think it's good for people to self-identify with that. The next is curiosity, like the amount that we don't know should be exciting to you. Um, and I think the people who really show signs that they are going to excel in this new world are excited, sometimes for the first time because they're young in their careers, and sometimes for the first time in 20 years in their career, because it's the first time that they don't know how to do something that they've been knowing how to do for decades now. Um, so that curiosity. And then the third, I think, is a sharing culture. Like it is so important to share negative, positive, success, failure, um, because we can't learn it all. And so those are kind of the things I'm looking for and that have been important to the team progressing. Um, but we're still on our journey.

SPEAKER_01 21:47

And um, and before we started talking, we were kind of chatting online, and I think you mentioned something around like it th the use of AI in the team kind of almost translates into more agency or independence as well. And that's Like more independent way of working, potentially?

SPEAKER_02 22:02

It does. I was talking with some of my um principals about this, and I was like, Do you think it's translated to more agency? And they said, Absolutely, in some ways. Um, absolutely. We find ourselves going through a lot less of the hurdles that we used to have. You know, we are in a unique setting right now because when I joined Opus, we were shipping every, what was it, like six to 12 weeks? We're now shipping daily, right? So uh it's not just AI that's changed, right? Um, but in pursuing that new release process and how do we iteratively build with our customers, I've encouraged the teams to throw away old things that weren't serving us, PI planning, um, for example. So yes, they have more agency, but then one of them said, and I completely agree, and I think anyone who's talking about like the way AI is changing things is documentation is becoming more important. Planning is becoming even more critical. The time that you used to do planning, you know, for a big project, especially anything that was strategic or had a lot of architectural complexity, it had kind of natural time in those projects for people to get on board and to learn and adjust. That's gone now. Like you're literally just iteratively building with, you know, your system. And so the documentation about what is the outcome that we need to achieve, what are the decisions that can be made with or without me in the room, if I'm a PM, for example, um, or if I'm even the architect, right? What are the architectural choices you can make, you cannot make? That becomes even more critical because otherwise you just build through without doing that planning work. And that, as anyone who's built software knows, is like how you end up with just loaded code, you know, stuff that maybe has gone off piece that's not really delivering what you need to.

SPEAKER_01 23:48

Yeah.

SPEAKER_02 23:49

Um, so yeah.

SPEAKER_01 23:51

Okay. And how does AI change how you think about scaling your team and organization?

SPEAKER_02 23:59

We talk about this a lot, my CEO and I and our exec team. Um, and the evidence is conflicting, right? So on one hand, um, I've got my partner who is a CEO, and they've gone, you know, built an AI basically platform for the product builders to build on and removing a lot of the requirements on engineers. And that led to, I think, about half of their engineers deciding they wanted to move on and do other things, and they haven't needed to replace them, right? Like, so that's a great example of they're getting efficiency and they're scaling without you know, bloat. But then you look at kind of the bigger stage kind of projections from the Gartners and the, you know, um kind of the consultancies, and they're still saying that, you know, 77% of jobs are cheaper to be done with humans, or that, you know, 40% of um agents, you know, corporate agent uh projects will fail in the next few years. And I think what that tells me, and it goes back to a lot of the principles that have guided me, is like do things because you have a strategy and they're on strategy. Don't do them because everybody's talking about them. Um so I would suspect, suspect, that a lot of those agent um or agentic projects will fail because they weren't tied into strategy. I've seen it in MA where I'll be talking to a prospective company where you were looking at acquiring, and I'll ask them about how they thought about AI. And oftentimes I'll hear answers that tell me that they're just kind of doing it. Um right. And that's not not there's nothing wrong with that, but I think it depends on your operating budget and what you've got to play with. And most of us have a lot of work to get done, and we need to get value into our customer hands. So I mean, I think the general principle that we've been going with is that we do believe that costs will go down, although right now also the the data is showing that you know AI expenditure budgets are up like 300% and 320%, I think, is the last figure I saw. And it's because agentic systems consume more usage, right? Um, we're still betting on the fact that the costs are going to go down, the more supply, more demand will naturally kind of work itself out. And so we we bet on, and we may be different than a lot of people, but we bet on humans in the loop. Um, we've just got too much expertise in our business, we've got too much um customer preference for having a human element still involved. So instead, what we're focusing on is how can the agentic systems that we're building allow people to, you know, do more. Um how can we, I mean, a really great example is we've recently launched AI in our hearing room and we've got you know nine agents sitting behind the hearings assist that's there. It's not so that we can automate and take people out of the loop. Like we want those same people who our customers love and trust to keep servicing them. But if that can help us start to service more of the remote hearings and it can allow our lawyers who are in the courtroom to not have to say, well, Lily said today that she was gonna show up and meet Tiama for the interview at three, but in the evidence that she submitted ahead of time, she said she was gonna be there at two. You know, is that do I have that gap correct? Like we try to remove the mental load and also think about how our technology scales in ways that it can work when we're not in the room if the customer doesn't want us to be in the room. And so we're kind of building for both. And we're gonna continue to have the human in the loop because our customer prefers it that way, quite frankly. Um if our customers get to a place where they tell us they don't want that, we'll we'll reassess at that time.

SPEAKER_01 27:36

And um, when you think about junior roles or like entry-level products, product people, engineers, designers, how are you kind of catering for for for bringing, you know, that because you got you know, as a as a growing company or as a company that wants, you know, to bring people up into the ranks, yeah. You kind of feel like you are responsible for training the next generation. I do, yeah.

SPEAKER_02 28:03

I don't know that everybody does, but I do.

SPEAKER_01 28:06

There's probably lots of people that don't feel like that. But how are you kind of handling that side of things?

SPEAKER_02 28:11

We're quite intentional about it. So we are continuing to hire juniors. Um, we're continuing to invest in, you know, bringing people who are earlier in their career in. And for me, this comes back to the diversity of thought thing. I've always felt this way that I think A, because I started my career in insights and data, and I saw how if you didn't have someone's perspective in it, it was very easy for an entire strategy, a product, a category, a campaign to go live with like old white guys' perspectives only. Um and I just thought, well, we surely want some young white guys, and what about some black guys and some Latino guys? And you know, like I just think, and women, women, please. Um and so I just think that for me, what do these younger or juniors bring? They bring a completely fresh perspective, right? Um, so they're gonna be challenging me to think in different ways, and they're gonna be challenging the engineers to think in different ways, and they're gonna be challenging the product people to think in different ways. So I would suspect, and I spoke to some of our um investors about this recently when we were thinking about our AI enablement strategy across the org. And I would think that actually there's a lot of argument to have people who are younger and who are not as senior, because I think a lot of people who are as senior maybe have a bit more, this is not a blanket statement, but maybe have a bit more to lose or feel like they have a bit more to lose. So for me, I continue to try to build diverse teams. I just think, and juniors are part of that diversity strategy.

SPEAKER_01 29:42

Yeah, that makes sense. And um I think one of the other things that I was curious about as well is how when you're bringing those juniors in, because the rate, the the rate at which they can then start doing the work and learning, you know, the learning is so fast, the support that they have around them with LLMs, you know, the the data and the knowledge that's available to them is way beyond what we had at the start of our careers.

SPEAKER_02 30:11

I wish we had had it. That'd be great.

SPEAKER_01 30:14

But then I think the the time spent doing the job doesn't necessarily like that's not there. Yeah. So is there anything that you're kind of doing to make sure that they're they're getting that um that time spent on the job, kind of like, I guess doing the groundwork of like how do you make a a good commercial decision? Like how do you develop that judgment, you know, and that understanding of the customer and and things like that without just kind of, I guess, cheat cheating your way.

SPEAKER_02 30:46

Yeah, I mean, um, we we work on this a lot. You know, one of my principals, you know, has recently hired um our first, I think since I've been here, our first junior, junior PM. And so we talked a lot about the onboarding plan and what was that gonna look like. And it's the same that I do for a principal or for a C, you know, a mid-level. Um, it's what are those 30, 60, 90 day things that I'm expecting them to do? And, you know, anybody who works with AI recognizes AI slop, right? Like we know when we get something and you're like, you're quietly thinking about this. No, you're not quietly thinking about this. That's like Claude writing that or the long dash, or you know, so I mean, like we're just using the same 30, 60, 90, and then, you know, kind of I do an on we do an onboarding doc for everybody of like what will value look like in your first year. And we just kind of rate back to that and say, take this again. Like, I want you to use your brain and think about this and come back to us and give this. And it's just the same, I think it's the same quality expectations that have been there for someone to come in and find out if they can onboard and they can get up to speed effectively. It's just we the onus is on us as managers to be even more diligent about spotting stuff that comes through like LLM versus Tony, my new PM, right? Like, and just really trying to encourage and create that environment where it's great that you're using technology to move fast. I want that, but I also want your thoughts. I want your opinion. I hired you as an individual for a reason and just pushing them to come back and have to present it back to us.

SPEAKER_01 32:26

Yeah, I think that transparency, like built like you mentioned it earlier, but building in that transparency of like how you're using it to support your work as well is really interesting.

SPEAKER_02 32:35

Yeah. I mean, I was thinking about it today. I'm setting up my um, I'm transferring my operating system into the enterprise model, and I was thinking I'm gonna just share it with my team because I think everybody is struggling right now. Everybody's struggling with how to adopt it the most effective way. I certainly don't have all the answers, but I think that transparency of sharing and back to that sharing culture is so important. So I plan to kind of say, okay, well, look at this is how I'm gonna do my board reporting. This is how I'm gonna do my competitive uh briefing, you know, and like remove the like, oh, this is an executive and this is what I do, and instead try to say, like, you can be doing the same thing. It may not be the board report that you're creating, um, but you should be looking at competitive stuff and like what is the reporting that you're doing, or how are you keeping your stakeholders informed? So there's lessons there.

SPEAKER_01 33:22

And you mentioned AI slop earlier, and I feel like one of the places this really turns up is with CVs, yes, and the whole kind of job application process. Now, I think the job market is really tough at the moment. It it seems to have picked up in terms of available jobs based upon my scanning of uh scanning of what's going on. But uh the what I'm hearing still is when people are applying for jobs, like you know, getting sliced out of that application process because of the the kind of AI tools vetting CBs. Yeah, but then also people using AI to um yeah, to basically write their CVs, and then it's just full of stuff that they're not they're not even really checking properly. Yeah.

SPEAKER_02 34:10

I think the the C V is the only place where my advice would be different on how to use AI versus like to my teams, I always say you write the first the 50%, you know, and like when the LMs first came out, they were like, Oh, it gets rid of the 50% page problem, like you can start with something. And I've intentionally said, no, like you write the first 50%. Please maintain your ability to create sentences and to think, you know, about strategically what are you trying to do, and then use AI to kind of fill it in around that and be your your forcing function and um kind of your operational partner in that. I think just because CVs are usually going through a bunch of algorithms that are somewhat withheld from the applicants and they don't know all the right terms and terminology, that that's probably the only time where I would say maybe you start with an LLM and say, okay, here's here's my job that I'm going for. Like when, you know, here's my LinkedIn profile. Like, can you write the first version of this resume for me? Um, that makes sense and paying attention to the words that need to be flagged in the systems that are going to look through this to give me a chance of actually being one of those candidates that's pulled out. But I think then what you could do is you could look at that first version and then rewrite that in your own words so that you don't have this stuff that's so telltale, you know, coming out of an LLM, but also so that as um a friend was showing me recently, she was uh they're reinventing the hiring scorecard for large, very large tech firm. And she was like, in what world does this level PM do this? Like this was so clearly then like spit out through an LLM saying that this person, she's like, Well, we would never need anyone else if like there was a PM at a level four that does this, you know? And so I think a good forcing function for anyone is to go back through what's come out and say, like, did I actually do that? Um, is that describing the way that I would like actually drive drove value? And that can help you kind of bring a bit more human back into it. Um, so that uh screening or hiring manager doesn't look at it and discount it, even though it's made it through the screening process, but it looks so artificial.

SPEAKER_01 36:17

Yeah. Well, Tiama, it's been so great to talk to you today about all of this. I just have one more question before we go. Yeah. Um, so for those listening, um, it would be great if you could give us your top tips for, you know, carrying on with a product career in this age of AI. Yeah.

SPEAKER_02 36:37

Um I think two things. I think start with who you are uniquely, right? So I we talked, we started this with talking about you know where I came from. And there's things that make Tiana uniquely me. And I don't try to be anyone else. And that's good because what is that like silly old thing like everyone else has taken, right? So be yourself and think about what that is. And I I it all started with for me like writing an operating doc about who I am, how to work with me. So I've gotten pretty good at thinking like there's certain things that make me special and I add value. And I also know the weaknesses that I have, right? Um, so if you haven't done that, I would say sit down and kind of answer those questions about what what is it that drives you and what do you feel alive when you're doing? And what does that mean if you're going into product? What does that mean about the type of product professional that you want to be, whether it's a builder or you want to be a manager or whatever it is. And then I think the second thing is to just build, really. Um, I think if you can be clear about who you are and the things that make you unique and you keep that as an up-to-date living document, and then you're building, you're getting your hands dirty. And that's like that's 99% of what a hiring manager is looking for. Um and it will keep you humble and it will keep you growing. So I mean it's really simple, but I would say, you know, make sure that you're building on a regular basis, not to your own detriment, please. Like you don't need to feel like you're always on the AI. Um, what is that thing people do to work out treadmill? Yeah. I was gonna say tread wheels. Tread wheels. That'd be bad. I I'm not I'm clearly not using a treadmill if I can't remember the name of it. Um uh but yeah, you don't want to do it to the extent that you feel anxiety or like you're falling behind, but you do want first hand usage and it will it will make you so much more confident in how to tackle things. So I think just don't lose your your core of who you are, those unique things that drive you and the things that make you uniquely well positioned to add value, and then keep your hands dirty.

SPEAKER_01 38:36

Yeah, amazing. Thank you so much for joining us. Thanks for having me. It was great, and thanks to the listeners. The product experience hosts are me, Lily Smith, host by night, and chief product officer by day. And me, Randy Silver, also host by night.

SPEAKER_03 38:53

And I spend my days working with product and leadership teams, helping their teams to do amazing work.

SPEAKER_01 38:59

Lou Ron Pratt is our producer, and Luke Smith is our editor.