Top Predictions for Product Operations in 2025

In 2025, Product Ops is positioned to play a critical role in shaping the future of strategic product development and operations. As the year begins, the Product Ops HQ community came together to share their top Product Operations predictions for 2025, and what challenges and opportunities lie ahead. Moderated by Clare Hawthorne, Senior Director of Engineering & Product Operations at Oscar Health, this meetup explored predictions, ideas, and strategies to help Product Ops professionals succeed in a fast-evolving landscape.

This roundtable provided a chance for Product Operations professionals to share their thoughts in an informal setting. Clare kicked things off by sharing her predictions for 2025, sparking a wide range of ideas and conversations. The focus wasn’t just on what Product Ops leaders should do, but also on how Product Operations can drive change and growth within organizations.

What you’ll hear:

• How to adapt to emerging challenges and leverage tools like GenAI.
• Strategies for driving product operating excellence across your organization.
• Insights into improving team collaboration and enabling new technologies.
• Predictions and real-world examples to help you navigate the future of Product Operations.

Watch the recording now to gain valuable insights that will help you stay ahead of industry trends and set yourself up for success in 2025!

Key Takeaways

1. Champion Outcomes-Oriented Tech Teams: Moving beyond the administration of roadmaps and OKRs, Clare emphasized the importance of being stewards of these processes. This means holding teams to high standards, asking the right questions, and ensuring that everyone involved understands the true value of a given initiative. It’s no longer enough to just track deliverables—Product Ops professionals must take an active role in making sure outcomes are being met.
2. Enable GenAI Adoption: As technology advances, one of the key opportunities for Product Ops professionals in 2025 is to become enablers of GenAI within their organizations. Product Ops sits at the intersection of many departments, making them the ideal role to help drive the adoption of GenAI. With their deep product knowledge, insights into customer feedback, and visibility into operational inefficiencies, Product Ops can empower teams to develop and implement GenAI use cases that can transform operations and decision-making.
3. Expand the Scope of Process Design: Clare also shared the importance of thinking expansively about where Product Ops can make a difference. Product Ops teams should not limit themselves to just product management processes. For example, at Oscar Health, the Tech Ops team took the lead in redesigning incident management and job monitoring processes. This is a great example of how Product Ops can apply its operational efficiency expertise beyond the traditional boundaries of product management, improving efficiency across multiple teams and departments.

Transcript

Product Ops HQ Meetup Kickoff

Ana Andrade: Hello everyone, and welcome to our Product Ops HQ Meetup, the first one of 2025! My name is Ana, and I’m your community manager here at Product Ops HQ. I’m also part of the Dragonboat team, joining you from Portugal. I’m super excited to be your MC today as we kick off the year together. While we wait for everyone to join us, please let us know where you’re joining from in the chat.

With me today is Leah, also from Dragonboat. She’ll be in the chat to help with any questions you may have and share some helpful resources with you.

I’m really happy to see some familiar faces today, as well as some new ones, too. So if you’re new here, a special welcome to you! Just to give you some context: our community was created to serve the growing needs of product operations professionals like you—people who want to meet, share ideas, learn, and grow together. We host virtual meetups covering a wide range of topics, and we also provide networking opportunities through our Slack community, where you can connect with peers, ask questions, and share insights. If you’re not part of it yet, be sure to check it out!

We have an exciting session today, and I can’t wait to get started. But first, a few quick housekeeping notes: This session is being recorded, and you’ll get the recording via email after the event. This session will also be interactive and will be guided by one of our community members. With so many of us here, please keep your mic muted while the moderator is speaking. After that, we’ll open the floor for everyone to share. You can either drop your thoughts in the chat as we go, or you can unmute and share directly with the group. We want this space to be one where everyone feels comfortable sharing and connecting, so the most important note is: have fun and stay engaged with us!

Before I hand the mic over to our moderator, a quick word from our sponsor: This community is brought to you by Dragonboat. Dragonboat is an AI-native product operations platform that enables everyone in a product operating model—from CPOs to product managers, to product operations and their teams—to set clear strategies, build portfolio roadmaps, and deliver products that delight customers and accelerate business outcomes. Dragonboat Product Ops Co-pilot has been running in production for over five years, supporting thousands of product teams across leading companies and managing over $52 billion in annual product and engineering investments. Dragonboat is more than just a platform—they’re your partner to achieve product operating excellence, which includes, of course, sponsoring our lovely community. If you’re interested in learning more about Dragonboat, Leah will drop a link in the chat for you.

And with that, let’s get to the exciting part! Today’s roundtable is all about what’s in store for product operations in 2025. We will have Claire Hawthorne, Senior Director of Engineering and Product Operations at Oscar Health, to guide us through this discussion. She will start by sharing her thoughts, and then it’s your turn, so get ready to contribute. So, Claire, without further ado, I’ll hand the mic over to you to introduce yourself and kick things off. Thank you!

Clare’s Prediction 1: Champion an Outcomes-Oriented Tech Team

Clare Hawthorne: Yeah, thank you so much, Ana, and thank you everyone for joining. When Ana reached out to me a few weeks ago, I was so excited to share my thoughts and predictions for what’s in store for product operations in 2025. I did just get a pop-up on my computer that my internet is unstable, so wish us luck. This might be a little bumpy, but I think it’ll be great.

I’ll also be trying to look at the chat if any of you have immediate reactions when I share. But, Ana, can you advance to my first prediction?

So, first up, I think product operations is well-positioned to champion outcome-oriented work. So, what do I mean by outcomes-oriented? I think this is where we’re not just administering roadmaps or having people write down their OKRs. We’re really seeing our role as stewards in that process. And this is a mindset shift that we had here at Oscar.

Clare Hawthorne: Bear with me, guys. It looks like there’s construction in my building, so let’s make this a little more cohesive. I think that being able to not just administer processes, but really see yourself as a core participant in these processes, is critically important. And so, I came up with a gif for each of these.

Why do we do this? Because it’s what champions do. We’re all champions in product operations. I think it can sometimes be intimidating, especially when you don’t know every nook and cranny of the roadmap, to feel like it’s not your job to push back on product managers. But we’re in a position to see the big picture, and we understand what those standards are.

So, depending on your role, you might be someone who’s running an OKR process, or you might be someone collating materials like the PRDs. But make sure that you’re not just moving that work around. You’re really using your expertise to make sure that the teams are hitting that high bar and standard in terms of outcome orientation.

Clare’s Prediction 2: Enable Your Company with GenAI

Clare Hawthorne: Now, my second prediction is that I believe very strongly that product operations will be the place and the people who unlock enablement for GenAI company-wide. Now, I know that some of you on the call might feel a little intimidated by this. You’re not an AI expert, you’re not an engineer, and there may be other people on your team who feel like they’re better positioned. But believe it or not, Ana, if you can share that next gif, I think this community should always have faith in themselves.

We are connectors, we are teachers. We lose internet from time to time, but in all seriousness, I do think we’re really in the crosshairs. Something exciting we did at Oscar recently was develop our own AI workshop. We had some internal tools, and we were trying to encourage the use of our LLM tools and AI within the product ops team.

A member of my team, Sofia, who may or may not be able to join, put together a walkthrough of our tooling. We had a live workshop where everyone picked a use case, did a prototype, and brought it back to their stakeholders or scrum teams. We got such great reviews that now we’ve turned that into a “train the trainer” program, and other folks within product operations are working with pods, scrum teams, or other parts of the business.

I think we’re in the intersection of understanding what LLMs are capable of, but also knowing some of the manual work being done and the processes that need to be designed. And a key piece of our DNA that not everyone else has is that we’re also mindful of the long term.

So, we know that if we just spin up AI every time anyone mentions the word, we’re going to create a lot of tech debt and process debt. I think we’re in a position to help govern and steward that process.

Clare’s Prediction 3: Apply Our Superpowers to More Tech Processes

Ana Andrade: I’m sorry, guys. Let’s just wait a second for Clare to rejoin.

Clare Hawthorne: I am so embarrassed that this is happening, and there will be a nasty letter to my internet provider after this call! But anyway, that was our chat. Unfortunately, my phone doesn’t have a hotspot, which is making this even worse. But let’s move on to the third prediction. Then we’ll open up the floor, and hopefully, it’ll be less distracting if I drop in and out.

My last prediction is that we, in product operations, will be able to apply our superpowers to [pause]. So, what do I mean by this? We are really good process designers. And I think some of the classic product ops use cases are things that we’ve seen in a lot of different places—roadmaps, OKRs, PRD templates. These are all classic ways that product ops get their foot in the door with these superpowers. These superpowers of design in the classic product management use cases. What we’ve seen at Oscar is that product operations is really involved in processes that are, frankly, more tied to engineering use cases. These are things like CapEx, incident management, job monitoring. In all of these cases, product operations were brought in as the process designer. We had something really manual and time-consuming, and we knew that at first, it felt a little intimidating to go into some of these more technical processes.

Clare Hawthorne: So, for the closing slide, Ana, you can bring up all three. These are my three predictions for the opportunities that product ops have in 2025: championing an outcomes-oriented team, enabling companies with GenAI, and applying our superpowers across more areas.

Terry Courtney Sharing Insights on a Proof of Concept for an AI Concierge Service

Clare Hawthorne: Okay, so that was my talk track, and thank goodness, because now I can engage with everyone else in the conversation. I’m curious to hear what folks’ reactions are. I see that some notes are starting to come in. Terry, I saw that you have a proof of concept for an AI concierge service. Can you share a little bit more about that?

Terry Courtney: Yeah, sure. So I’m just starting this journey. I’ve only been working on it for a couple of weeks, but the results so far are encouraging. Basically, when I joined the Product Ops department about 2 and a half months ago, I transitioned from product management. I’ve been doing some discovery work, linking it to my previous experiences. There was a lot of feedback about how difficult it is to find out who’s responsible for certain features, who the tech lead is, and when updates are scheduled. It’s time-consuming to gather all this information—spending 5 minutes here, 10 minutes there. So, I saw an opportunity to use AI for this. Initially, I’m focusing on fetching information, but I also plan to explore generating things like simplified roadmap presentations based on our current roadmaps. It’s early stages, but that’s my goal.

Peter Boersma Asking If Product Ops’ Superpowers Are Shared with Other Specialized Ops Teams

Clare Hawthorne: That’s awesome, Terry. Thank you so much for sharing. Hopefully, that’s prompting some ideas for others. We are trying this on my mobile phone. I did lose the chat history, but I’m excited for others to jump in and share either their reactions to the predictions or their own predictions. Okay, I see a question in the chat here. Do you agree that the reasons why product operations will succeed are shared with other Ops teams like design, app research, Ops, PMM Ops, data and analytics, and even DevOps? Peter, I’m not sure if I’m entirely following your question. Do you want to come off mute and explain where you’re coming from?

Peter Boersma: Sure. Hi, these are the three reasons why product ops will succeed, and how we can contribute to fantastic teams. The caveat here is that it depends on your definition of product ops. But…Do you agree that maybe other specialized ops teams have these same reasons? They have the same capabilities, the same superpowers? They might also be introducing AI, etc. Do you feel this is a shared effort, or is it specific to product ops? And here I say, product management.

Clare Hawthorne: Yeah. And so I think that—thanks for the clarification. For those of you who’ve seen me around the community, you probably know that I have a very broad definition of product ops, probably one of the broadest. Internally, we even call ourselves tech operations, not just product operations. We see ourselves as stewards of the entire tech team. We’re here to help people ship their roadmaps, not just product managers, designers, or program teams. 

So yes, I completely agree that all of these traits are shared across operationally-minded people. And that’s something we’ve been trying to make sure people understand. We have a different set of superpowers than our colleagues, whether they’re product managers, engineers, designers, or security professionals. There’s a really lovely complementary relationship between us and the “maker” functions. While we make things, we’re focused on different things. Personally, I get a lot of satisfaction from being that enabler, connector, and long-range thinker. I think many of you can probably resonate—if you’ve worked with product managers, engineers, and designers, you know they’re often so excited to get something out the door, and then it’s like “squirrel!” and they’re off to the next thing. We’re here thinking not just about launch day but launch day +30, +60, +365. I think a lot of the operational sub-functions you mentioned share that same DNA.

Peter Boersma: Thank you. That helps.

Rukmini Baruah Asking: How Technical Should Product Operations Be and What Are the Top Skills for the Future

Rukmini Baruah: Hi, Claire, I have a question. This is Rukmini, joining from the San Francisco Bay Area. Claire, I wanted to ask you—I’m very curious. Given your experience and the current situation with AI, the tech stack really reducing, and companies changing their ways of working, I have two questions. First, how technical do you think product operations managers should be? When I say technical, I mean really SQL-heavy, understanding AI, product management, and that kind of tech stack. How important is it to upskill in those areas?

And number two, what do you think are the top 4-5 skill sets that product operations folks, like all of us in this community, should have going forward?

Clare Hawthorne: Two great questions. Let me see if I can sort of decouple them and address both. The first one, in terms of disruption and technical sophistication… I think that’s the first question, right? At Oscar, we’ve recruited people who have high technical aptitude. We haven’t in the past required SQL experience, but the expectation is that you’re curious enough and willing to pick it up if it’s important in your day-to-day. And I say it is important for a lot of people’s day-to-day.

That said, I also think AI is a huge democratizer in terms of leveling the playing field. I know that in our product operations team, we’ve had folks use AI tools to help generate SQL queries or even write lightweight code. Just last week, I participated in an onsite where we used a tool called Replit, which is a prototyping tool you can get a license for. We have one at Oscar, but you can also play around with it for free. I was incredibly impressed with how sophisticated this prototyping could be. I was paired with one of our senior directors in engineering, and I turned to him at one point and asked, “Is this the way you code?” There was a coding assistant and all this other stuff, and he said, “Claire, I haven’t written code in a year.” So even engineers are learning new ways of working and adopting new patterns.

To phase into your second question a bit, when I think about skills for product operations or other tech operations-related roles, a few things come to mind. One is problem-solving and pattern recognition. I think one of the things that sets us apart is our ability to recognize patterns in different contexts. For example, maybe you’ve had issues with bugs in one place and, in another, there are stakeholder requests coming in on a queue. It seems like a complicated problem, but I think we’re good at identifying that these are actually the same issues. Others might look at them as different—one’s bugs in Jira, the other’s stakeholder requests in Slack—but we can recognize the underlying pattern.

Another key skill is scalable and long-term thinking. When designing a new process or solution, you’re not just solving for today, but thinking about how it will scale going forward. We’ve had instances on our team where someone identified a recurring issue—like documents that need to be digitized every month or quarter. Instead of just handling it on the fly, they thought ahead, invested a little more time up front, and ensured it wouldn’t continue to drain resources long-term. Thinking about how solutions will scale over time is another thing that sets us apart.

Rukmini Baruah: Thanks, Claire.

Jenny Wanfer Sharing AI Use Cases and Processes That Work Well

Clare Hawthorne: Of course. What processes have you already delegated to AI? What works and what doesn’t? This is one where I’d also love to invite folks to come off the chat. So, Jan, thank you for such a great conversation starter. A few things for me personally—I’ll tell you, I’ve started using it in all sorts of different ways. We’re very lucky to have an internal-based tool that allows us to use it for any professional use case, but we can also pass through sensitive information, so we don’t have to worry about what’s going out to the cloud or training someone else’s AI model.

A few of my favorite uses recently: We were evaluating a vendor solution related to incident management. I asked the AI, “Hey, I’m trying to evaluate these two vendors. Can you give me a starting point checklist for this evaluation?” And guess what—it wasn’t 100%. I still had to modify it and tweak it. It prioritized some things that we didn’t think were important and was missing some things we did think were important. But, oh, can I tell you, it got me 85% of the way there.

My suggestion to you is, whether it’s summarizing meeting notes, highlighting key themes in a survey, or translating something very technical into something non-technical for another audience—those are great use cases. One example: an engineer asked me, “Why does capitalizing software really matter?” I had an accounting definition, but I was like, “Help me give a 3-sentence response to a senior engineer that explains why CapEx matters.” So, translating for different audiences is something I find really useful.

Does anyone else in the audience have ideas on things that have worked well for AI use cases?

Jenny Wanger: I mean, I’d say, continuing on the thread of individual use cases, as opposed to broader ones for the whole team, one thing that Hillary—good! I’m going to mess up her last name—but she emailed out a prompt. She created a little bot called Slacky. What she does is, every time she drafts a Slack message, she feeds it into Slacky, which then formats it. It adds bullet points and just cleans it up so that it’s a much clearer piece of communication.

I still use it since she sent that prompt out. I’ve started using it myself. It’s a very simple thing to integrate into your workflow. Every time I’m about to send out a message—I’ve started using it for emails, too—I just throw it in there, and it cleans it up. It’s usually very prompt to maintain the original tone and language and not try to “AI-fy” it. It’s really excellent.

Chris Butler Sharing Insights on AI Challenges in Summarizing Complex Status Reports

Clare Hawthorne: I love that one, Jenny. 

Chris, I know you mentioned in here that there are places where AI usage isn’t optimal. Do you want to share a bit about your experiences there?

Chris Butler: Yeah, sure. I guess the thing I’ve started to notice is that there was a time when I was trying to help summarize an entire team of PMs’ status reports for a leader who would then pass that on to the next leader up. I just found it to be really error-prone. You have to put in a lot of effort to make sure the AI doesn’t create things or word things in ways that don’t make sense. Also, there’s a bit of shaping that happens in that messaging between leaders, and so… yeah, anyway. That’s something I’ve tried a bunch, but it just didn’t give me the results that were quite right for that type of task.

On the flip side, I do think that AI is great for first drafts. It’s just that if the task requires a lot of factual information, it does need either a lot of prompt engineering or you need to double-check it.

Kiran Sharing How AI Streamlined Guidebook Creation and Proofreading

Clare Hawthorne: No, definitely. Karen, tell us about guidebook creation. What’s the use case there?

Kiran: Okay, so when I was creating guides for my company in my previous job, I had to keep the tone consistent throughout the entire guide. Every time I updated it, I had to think, “What tone did I use before? What terms did I choose?” I couldn’t change the tone for the whole guide. So, what I did was create a prompt in ChatGPT that would maintain the tone for the entire guide. I set it up early on.

I kept a separate folder with ChatGPT prompts just for proofreading the document. Every time I needed to update the guide, I’d type it into ChatGPT, and it would automatically update, proofread it, and give me the exact terms I needed. I’d just copy and paste them into my guidebook. This saved me half an hour or even more of work each time. It helped keep things focused and simple. So, yeah, that’s how I used AI exclusively in that case.

Peter Boersma Explaining His Preference for AI Features Built Into Tools Over Standalone AI Tools

Clare Hawthorne: That’s awesome. Thank you so much for sharing that, and I’m guessing other people are also walking away from this conversation thinking, “Oh, we could use it here, we could use it there.” Peter, you mentioned that you prefer some of the integrated AI features. Can you tell us a little bit more about that?

Peter Boersma: Yes, sure. So, I’ll start by saying I’m not a fan of standalone AI tools that just say, “Hey, just start typing, and we’ll give you some useful answers.” I see too much hallucination and wrong answers happening with those. I’d prefer tool makers of tools that we use anyway to explore where AI features could be helpful within their tools for their use cases, and do it right—not using some generic large language model, but using data trained specifically on their data and not just random articles from the web.

This is why I prefer AI features built into the tools we already use, like Miro’s summarization tools or Copilot in Excel. For example, ServiceNow has plenty of AI features where it suggests proper answers or good responses to request messages. That’s my preferred way of using AI, because it means I don’t have to go out and look for specific AI tools. They come to me through the makers of the tools I’m already using.

Jana Debusk Explaining How AI Enhances Support Ticket Trend Analysis and Actionable Insights

Clare Hawthorne: Love it. Thanks so much, Jana. I think your AI example is really relevant to a lot of folks. Could you tell us a little bit about the support tickets and how you’ve used AI there?

Jana Debusk: Sure. So, I’m one of the people who wouldn’t say I’m the most technical person, which is why I believe in connecting the right people. But I manage both of our help centers. Along with product operations for the product team, there’s also the aspect of managing our help centers. Our support team is now part of the product team, which is a relatively new development. We use Zendesk for support tickets and our help center, but it has some limitations. A major limitation we found is that it doesn’t pull out trends from the textual interactions of support tickets, which is the most important part.

So, I worked with an engineer who’s extremely knowledgeable, and she’s started building an internal tool that recognizes these trends. It’s in a ChatGPT format, but it can pull out trends from these textual interactions, and we don’t have to worry about the PII (Personally Identifiable Information) because it’s an internal tool.

What I think is the most exciting part is not only pulling out those trends, but also having something actionable to do with them. For example, we can point it to Help Center articles, giving it the layout of our Help Center and saying, “What areas should we start with first?” Or, if it’s maybe more related to guides or walkthroughs, we can focus there. Having that information available and directing us to where we’ll be most impactful is, I think, the key part of the AI portion. We’ve done this manually before—I’ve read every single support ticket over the last few months and created my own insights, but that takes too much time. It’s just not efficient.

So that’s where we’re headed with it. Like I said, I got to a certain point, and when we were working on the prompts, I realized, “This doesn’t feel like my expertise anymore.” So I connected with the right people to keep it going.

Clare Hawthorne: Yeah, and we’ve also seen at Oscar, we had a case where someone prototyped in this workshop. I think Sofia might have some other examples in the chat, but they came up with a manual process. They were like, “Oh, we think that the LLMs (large language models) would be great for this.” They brought it back to the scrum team it impacted, and the engineering manager said, “We shouldn’t use AI; we should just build this into the tool.” So I think that we can help surface opportunities, whether it’s for content generation in product operations.

And I do want to jump ahead in the chat because I think the enablement and help use cases are really good ones. So, thank you so much, Jana, for bringing this up. But I also think we can look not just at what AI can do but also if AI can help us identify a feature gap in our tool set or a process that needs to be formalized. Really love that, Jana.

Karim Wagdy Sharing How AI is Used to Create Product Enablement Podcasts for Efficient Knowledge Sharing

Clare Hawthorne: So I’m going to skip ahead. I think Karim brought up a really interesting use case around product enablement podcasts and other supplementary information. Karim, can you share a little bit about what you guys have done?

Karim Wagdy: Yeah, sure. Hi, everyone, this is Karim. I’ve been in Product Ops for around 2 years now. Can you hear me quite well? Yeah, so I’m actually based in Italy, but right now I’m in Egypt, in my apartment in New Capital. Regarding this point, what we’re doing right now is trying to integrate different parts of AI. One of them is AI-focused. We’re doing this release session, and we want to complement or supplement it with something else. We’ve chosen the AI pod. It’s around a 5-minute summary of everything in the release sessions or any sessions we host for people. I believe it does a great job summarizing, and it’s something they can listen to really quickly. Plus, it doesn’t involve anything extra from us to do it.

Dan Mishra Explaining How IBM’s Experience Orchestrator Augments Customer Experience with AI

Clare Hawthorne: That’s amazing. I love that. And a prompt for the audience: I know that a lot of us have been talking about use cases where we’ve used AI to help streamline our work. Does anyone have examples of places where we’ve either augmented a customer experience or augmented an internal team’s workflows, in terms of how they might be leveraging AI to do their work faster? Dan, this might actually be a good segue for you in terms of Gen AI assets and agents. Maybe you can kick it off, and we can see if anyone else has other examples to share.

Dan Mishra (IBM): Sure. So, we have an entire portfolio of 38 offerings, but I’ll focus more on the question you asked. In terms of customer experience, we have something called Experience Orchestrator, which asks about 30 questions to our customer based on what kind of front-end customer experience they want. And with those inputs, we train the model on the customer’s products, brand, ideal target buyers, and their approach. Against that, we give them a blueprint on how they should design their front-end experience. For example, if they’re building a commerce site, we provide guidance.

This is an example of a customer-facing experience manager. You can take it to the next level, too. For instance, if they want a multi-device or single-device experience, they feed those questions in, and the model will give them widgets on how to assemble their website or mobile experience. One of our marquee clients, for example, has us completely build their front-end experience for their flyers.

Sofia Sharing How Oscar’s Product Operations Team Uses AI to Automate Processes and Improve Product Feedback

Clare Hawthorne: Very cool. Thank you so much. Anyone else who has used AI to enable others or scale impact across your tech team?

Sofia: I can share a couple of examples from our product operations team at Oscar. When we think about enabling others, one of our team members, Mayhole—he’s on the call somewhere—was able to leverage AI to heavily automate a previously manual process. We had to take heavy data from a wonky PDF format and translate it into a very formulaic Google Sheet format so it could be ingested into one of our tools. At the time, this was owned by the tech team because it was a complex process. But with AI, Mayhole was able to heavily automate it, cutting the time spent on that task by more than 30%. Because the process was simplified, we were able to transfer the responsibility from the tech team to an operational team, empowering the operational team to own the process themselves. That was a really cool example of enabling others.

We also have another example of improving the product feedback loop. A team member, Kelsey—I’m not sure if she’s on the call—was able to use AI to go through a complex and high-volume help channel and identify trends related to specific tooling issues. On top of that, she used AI as a brainstorming partner to think through potential solutions for those trends and issues in the tooling. So, those are two cool examples of how members of our team have used AI to both enable others and improve the product feedback loop.

Karim Wagdy and Chris Butler Sharing Predictions for 2025: AI Advancements and the Evolving Role of Product Ops

Clare Hawthorne: Cool. Thank you so much, Sofia. I know we’ve been talking a lot about Gen AI. I’m curious if others have any predictions for 2025 and the state of product operations before we wrap up.

Karim Wagdy: I believe I can add something here. With the new generation of AI tools, especially in video, I think this will be a major thing for us to use, especially in Product Ops. In the past few months, I’ve tried the new AI tools from Google and Microsoft. However, they’re not yet available in Europe or anywhere else. But these tools—when you add a certain level of prompts—they can generate a whole video in just a second. I think teams could benefit from this. So, if anyone in the U.S. can try it, maybe share their experience with integrating this into their work.

Clare Hawthorne: Yep, and Chris, I saw in the chat that you mentioned more program management work ensuring coordination rather than improving collaboration. This isn’t the most optimistic prediction for 2025, but do you want to share a little bit about what you’re seeing and why you think it’s going to shake out this way?

Chris Butler: Yeah, I just think that as people and teams become more reactive, with all the uncertainty going on—like layoffs and additional roles being created that require more coordination—we might lose focus on improving collaboration within teams. I’m already seeing this in my role. I think it’s a trend that we’ll unfortunately see more of. And, I hate to say it, but I’m also wondering: What is Product Ops for if we’re just doing that? I have a hard time explaining the difference between Product Ops, TPMs, and program managers, as it is.

Clare Hawthorne: Yeah, I think we’ve seen this on a micro scale at Oscar. We’ve gone through some changes in terms of our program management resourcing. Over the last year, we reduced the size of that team, so we saw an uptick in what we’d call project management within our embedded Product Ops folks. It wasn’t really complicated programs, but we saw some coordination move to the Product Ops team. And for those without a Product Ops team, it often moved to engineers or product managers. So I think everyone’s feeling a bit of that squeeze, to your point, Chris. 

For those with Product Ops teams, I think it’s a natural place for that work to fall, but it’s a choice organizations need to make—how much of that aligns with their goals and where they’re delivering value. At Oscar, we’ve been pretty pragmatic, and we see it as a way to further our vision of advancing the roadmap. But I know this doesn’t always align with the goals of Product Ops teams.

Chris Butler: If I could add something else—I say, “PM’ing the PM experience,” and I know not everyone in the community agrees with that. But I think this is one of those hazards of focusing too much on the end customer rather than the product group or the product itself. I think that once we start viewing the end customer as our customer, we end up doing more of the “bullshit” work that PMs don’t want to do. And that’s my concern—if PMs are stretched even further, we might get more of that work where it’s like, “Can this Product Ops person just handle this for me?”

Jenny Wanger on the Importance of High-Leverage Work and Scalability in Product Ops

Clare Hawthorne: Chris and I have often had lively debates about the end customer focus versus the PM experience focus. So, for those of you who are going to be in the community, stay tuned. Chris and I are two sides of the argument there. That’s true, and I think Jenny wanted to add something too. Oh, yeah, Jenny?

Jenny Wanger: Yeah, I was going to add that one of the risks I see is that the more we spend our time in spaces where we’re not high leverage—where we’re doing more execution work—the more likely, depending on how it’s spun, it could be very high risk. Or, it could just be where things are headed. The other side of it is, “Why do we have a second person on the team doing the work of a product manager when we already have one?” That’s redundant. 

The only way to grow Product Ops and grow our influence is to add more people to do this work across more teams, but that doesn’t fly in this environment. So, I encourage people to focus on tasks that actually unlock value across a minimum number of people but it needs to be scalable—like one person supporting 20 people. Otherwise, if we can’t differentiate ourselves from TPMs or program managers, that’s not good.

Closing Remarks

Clare Hawthorne: I’m getting nudges from Ana. We’re approaching the end of our time, and Karim, you left us with a prompt that I think we should take to the chat. So, Karim, if you could post this in the Product Ops HQ Slack channel about ownership for Product Ops within organizations? I’m assuming you mean reporting lines, organizational design, and structure. I think this is a great reminder that the Product Ops HQ Community is here, and I’d love to keep some of these thoughts going as we move into 2025.

Loved this conversation. Thank you all for bearing with my technical difficulties. Definitely feel free to get in touch with me on LinkedIn—I’m always excited to hear from folks doing Product Operations in different contexts, environments, and countries. Have a wonderful 2025, and I’m excited to see some of you around these virtual communities.

Ana Andrade: Yeah, thank you, Clare. I think you’ve said it all. I would love for all of you to join us in Slack to continue this conversation. You can also get in touch directly with Clare on LinkedIn. And last but not least, I’d love for you to help shape the future of our community by taking our annual survey. It’s still live, and it only takes 5 minutes of your time. I promise! Your feedback will help us uncover trends and best practices that will benefit everyone in the community.

And before we say our final goodbyes, I’m also excited to share that we’ll continue this AI conversation in our upcoming meetup. We’ll dive deeper into how AI is reshaping Product Ops, so stay tuned. The registration page will be launched shortly, and I want you all to register and join us. Thank you once again to everyone for joining us today, and a huge thank you to you, Clare, for leading this fantastic session. See you all next time, and have a wonderful rest of your day. Bye!

Clare Hawthorne: Bye!

Reference

• Product Ops HQ: https://www.productopshq.com
• Dragonboat: https://www.dragonboat.io
• Slack: https://slack.com/
• Oscar Health: https://www.hioscar.com
• Jira: https://www.atlassian.com/software/jira
• Replit: https://replit.com/
• Zendesk: https://www.zendesk.com
• Microsoft Copilot: https://www.microsoft.com