In this episode, we have a fascinating conversation with Marcus O'Brien, VP of Product, AutoCAD, at Autodesk, on the evolution of AutoCAD. Marcus takes us on a journey, discussing how AutoCAD has evolved since the 1980s and how it is now a go-to tool for architects, engineers, and designers around the world. From creating 2D and 3D objects to becoming an extensible platform, Marcus shares insightful details about the product and its wide range of applications.
In this episode, we delve into a fascinating conversation with Marcus O'Brien, VP of Product, AutoCAD, at Autodesk, focusing on the evolution of AutoCAD. Marcus takes us on a journey, discussing how AutoCAD has evolved since the 1980s, establishing itself as a go-to tool for architects, engineers, and designers worldwide. From creating 2D and 3D objects to evolving into an extensible platform, Marcus shares insightful details about the product and its wide range of applications.
We also have the opportunity to hear Marcus's extraordinary personal journey from Ireland to America, and his transition into product management. Marcus enlightens us on the evolution of product management, discussing the industry's macro shifts that have influenced products and the strategies to enter product management today. Additionally, he shares his thoughts on what makes a great product manager, how AI and ML are utilized in product management and his experience in building models for Autodesk's products.
We conclude the episode by exploring the use of LLMs in 3D modeling and design, along with the capabilities of AutoCAD products in generative design. Marcus offers insights into onboarding customers and highlights the available tools for individuals interested in learning 3D modeling and design. Tune in to this insightful episode to learn from an industry expert and explore the world of AutoCAD and product management.
Key Quotes:
Timestamps
(1:45) - The journey of AutoCAD
(7:56) - Marcus’ journey from Ireland to America
(12:05) - Taking the technical route to product management
(19:20) - Bringing GenAI and product management together
(25:26) - LLMs in 3D Modeling and Design
(29:56) - Goal Setting and Adapting in Product Management
(37:15) - Quick hits
Links
Check out the Figuring Things Out podcast
Madhukar: [00:00:00] Hello and welcome to How We Made That App. I'm your host Madhukar Kumar. After starting my career as a developer and then as a product manager, I am now the Chief Marketing Officer at SingleStore. Today I'm thrilled to be joined by Marcus O'Brien, VP of Product AutoCAD at Autodesk. Marcus is an expert in product scaling and understanding advancements in machine learning and AI to bring groundbreaking results.
He also hosts Autodesk, the AutoCAD podcast, as well as co hosts his own tech podcast, Figuring Things Out. Welcome, Marcus.
Marcus: Thanks very much. It's great to be here.
Madhukar: So, Marcus, let's start with a question that I get asked quite a lot. How do you pronounce your name?
Marcus: It's a silent N. Arcus, I go by Arcus to people who know me well, but I'm happy for you to call me Marcus today, if it's easier for you.
It is it's a funny one to ask because being Irish, when I lived in the UK and lived in England, with my accent, for whatever reason, when I would answer the phone and say, hi, it's Marcus here. How, you know, how are you? They would always hear Max. So for the longest time in the UK, people called me Max.
And, and in the office, it was a running joke that my name is Max because I say my name too quickly. So I'll answer to Arcus, Marcus or Max today.
Madhukar: Yeah, and I, and I, you know, started off as a developer, so I see it as a primary key, so as long as I know people are addressing me, I'm fine with it. But I've come across a lot of friends who would call me every other name except my real name, but I'm used to it.
Marcus, tell our audience what AutoCAD is and what is it primarily used for and by who?
Marcus: I mean, AutoCAD is a product that's been around since the 1980s, you know, first released in 1982. It was a real secular shift product [00:02:00] before AutoCAD designed for buildings. You know, it's used by architects, engineers throughout the world.
Fundamentally, it is a It is a key design tool for creating buildings, bridges, infrastructure, products, your watch, your car, these glasses, all of those things are likely have had AutoCAD at some stage in the design process, you know, before AutoCAD in 1982, things were done on, on paper with, they were very, you know, drawing boards, paper.
Iterations were really, really slow. It was very expensive to make changes. And then Audic had really changed the world at that time by, you know, democratizing design tools and making it available on home computers or for office computers. So really reduced the amount of time it takes to iterate and increasing innovation.
So, you know, 42 year old product, still going very, very strong. We've got a lot more products than we did. In the 1980s, as you can imagine, but still targeting the architecture, engineering, construction industry, as well as, as well as the product design and manufacturing industry. [00:03:00]
Madhukar: If the software is from the 80s.
Was it used to design the ruins of the Roman Empire as well?
Marcus: That even predates AutoCAD, I would say. Yep. That is one of those things that was on the old system. If
Madhukar: you believe that we are living in a simulation, I would believe that it was, the ruins were designed in AutoCAD. But do you, do you also call it as, as a Swiss army knife for designing any 3D objects?
Marcus: I've heard it being called that, you know, 2D, 3D. I think the thing about AutoCAD, I think part of its success story is that from the beginning, it's a very buzzword now to be a platform, but AutoCAD was, was kind of, One of the original platforms for design, a very extensible product where people built capabilities on top of it.
And as a result, we have, you know, AutoCAD architecture, AutoCAD mechanical, AutoCAD MEP, AutoCAD electrical, AutoCAD plan 3D. A lot of these vertical focus products that are built on top of the AutoCAD. core platform. And [00:04:00] as a result, you know, customers do refer to it as the Swiss army knife, because depending on what you need AutoCAD to do, there's a specific tool for that.
It's not something I refer to it as, but I've definitely heard it before.
Madhukar: So if it's a Swiss army knife, do you use it to open a bottle of wine at the end of every day?
Marcus: Not every day, not every day, but I don't think that would be healthy. I think when we think back to the 1908, I mean, you probably could with those, the first versions of AutoCAD that came out on nine, three and a half inch floppy disks, you probably could open a bottle of wine with them, but nowadays, no, not so much.
Madhukar: Very good. I, of course, joke, but my son is doing his mechanical engineering and he uses AutoCAD quite a lot. How many users do you currently have? And in terms of ratio, where do you think it's maximum uses? Is it among architecture, mechanical engineering, prototyping? Yeah. What does the overall spread of users look like?
Marcus: Yeah, it's a great question. So we're, we [00:05:00] don't publish our exact monthly active users or user base. What I would say is it is one of the leading products in engineering, architecture and design. So as well as product design and manufacturing large, so, you know, a global use in terms of the profile, US, Europe, a lot of emerging markets using it, Brazil, India, and China.
Big user bases in those places as you'd expect with a product that is, you know, so readily accessible on desktop, web, mobile, et cetera. And I would say if we look at the AEC space, the, the architecture, engineering, construction space, I think that's our primary use case right now. We've probably got.
More than half of our users in that industry. And then the other half, or just less than that, identifying as being product design and manufacturing. So, so your son would be in the PDMS, the product design manufacturing space that we talk about as a mechanical engineer.
Madhukar: And we will get to AI and how that is affecting, but I'm assuming to use AutoCAD, you'd need [00:06:00] training.
It's not something that you just wake up one morning and then you understand how to use it and. Start building designs.
Marcus: Yeah, it's probably somewhere in between. You know, if I think about, if I think back to AutoCAD certainly legacy AutoCAD, 1980s, 1990s probably a steep learning curve. One of the things that we do to help with that steep learning curve is that we make AutoCAD available to university students for free.
So that during that steep learning curve, they can actually learn how to use the product. And by the time they get into the market, that they are effective at using it. But I also think with advancements, we're going to talk about AIML and, and You know, just general experience enhancements in the product over time.
We're not talking about a product from the nineties anymore. We have very, very modern, very, very cutting edge capabilities in the product. And I think part of that is onboarding is, is feeding information to customers at the right time. We do have specific capabilities targeted at people to help educate them, use the product as they go.
AIML capabilities that will actually. You know, watch the way you do things [00:07:00] and then offer up information to you around how you might do that better, how you can create macros around it and teach you how to do it as you go. So, I think that steep learning curve is actually flattening over time. It's, I mean, considerably, the slope of that curve has flattened over time.
And we're continuing to do that with the different products we deliver today.
Madhukar: That is very cool. That sounds exactly like product led growth and the next version of product led growth, which I talk about quite a lot. You and I were talking about this prior to the show as well. Tell us a little bit of your journey from Ireland to London to America.
How you got here, how you got into AutoCAD and design.
Marcus: So I moved, I was, you know, lived in Ireland till I was like 30. So I did mechanics, same, same as your son, mechanical engineering undergraduate. After graduating, I went and joined Arup, who are a international design engineering design consultancy.
And I worked on designing buildings, designing art galleries and airport, a railway station, various infrastructure projects like that. afTer a while, I, I stayed there for [00:08:00] seven or eight years. I moved to the UK. My wife was doing a master's. We went to Oxford for her to do her master's. And I joined a really interesting company who was doing, they were doing data centers at the time.
This was 2010, Amazon web services was not as big as it is today. There was a big arms race in terms of everyone was building their own data centers, and there was a huge market for it. And I worked on a really innovative. product, where it was about manufacturing a data center that could be shipped anywhere in the world.
And it would be really energy efficient, have very low PUEs, a power usage effectiveness, which was, which was a big buzzword at that time. I worked there for three years, but during my time designing these data centers, the types of customers I worked for, as you can imagine, were a lot of tech customers, it was, I can't name them on this podcast, but you can imagine who the data center players are and were at that time.
So I decided I really wanted to actually go onto that side of the fence, be one of those people, as opposed to being the engineer designing the systems. I went back and did an MBA in 2015. I went to London business [00:09:00] school. I was older for an MBA. I was 32 when I did it. I did it full time. So it was a big opportunity cost.
And afterwards, me and my wife were looking to move to the US. My wife is American and I had a couple of job opportunities. And one of them was at Autodesk to be a PM on AutoCAD. And I took this job and. I moved here in 2015 and did the NBA 2013 to 2015. Moved here 2015, and I've been at Autodesk at the auto on the AutoCAD team ever since.
So I would say, you know, the move to the move to the uk Yeah, I'm, you know, maybe, maybe we're gonna talk about it a little bit, but I think, yeah, moving to the US has been, has been fantastic and I, I've loved the journey.
Madhukar: What was your biggest cultural shock or thing that you noticed as soon as you got to the US the first time?
Marcus: I was expecting a massive cultural shock. Maybe it was because I'm married to an American. By the time I moved here, I was married to an American and she had rounded off some of my edges. I think, I think, I think the irony here is when I moved from Ireland to London or to Oxford, I moved [00:10:00] 400 miles, you know, you know, Ireland is an island.
It's different to the UK for, for some folks. And I moved there and I was expecting things to be really similar. And I wasn't expecting to have the amount of culture shock that I did have moving to the UK. So I found that a big adjustment living and working in the UK for five years. Just how different our sense of humors are, how different work, work is and how, how, you know, just some significant cultural differences.
So when I signed up to move to America, I was kind of expecting, I was like, wow, this is 8, 000 miles, it's going to be like this, but 10 times. And I really didn't feel like that. And I think it's, I think it's because I moved to California, honestly. And you know, the team that I immediately worked with and everybody who was, you know, all the friends I was making when I first moved here, actually not many of them were from California.
So we were all having these, it was like this, just this kind of, this soup bowl of everybody bringing a different ingredient. And it wasn't really that I was joining some other culture. It was about kind of creating culture together. So not so bad, honestly, as far as I, as far as I remember.
Madhukar: The one thing [00:11:00] that I noticed was just the portion of food sizes when you go to a restaurant and I was so impressed.
I was like, wow, you get so much for so little, and then you get to carry that food back home too. So I love that.
Marcus: Did you put on your, what do they call it? They call it the US 15 or something. Did you put, how many pounds did you manage to put on in your first year? It's a thing for sure, you know? Yeah,
Madhukar: I think I must have put on at least 50 pounds.
Marcus: Yeah, that's, that's about right. I think you did it properly. So, yeah.
Madhukar: Tell me a little bit more about how you got into product management. Actually, I do see a lot of hope for my son as well, because he's doing mechanical engineering. And I do see that he could become a podcaster like you.
Marcus: Yeah, you set the dream at some stage, get out there and be a podcaster.
That's I mean, I think, I think mechanical engineering is a great path into product management. I encourage technical people to be PMs if they're interested in the business aspects as well. I think, I think going through the technical route and then getting into [00:12:00] product management later is a. Is a really strong foundation in being able to understand some technical engineering concepts.
And then you can kind of scale yourself, learn a bit about strategy, but, but be rooted in the technical side, I think is one of the things that makes you really successful. I think the other thing is, I don't know how you feel about it. If you've noticed it yourself, but I think when I look at founders, if I look at all the VC investment that's happening at the moment, it's for a more technical founder base.
So I think the Wild West of, you can just go to a. to a VC and, you know, you're, you've got a business plan and you can talk the talk. I think those days might be over now, unless your, your company name ends in AI, but there, there tends to be more of a, of a technical bias to, to these positions now. So I think anyone coming in with a technical background and then switching to PM, it's a good route.
So for my, for myself, you, you, your question was how did I get into it? You know, I, that was the route I took. I was, I was technical for a long time, you know, I did all the really useful languages like Fortran when I was at [00:13:00] University , you know, which was a dead language at the time. I learned it. I did some c plus plus, things like that.
But, you know, I'm never gonna be a developer. I'm never gonna be very good at that. But I have it as a foundation, and I think that that was a good start. You know, so
Madhukar: if you were to. Compared and contrast when you joined Autodesk and you took AutoCAD product management role to where it is now, what are some of the big differences, both in terms of the product, but also the way we do product management and we do engineering?
Marcus: That's a great question because there really has been a massive evolution. I think when I joined, even in 2015, PM was not a very well understood. Discipline. It was somewhere between program management. Some developers thought you were program managers and you were just responsible for a schedule, which you're not.
Some people said you're the CEO of the product and that you're responsible for all aspects of it, which you're not. You know, I, I, I don't line up with that one. I think, I think it's really well understood now. I think things like Marty Kagan setting the [00:14:00] standard for, for what PM is and people understanding it from a, you know, as a, as a career.
Path, I think is well defined now. I think, I think the FANG companies have done a really good job of standardizing on it with the different levels of PM. What's it mean to be a PM or a senior PM or a group PM or a director of PM? Like they're all really well understood now. And that had to happen over time.
In terms of the product changes that happened. I think there's been a lot of like macro industry shifts that happened. So, you know, when you and I grew up cell phones were not some, well, it certainly wasn't something that I had when I was a teenager, you know, the iPhone came out in my late 30 or in my, in my late twenties, you know, and, you know, those paradigm shifts that happened, at a macro level have changed products. So when I look at AutoCAD's journey, the first 20 years was about, was about building automations on desktop software. Maybe, maybe, maybe, you know, after 20 years for the next 10 years was about acquiring vertical products or building vertical products and bringing them to [00:15:00] market to target specific niches.
From 2015 on, you know, from 20, from 2010, maybe to 2018 was more about multi platform, about creating AutoCAD that is truly everywhere, where it's desktop, web, mobile. If you've got, we've got AutoCAD design automation API in the clouds is that if you want to run. Automations, if you don't want to use your GPU, if you want to do things online with servers, we've developed this full third party ecosystem of developers who develop capabilities on top of AutoCAD.
I think that was the kind of push. And certainly this last number of years for PMs, it's been about machine learning and AI, about really. Taking out repetitive tasks making smarter automations, not just automations that make you faster, automations that teach you things and help you design things in a way that are different than maybe you had thought going into a, into a concept design.
All of those things really shifted both the PM. Discipline and the products that we shipped as a result.
Madhukar: So if somebody looking [00:16:00] to get into product management today, what would you advise them other than having a big, massive ego?
Marcus: It helps. It, I feel you're, you're 70 percent of the way there. I, I, do you think PMs with big egos do well?
What's your question? I mean, what's your experience
Madhukar: on that one? Yeah. I didn't comment about that. I just, I just talk about, you know, what it takes to become a PM. But of course, I joke. I've, I've, I've been asked this question quite a bit. Like, how do you become a product manager? You know, there is no product management school.
The last I checked, it's not offered.
Marcus: There are some you can do when at Berkeley, they have a, they have a product management course at Berkeley, and there's a number of strategic courses you could take, I think you need to learn it on the job, if I'm honest, maybe I'm a bit old school like that, I would push back on the ego and I actually think the most successful product managers are humble.
And I think that is one of the qualities you look for. You want [00:17:00] table stakes, you need the smartest person, you know, super smart people. My personal preference is a strong bias for action. So somebody who doesn't have to have the idea, but is the person who wants to get traction and make progress with the idea.
Incredible communication skills, both written and verbal. You have to be one of those people who just enjoys it. There's people who can learn to communicate, but you gotta be one of those people who likes talking to people because you know what, you're going to have to say the same thing over and over and over again to different people.
Whether you're trying to convince the engineering team, you're trying to convince customers, you're trying to understand customers, you're trying to convince your stakeholders to invest within the company, you're saying the same thing over and over and over again, you gotta love it. You got to love doing that sort of stuff in order, in order to be successful.
I think the people oftentimes who have the major egos or, or, or are too hung up on a single idea, like this is the way we need to go. This is true North. It's great having a true North. It's great having a sense of direction that can be very powerful, but sometimes you got to disagree and commit as well.
[00:18:00] Or, you know, you need to flip flop. You need to be wrong. You need to admit that, okay, that wasn't the way to go. And I think, I think that takes a special skill set too, to be able to. to roll with that. Yeah, absolutely.
Madhukar: In my, in my experience, PMs with a very big ego eventually end up being industry analysts and earn lots and lots more money.
Marcus: That's right. Okay. Yeah.
Madhukar: Yeah. Let's, let's talk about AI because you wrote a paper about it, I think on LinkedIn a few months back. How did you get into generative AI or? Yeah, in general, how does it relate to product management today? Where do you see this going in the future?
Marcus: Good question. There's a lot in there.
I, a few years ago at Autodesk, I was lucky enough to be in a position where we decided to invest in machine learning. We, we, you know, I kind of put a pitch forward that we, that the next set of capabilities we need to be were involved a ML AI. And I think there's a lot of confusion. We can talk about it in a [00:19:00] bit about what that is.
Is it generative AI? Is it LLMs? Is it, you know, what are we talking about there? But for me, it's not so much the technology. It's more about the smart capabilities to help customers do more. And we made a, we made a, an investment in terms of retooling. We had to do things like we hired a bunch of new people to do it.
We also gave people internally opportunities to, to retrain and to change over, change positions, to be able to do this new work we wanted to do. We needed a better understanding of data, both, you know, we've already, you know, very careful since GDPR, et cetera. We're very careful with data, with customer data, but like.
Data pipelines, what you can and can't do with data is very, very important. And we're very, very strict about that. Labeling of data, getting clean data. I think a lot of people don't realize that to train an ML model, you need clean data. It's not just data. We need to know what that data is that you're training it on.
And that can be a very manual process, a very slow process to start. So we had to get better and more data savvy, understand our data better. We had to understand how to build ML models [00:20:00] better. We had to. You know, build, measure, learn in a true sense and see, see, these models live into existence that they go on into perpetuity.
And then we need to know what did it mean to maintain these models? Cause there is such a thing as model drift and a model that's working really well today over time may not be working as well. So understanding all of those capabilities, all those skillsets we needed required us to retrain, not only our development efforts, but how we thought about analytics, how we thought about data, how we thought about product management, how we thought about delivering capabilities across.
Multiple years, you know, these are, these are major models that we work on. So very interesting changes, something I felt really passionate about and was very lucky to be in on the ground floor and kind of involved in, in in creating this at Autodesk.
Madhukar: So did you end up building your own models for your own products, which is based off.
Retraining or fine-tuning existing large language models, or did you just build it from scratch?
Marcus: So talking about LLM specifically, so, you know, we, on, on the scale of things, just to make [00:21:00] sure we're on the same page, there's like, there's a lot of solved problems in machine learning today, you know, Google has been working on raster recognition for years.
You know, you, you show up, you have a photo. Is this a cat? Is this not a cat? Is this a hot dog or is it not a hot dog? You know, there's entire apps based on it, but it's a solved problem. Raster recognition where you have an image and decoding what's in that image is a solved problem. So if we want to do raster work, it probably wouldn't make sense for us to build our own raster engine.
Those are the types of places where you just want to partner with an API. Just use an existing public API, open source API even, and leverage that. Same for object character recognition. If you're doing things like recognizing text or font. Handwriting, things like that, it makes sense to use third party or open source.
But there are things around, you've talked about LLMs, there's an in between as well which is vector recognition. And by vector recognition I mean, I work in the design industry where vectors are lines that have start point, end points, orientations in space, thicknesses. And while there's an A wealth of, of well [00:22:00] understood solved for problems with raster images.
There is not that same level of, of work that's been done on vector. And that's something where we've been investing, something where we have been creating our own models, our own capabilities and training our own data on things like that, because we feel like that is a unique advantage we have that we can offer to our customers where it doesn't exist in the industry today.
Talk, going back to your question about LLMs. You know, that's another paradigm, something else that, that, you know, integrating chat GPT or BARD or whatever your flavor of LLM is into products is definitely something that I think a lot of companies are looking at right now as a way to check the box on AI.
I think it's possibly one of the less exciting things. I think if your company is solely reliant on LLMs to check your AI ML capabilities, you're probably missing a beat. I think, I think the companies that are looking at more broadly beyond LLMs, maybe have a little bit more strategic advantage and more value to offer to customers ultimately.
Madhukar: Makes sense. So, let [00:23:00] me go back to product management a little bit and then we'll come back to LLMs and 3D modeling and such. When you go back to product management, you know, the old school way, and I did this quite a lot as well, is to go talk to customers and constantly be in sales call, but also be in support calls and just gather feedback.
A lot of that now has also been just telemetry, where you collect all of that information, although anonymously in GDPR to get a, you know, get a feel for what exactly the customer is doing. Have you seen that change? And do you think LLMs being used in that to then surface insights of what features to double down on and what features to De prioritize.
.
Marcus: You know, on this one, I I can't really talk about our own products, but what I can say is that I, I have seen that shift in the industry where, you know, customer sentiment is, is aggregated and you know, models do that for you [00:24:00] now net promoter score is more automated. Like there was a time when I first started that NPS was, you know, NPS for folks is your, how happy customers are with your product or not, was something that was a report that was generated once a year and you would buy it off a third party vendor or something like that, like, you know, things are more real time now.
We can check on specific capabilities, specific actions, et cetera. And I, and I see that in other products today. And
Madhukar: now switching over to the LLMs inside of 3D modeling or design, do you see LLMs being used to build new design based on, you know, vast amount of previous design data that already exists?
Like, would somebody be able to just go into a. Tool either today or in the future and say, draw me a 3D model of a monument or something like that. Or of the
Marcus: Roman, the Roman ruins create the Roman ruins for the fifth Roman. Yeah, exactly. So the short answer is. You can't do that without the underlying technology in place.
So [00:25:00] what what an LLM is good at is, is understanding context. It's good at training on like a great use for, for LLMs would be for product support cases or for help or for training. You're not, you, you said, you know, how difficult is it to use AutoCAD a way of onboarding a customer based on 40 years of documents that have been created on how to use the product and what buttons do what, you know, LLMs are great for that.
Because you can train it on existing content that's available. And, but for what you're talking about, about doing design, you know, generative design basically is what we call it and offering designs up to customers and helping them in their design process. It's something that we've done at Autodesk for years and various products where we do generative design.
We, we, you know, we have a great product called Forma, which is on the market now, which will model designs for you of buildings and give you lots of different options as well as modeling things like sunlight, heat gains noise levels, all of that sort of stuff. So we have those capabilities in all of this today, but, but just to step back from the actual [00:26:00] products and say, like, from a technology point of view, the product needs to be smart enough to understand raster, to understand commands, to understand inputs.
And those individual little ML models that we have in place that we've been training for the last couple of years. Are the things that when you layer them with other capabilities, like LLMs, et cetera, then you can do more. So we have a capability in Audica, just to give you an example today. Where if you're using AutoCAD and we see you do three or four things, supposing you're designing a building and you put a window in and we see that to draw the window, you draw a square, you draw another square inside it, you put a little blind on it, not, you know, we see you do that and we'll say, you know, we've noticed that you do these things over and over to create this window.
Do you know you could create a macro for that? So we'll put a little insight in and we'll say, instead of doing the eight things that you did to create that window, why not use this macro instead? And that saves. That user, instead of doing eight commands, eight times, we create that macro for them where we're constantly training these [00:27:00] macros in ML to make sure that they're smarter, better that we're not just interrupting workflows, that they're done in the right way.
And we're not, you know, clippy, the old Microsoft clippy, but you know, those are the things that are really useful in the context. It's all about in context of the workflow that you're in offering up faster, better ways of doing designs you know. They're definitely the first steps in, in what we're talking about here.
Madhukar: So somebody who's looking to learn just again, design, 3d modeling. And if they were looking to just get their hands on AutoCAD and learn it on their own, what would your advice be? Just go to YouTube, watch videos, like what would that process look like for somebody starting scratch?
Marcus: I think there's a bunch of free courses, honestly.
I think you would go to YouTube. Autodesk creates a lot of content on how to use specific capabilities. We have a lot of. We have a lot of capabilities built into the product, so if we, you know, when you, when you open on a start tab, we've got some learning capabilities. I'd probably start if you were looking from [00:28:00] scratch and you were like, I want to learn AutoCAD.
Now I've no, no experience at all in AutoCAD. I would use the web app first. It's a very lightweight. WebAssembly product web. autocad. com, fantastic product. Or if you prefer iPad or Android, do it on your iPad. Do it, use the mobile products that we have, but they're, they're a lighter weight version of our desktop product because, you know, GPUs are smaller, et cetera, and, and we're trying to build light.
products for the web, but, but, you know, one of the things that was said at the Google I O conference, when we, when we launched the web product was it's the most robust desktop software running in the cloud. So, it really impressive AutoCAD very, very small size caches in your browser and a very good way to learn how to use AutoCAD from a basics, the ground up approach.
Madhukar: And going back to, you know, people looking to also become product managers People who are learning to become product managers, what, what, going back to maybe your team, how do you define goals [00:29:00] for your team? Like if I'm a product manager and I was working in your team, what would my goal be? Would it be more product adoption?
Would it be to get more users into the product? Would it be to do more releases? How do you measure product management in your work?
Marcus: Yeah, great question. So luckily these days I'm in a position where I have one head of product who works for me, a director of product. And she is responsible for managing the entire product management org.
I used to do that role up until a couple of years ago, but, but that is very much in her view. When I go back to a couple of years ago about just running the PM team I would always, I think, I think you need different PMs for different gigs. We've got about 15 or 16 different products with AutoCAD in the name now that we manage.
And whether it's a, you know, developing a desktop product, developing one of the vertical products, developing a web product, developing a a developer product, you know, for in the cloud, creating APIs for third party developers require different skill sets because you're talking about. different shipping [00:30:00] cadences, like do you, does it go, do you update it every day, every week, every month, every quarter?
Each of these products has a different shipping cadence. And I think as a result of that cadence, there's a different level of what's done is done expectation. So if you've got something that ships every quarter, I think you need to have higher levels of testing and getting it right. If you have a product that you update every day, I think there's more room for you know, dual track agile development where you're doing experiments in the morning and you're, you're kind of making decisions in the evening.
And so, so depending on which role you're in, I think there's, I think it's a good one to rotate, if I'm honest, I think the people who are best at this get an opportunity to do all of those things. Cause they're all different and there's no. One right approach, like going back to the thing, if you're dogged on, this is the one way you do PM and this is how you be the best PM, it's because you haven't experienced the other aspects of product management as well.
And it's just going to take you a little bit more time to get there. So, in terms of setting goals for my team, what I always did was I'm a big fan of at the start of each year, we create our own OKRs, our personal objectives [00:31:00] and key results. Objectives and key results is a kind of a well understood Google framework around setting goals for products.
You usually do it on a quarterly basis. The leadership team creates OKRs and they cascade down through the org. So everybody has an OKR that lines up with the leadership one. But we do that on a personal level as well. And they're not published. They're just individual, but it's like this year, what's your objective for this year?
And I usually do a present tense statement in November. So in January, we say, what are you going to say to me in November? That, that you can say in the present tense is true. Something like I'm a really well recognized senior product manager who has had success releasing or doing a zero to one new product release.
And then you go, okay, well, you want to be able to say that in November or in January now, what are the three key results we need to work on to get you there? And so it's a really good way of defining like, well, we need to look for these opportunities where you can work on a, on an. Incubation project or creating some new capabilities or [00:32:00] some new features for customers.
We need to make sure you're better known. There's a recognized part in there. So what are you doing about networking or how is your org savviness, who you're having meetings with? And it really kind of teases those things apart, but I love doing the one year timeframe present tense, because it really makes it really like, we're going to say this in November.
Let's see how close to actually being true about this statement we are when we get there and it, and it makes it more tangible. Well, good
Madhukar: thing I'm not in your team, because if you were to ask me to write it, I would be writing, I am a millionaire. I just got a 200 percent raise and my boss just loves what I do.
Marcus: Let's break it down. We just need three things. We can do that.
Madhukar: Going back to building products, of course, you're a creator. You've been building products all your life. If you were to Start from scratch today and build your own dream product. What would that be?
Marcus: I do think right [00:33:00] now, so, you know, this is a, this is, this is a longer answer that maybe, that maybe you want.
But when I think about the last couple of years, I tried to get excited about NFTs. I didn't get it. But I was like, there's something about blockchain and cryptocurrency and Ethereum. And, and there's something there, there's value there. I don't think bored apes are it, but I think there's value there. I just don't get it.
And then there was the whole web three. It's actually web three is the value. It's the, the decentralization, the democratization of the internet, you know, Google and, and Facebook, et cetera, one for too long. And it's about democ But it was very hard to get excited by it. There was a lot of people getting excited by that and felt like this is the new shift.
And it's like, Is it though, is it the shift? I think, I mean, you know, this is not going to blow your mind, but, but I do think what's happening right now with AI ML and this, this must be what it was like. In the late nineties, I mean, you know, I'd [00:34:00] speak to your experience on this, but, but it feels like this is, there has not been an opportunity like this since the late nineties, I would say, for there to be a true one way secular paradigm shift, whatever you want to call it, where things are going to be done differently.
So if I was a PM and, or I was starting to be a PM now, I think there's a couple of things that we need to think about here. Like, I think. There's, there's the, like, from the kid's point of view, as a parent, I've got two kids, and I think that the way that I raise my kids needs to be different now because I need them to be comfortable with working with AIs.
I think that that's going to be their, their childhood. They're going to grow up with AIs. We, just in the last few weeks alone, Co pilot was announced for Microsoft where they're integrating chat GPT across all their product offerings. So you don't have to open a Word document to create a CV anymore.
Like we did, like you just, you say, you, you, you use co pilot to do it with you. Google has done the same with Duet. They announced it like three or four weeks ago across their workspace products. [00:35:00] Just yesterday, Google announced BARD would be integrated into, or was it this morning? It was this morning, Google announced BARD was now integrated in search.
So when you search BARD, you can use BARD and it will be, they're one in the same thing. So like, I think we have a role to play in training, you know, teaching our kids how to. How to get the best from AI in the way that we had to learn how to use iPads. They're going to have to learn how to, to work with AIs and they'll take it for granted.
But I also think tech leaders, I think, I think people coming out of university will get it faster that things are not being done the way they used to be done anymore. And I think it's more likely you and I and other people who have done it a certain way need to be flexible enough to realize that that's not necessarily what's going to win in future.
And we have to be flexible, fast moving and adapt. Well, to this change, because as I kind of alluded to earlier, if your product strategy is, I'm just going to integrate an LLM into my product, you're probably not doing enough there. It might be enough for an earnings call, [00:36:00] but I, but I think there's more value in doing true ML and, and really changing how you develop products for customers.
Madhukar: All right. So we are now in our quick hits section. So I'm just going to ask you four or five quick questions and you have to answer without thinking.
Marcus: Oh, no problem. I do that all the time.
Madhukar: Me too. That's why I have this section. So the first one, how do you explain your job to someone who is Not in technology or app development, for example, to your parents or your loved ones.
Marcus: Or even mywife, I haven't quite got there yet. She's cheap. My wife definitely doesn't know what I do. I have, as much as I've said, PMs aren't the CEO of a product. I have used that analogy for my parents and being like, you're kind of responsible for everything. You don't have the, you don't have the like, people don't report to you directly. So you've got to be much more convincing and persuasive, but it is something on the lines of you are, you are the, the glue that keeps everyone together, the gap filler, the you know, the coffee [00:37:00] boy, all of those things, all wrapped into one. It's not as glamorous as getting to be the CEO, but it, but you do get exposed to all parts of the business.I like the coffee boy..
Madhukar: That my parents get immediately. Yeah. You also host your own podcast, the AutoCAD podcast and figuring things out. Where can people listen to these podcasts?
Marcus: Yeah, so Autodesk, we have our I'll, I'll send a link if so that we can put it in the description. So Autodesk, we, we ship our own AutoCAD podcast where we talk about all things ML, AI, all the new capabilities that we have in AutoCAD.
So some of the things I touched on here, we go into much more detail. I host some of those, but I also, some of the PMs on my staff host them as well. So there's a lot of value there. The other one is a personal project, figuring things out where I, me and my co host Rami Banna who works at Stripe he's a senior product leader at Stripe.
He, we talk about all things technology. We're just passionate about it. And, and as these cycles come in and out and as things get hot and new, we like to talk about it on a Wednesday evening. So I'll put a link to that as well. Very good. [00:38:00]
Madhukar: What according to you is the best on site conference to attend to better understand AI?
Yeah. Absolutely. And the right answer is SingleStore Now
Marcus: next week. I heard Single Store Now has got a conference next week. I will be at Single Store Now next Tuesday. I'm looking forward to it. I'm not there's, I, I'm actually going to a couple, even in the next few weeks, there's the AI Pioneers Conference.
I've got the AIX Business Summit. I'm very excited about Single Store Now as well. But, but honestly, I think there's a, you know, just putting it out there. I think the things that we talk about at Autodesk University as well are really impressive. Anything with Mike Haley, who is our AI guru at Autodesk. I think anything he has to say about it, people should really listen because he's, he, he knows what he's talking about.
So. Is he also on YouTube? he, there's a couple of podcasts that he's on on YouTube. Yeah, we can put a link into one below.
Madhukar: What app do you use the most every day?
Marcus: What app do I use the most every day? I've got, this is going to when I look at [00:39:00] my home screen, so I'll put it. I'm happy to share. I've got I know most people won't do this on video, but I've got on the bottom in my iPhone, the place I save, I've got text messages, camera, macro factor for tracking calories.
We talked about putting on the U S 15 pounds. That's to help me with that. And then chat GPT. So I do subscribe personally to ChatGPT. I use ChatGPT for all the time. I use it like many times a day to help me with all sorts of things. So it's definitely one of the most used apps I have.
Madhukar: And the last question, what do you do outside of work and outside of product management and outside of podcasts?
Marcus: Podcast is a passion project for sure. I got little kids. They, they take up most of my waking hours and a lot of my sleeping hours. A real passion of mine is, is, is building stuff. I make, I like to renovate old motorcycles. Right now I'm working on a 1979 BMW. R80 slash seven, an old BMW Airhead motorcycle.
I've taken it apart. I mean, literally apart, every single nut and bolt [00:40:00] into every single piece individually, cleaned all of those little pieces and I'm meticulously putting it all back together again. So there's definitely some OCD slash control freak stuff going on there, but it's, I see it as like Legos for grownups.
Madhukar:Very cool. Well, thank you so much, Marcus, for spending time with us and our audience. And I really appreciate. Just you dedicating your time to talk about product management and the product and AI in general
Marcus:It's been really great talking to you today. I really appreciate it. And I'm glad you got the pronunciation of my name right in the end.
Thanks very much for that. It means a lot to me. I think so. All right. See you. All right. Good luck.