Martin Miller: The Art and Science of Building a Successful Startup

This transcription is AI-generated and may contain errors or inaccuracies.

Matt

My name is Matt and I will be talking with Martin Miller about bridging the gap between teams and keys to startup success. Martin, I don't like the long intros and I always like to get the nitty gritty details and your lessons learned and your experience. So the first question that I really wanted to ask you because you work for quite a while in tech, so you have experience on a bigger organization or startups. So you've been there, done that multiple times. And from the startup perspective, I have a feeling that having the non technical founders, it's really hard for them to work with tech teams. Right. So I'm just wondering what is your experience here?

What are the pros cons? How do you approach it?

Martin Miller

Oh, great question. So I work with a lot of non technical founders along with technical founders and I myself being a technical founder, been doing this for more than a couple days, maybe even a couple decades or more. You know, it really breaks down to some simple things to help a non technical founder come up with is understanding what the concept of collaboration is with technical people and non technical people. You know, you got the different Personas and different type of personality types and how they like to communicate and get information. Obviously that communication channel is so important. Voice goes a long ways. But writing, written communication, how we track things, how we match things to expectations, which brings up the next part of that journey is expectations.

You know, there's a dream, I, I want to build this, I want to make this, I want to make something better and how do I get there? Can I go step by step and set up some realistic expectations? Do I, do I just shoot for the moon? I mean, I'm not going to launch a rocket in 30 days without having a rocket in hand. So let's, let's, you know, set up some real expectations for how to get there, you know, and then also do you have the right people and you're in your own skill set to, to get there and do you understand what you don't know? And understanding what you don't know is super powerful because you don't pretend to be an expert in an area, you have no domain expertise. Which leads into people coming to me looking for domain experts for various modalities and whether it's an optimization for a manufacturing or it's developing an E commerce strategy from a technology perspective, an ad delivery platform, a video delivery platform, you come to the right expertise and people that have been there, done that, or can they adapt something new based on prior knowledge of how to do other puzzle pieces.

That's really important. And you know, in that journey, you also want to build respect. You know, I talk about respect on my own podcast in artificial intelligence and delivery, just to kind of put that out there. And then you're always learning and adapting. I mean, as a founder, as an executive, as a lifelong learner, you just have to be able to adapt. And change is the number one rule, regardless of it's a startup or It's a Fortune 10 company by any measure of a doubt. And then how do you work on partnerships and, and delivering to those partnerships?

I think that's a, that's a quick one for you.

Matt

Martin. And another question that I really wanted to ask you, and this is what really caught my attention while I was doing my research on you. So I was like, you know, looking around the Internet and finding everything about you. So I don't be, don't be scared about it, but the thing is that I found on the label. So you worked for Levy Strauss, right?

Martin Miller

So, yeah, Levi Strauss.

Matt

Yes, sorry, Levi Strauss. And like, in my mind, those are the genes, right? Like, long story short, jeans complain. And you, you are an expert in AI machine learning and you worked for those guys in the field of AI and machine learning. And I tried to connect those spouses, but it was really difficult for me. So maybe could you elaborate on that? So what those two things have in common and brief overview on what have you been working?

Martin Miller

Well, let's, let's, you know, I have a book on artificial intelligence in a weekend, an executive sky that I go over, you know, kind of how someone would approach problem solve. Let's, let's connect the dots for Levi Strauss or just be any company that's in retail or manufacturing. Because if you think about it, what is a clothing. Clothing company, if they source and build their own product, they deliver it, they forecast it, they plan it, and, you know, track it. Right. And this is all elements that have data associated with. So artificial intelligence and machine learning are a lot like blue jeans, if you think about it, or believe it or not, you got to think about what does it take to make this, this pair of denim items, whether they have rivets or unrivited, which is actually the name of my podcast, Just doing that selfish plug.

And genes need to be broken in, just like AI, you know, artificial intelligence. You don't just build it, walk away from it. You. You build it and maintain it. You adjust the training of it. The training is how machines learn. I just want to put a statement out here for the audience that machines don't really know.

Machines learn and they apply patterns from learning. And so unlike a human, where humans can know something, machines don't know it. Just keep that, keep that in the back of your mind. Now let's come back to the, to how Levi makes sense here. You know, it's, it's a retail company, it's has its own direct consumer. It also has its wholesale business. And there's a lot around how you monetize the, the build and optimize the build a product and whether that's the sourcing of the threads or the, the raw materials and how you forecast.

I'm going to build so many pieces of this item number or SKU as we call in the US or upc, you know, the code that associates with it. You know, each item on size has a new number. So these are numbers. And what are machines really good with? They're really good with, you know, doing repeatable things. And so we want to be able to look at numbers objectively and be able to do repeatable actions. Like I want to forecast based on the geography, what kind of sizing patterns might be best for that geography.

I'm not going to ship a bunch of skinny jeans to the state of Missouri in the US where in California I may have more skinny jeans. I'm just using the demographic information. We may have data sets on demographics and you know, the demographics may say the population is generally taller or heavier. You know, you can do this kind of adjustment without a clipboard, without an Excel spreadsheet, and let the machine do the talking. And you know, it's not just beauty that we, we can discriminate against for machine learning. We can actually use types of material, you know, and, and do analysis on the material. We can look at color, texture.

There's lots of fun things machines can do. It's not necessarily that unique to a riveted or unrivited company with denim, but it, it does apply to all sectors that build, build a product.

Matt

And another question that I wanted to ask because you have a lot of different, I would say, approaches, routines. So you did a lot of stuff. And I'm always looking for something that could be inspirational for other tech leaders. So if you could share some not obvious things in tech that works for you, but maybe are not a common knowledge and are not really widely used. So what would it be?

Martin Miller

I'm coming in a little facetious and I might have to explain that word to the audience, but I'll just leave it at this. That common knowledge as you might think of it is not common. It's your frame of reference, not necessarily the person you're communicating.

Frame of reference. So when it comes to non obvious tech types, you know, you start saying what's the best way to keep your computer awake during a, you know, a video meeting with, with your team. And for some people they, they have the mobile device and I hold mine up and you know, you don't see them, they put it in front of their monitor and they're watching YouTube cat videos while they're on mute because that's how they can stay awake during the meeting. And I kid you not that, you know, attention, attention span, the alleged focus or lack of focus due to multitasking can be all over the place. So don't take it for granted that people are truly paying attention. You want to have meetings with purpose. That, that, that seems like a given and you want to be careful about.

How do I want to say this? Tech doesn't solve all problems. Human communications can help you solve those problems on its own. Use tech enablement. And if, if people aren't truly paying attention and focusing, you know, you tell them that hey, let's turn off the WI fi. And then you realize everybody falls off the meeting and this kind of, you know, recording here falls apart because I need my wi fi to be able to communicate with you. So it's, it's, I'm, I'm not truly facetious here, but I'm just want to put in perspective that tech doesn't replace thinking and general cognitive thought.

So let's step back and say is this good for humanity? Is this good for the business? Is this going to monetize? Is there a return on investment? How's that?

Matt

To me it makes sense. Pragmatic approach. And we had an interesting conversation during our last talk about the followers and leaders in tech industry. I hope you recall it and maybe you could elaborate a few words about it.

Martin Miller

So you know, it's funny is if you look from just the hype, you can see, you know, there's this double air quoted unicorn company and you see a lot of this because I'm out of Silicon Valley. And so you see a lot of this in my area. You hear about the billionaire and you don't, what you don't see is the path of how they really got there necessarily, but you see a lot of the hype. And you know, in the tech jungle there are leaders that think outside the box. There's leaders that turn the box into an app, you know, so you See, their idea becomes reality and they create a, an opportunity for monetization. And it's like, you know, some, some people call it a glitch in the software where another person sees that as a feature. And it's funny how accidents lead you into prosperity and the ones that actually know how to operate a video conferencing tool set is like keeping a sailing ship, you know, turning the right way.

I'm just trying to get you oriented that there's a lot of accidental success, but there's also planned and intended success success and leaders should follow a course, you know, using their digital fuel. Think of it like a mix of old world and new world. I'm going to make an analogy difference between a visionary and a follower. Think of the, if you go back in time to when the TV show in the US called Star Trek was released, you know, there, there was a spaceship, it was called the Enterprise. Funny, funny ironic thing that we call it. It was called the Enterprise. And the Enterprise to be successful needed both Captain Kirk and Scotty.

It's the teamwork between Captain Kirk and Scotty and then the greater team to keep the commanding of the bridge in order and the engine room working.

Matt

Let's get back a bit to AI machine learning. So for me the AI and machine learning gets a lot of R and D work with not very clear outputs, right. So you try, you break things and it's really hard to get something really tangible and working. And a lot of companies, organization now are crazy about the AI machine learning. But when I talk with the, with those guys, they really are afraid of investing in those projects and they are really, you know, really cautious about putting any money because they don't see a really clear roi. So I don't know, like what is your view on, on that maybe as an engineering leader, as somebody who is running the team and is responsible for the budget.

Martin Miller

Yeah, that's, that's a really honest and fruitful question to, to, to respond to. And there's the hype cycle. So I think a lot of people got caught up in the hype cycle and following others investing in the same part of the hype cycle. So in the, in the beginning we talked about intelligence or we talked about statistics and trends and following following these. And then we got into some ways of doing business intelligence and then leveraging beyond business intelligence, the data sets that become larger and larger and difficult to comprehend. We, we got into what's called data science and we're trying to predict the future from information from the past or do Things for decision making based on information. We have a couple of the challenge challenges in this whole hype cycle were, were how much to invest, what to invest, do we have the right people, the right tooling, and what's the difference between a data scientist and a software engineer?

They're quite different. And I bring this up because I attribute a lot of the lack of ROI as a person working in a lab like an alchemist, which is the data scientist, like a soft piece of clay, making an analogy like an artist. And if the clay never hardens, they can keep pushing on it, moving it and shaping it. Where let's take it over to the software engineer side or the engineering side in general, in that engineers deliver something that's, that becomes tangible, touched, used and operated on. And then you can track defects and repair and complete the cycle of improvement. So that bridging gap between the artisan to the delivery and proof of perfection and ROI is a huge gap. And many, many organizations were burned by this in the last five years.

If you, you know, you can watch the numbers people reported on their investment into the area, you can watch with their ROI that they claim to have made. And there's a misalignment. A couple of the problems were the tooling wasn't appropriate for the time, the investment wasn't appropriate for the time. You want to start small, Very, very small. Don't go big when you, those, everybody that went big has paid the price and you know, their, their investors know, know the truth here. And the truth is the return on investment wasn't there. So not to throw every company under the bus, but I'm going to go way out on the limb.

Those that have shown the best results started small and then organically grew that practice. And part of that practice is creating the data pipeline, making the data pipeline accessible, repairable, resilient, and working through obstacles in the data pipeline. Because without good data, nothing in the AI world is even worth anything. Because a model in AI is not just code and it's not just data. It's a synergy between the two.

Matt

I think it's a great lesson learned. So like start small. It's like, so it's so simple, right? Do you have some kind of budget that you just, just simply put there and you give it a try, right, and it could work or maybe not. But if you don't try it, you could be behind your competitors pretty easily. And Martin, could you describe me a time when you were a part of a controversial engineering or leadership decision? What did you do and what have you learned from that?

Martin Miller

Oh, I've been, I've been through several I would consider challenging or controversial. I mean it all, it all goes in various directions depending on where you want to pick up on the controversy. Controversy would be challenging. Is this a good idea? Does this monetize in the face of the pressure that it's a good thing to do and there's no validation of it. So I try to let data describe my point of view. I try to be objective and accepting and open.

You know, there's many, many data points I can share. I'm not sure I want to share some of these publicly, but I'll just leave it at this level.

Minds will prevail and stay level minded.

Matt

And you are in the heart of big tech. So you're based in California. So you have seen a lot of, a lot of stuff for many years and now you're struggling through the recession. But yesterday I read I think interesting report that the May was the, was the, was like had the lowest number of layoffs in one and a half year. So it's kind of like give a positive kind of signal. But for you as an insider, how do you, how do you see it? Have you noticed something?

Are you optimistic about the rest of the year and upcoming years or not?

Martin Miller

Yeah, let me, let me kind of replay a little bit of my, my perceptive perception and perspective on this in that I think what we had is, you know, you have to overlay the pandemic and what I call the impulse effect and ringing out or signal, you know, amplification because of the pandemic. And the pandemic caused a lot of closure, shut down and then pent up demand and then coming out of it a little over investment. And I don't want to say they're linked to the AI cycle because it had nothing to, you know, pandemic had nothing to do with that. But the two kind of, if you watch coming in it was on a growth, growth, growth, growth, growth, flatline coming out. Then it went boom, it shot up again. And so there's a lot in the tech sector that grew quite rapidly just post pandemic and then a little bit too much investment, a little overzealous investment, you know, trying to keep up with the Joneses and, and be the best on your block with the fastest cars and that, that eventually you know, has to curtail to where profits being made. What is, what are the trends, what are the other overlying influences that hit the macroeconomic equation?

And that's, you know, Price of oil, what's impacts, price of oil, price of energy, little things like that, you know, a little war here, a little war there and you know, political pressure, it all drives together. So in California in general, you know, it reflects a lot of the, you know, the pain points of being over invested in some areas, but resilient overall. Those that are able to adapt to change tend to do well. Those that don't adapt to change don't do well. It's really, that has little to do with tech on its own, but that's general life lesson.

Matt

But do you see like now like more optimism versus pessimistic view like last year maybe?

Martin Miller

Yeah. So if you watch there are trends, you know, let's, let's, let's jump on the, the, I would call it the elephant in the room problem, which is, you know, OpenAI opened up chat GPT for the world over a year ago, closer to a year and a half ago for general consumption. And what has happened since that is like another gold rush. So where there was a vacuum, there's now another gold rush. And if we step back and we think about, oh, we'll talk about that denim company, Levi, Levi Strauss. You know, during the California gold rush, Levi Strauss did quite well, even though most of the people going for the gold did not. The people that did well were making picks, shovels, clothing, selling food and services to those that were trying to find the gold.

It's very much analogous to what's going on today. Those that build the tooling, those that build the delivery services and have ROI associated with, they're all going to do great. I mean, maybe, you know, Nvidia will have a plateau. You know, that's the company behind a lot of the silicon that's used to help power some artificial intelligence solutions to run faster, more efficiently. And so maybe we'll see a plateauing there. But if you watch just them, they're, you know, they're on like a skyrocketing pace. So if that helps take this conversation in a direction, opportunity is there.

It's a question of how, how to make the best use of the opportunity.

Matt

And each year we learn something new as the leaders. This is what I feel. And there's like one change in the mindset. So I'm wondering In your case, 2024, your challenges, your pain points, what are you learning this year or what do you need to solve this year?

Martin Miller

You know, interesting. I spend a fair amount of my personal time, I'm a fractional cto, you know, a podcaster, and I also have my book AI in a Week and an executive's guide. So I have myself spread like, you know, butter on bread. And so I need to uniformly, you know, distribute myself. And I spend a lot of time with early phase startup founders, early startups in monetization path. To build value, ultimately you have to sell something. If you don't sell something, you have to ask yourself why you're building it.

I step back to a lot of fundamentals and then I get into, you know, the 2024. I think overall will be an okay to good year for many. For some it may not be as okay and not as good. And I'll attribute that to your ability to adapt, learn and change.

Matt

And the last question that I wanted to ask you, and I ask all of my guests, could you recommend any books, podcasts? Of course. Yours? It's one of those that have been influential for yourself particularly.

Martin Miller

Sure. There's several books actually. I list them in the references of my own book which is co authored John Sukup, AI in a Weekend, An Executive Sky. It's only offered as an ebook and the reason we're doing that is that we don't want to destroy plant matter in the process of creating a book. And most of our audience reads digital. So once again, it's available on every platform. Amazon, Barnes and Noble, itunes, everywhere you can get a digital book.

I think it's even translated for us. I, I don't have access to that information. But you're, you're welcome to offer special deal to your constituents to purchase a book and, and I could send them a digitally signed version if they with a proof of receipt. As far as podcasts, you know, I have my own podcast, it's called the Unrivited Podcast. So you can find us on YouTube and you can see see us in person on YouTube or you can listen to us also on every audio channel.

Matt

Awesome. Amazing. Thank you, Marlene, for today's talk. I really appreciate it, Matt.

Martin Miller

I had a great time. As I, I say all the time to people. Make the day great every day.

Matt

Follow Matt and Leszek on link.

‍