Product-Market Fit, Persevere vs. Pivot & Lean Build, Measure and Learn - E527

Jeremy Au examined the evolution of technology marketing and communication, emphasizing the transformative impact of AI, the importance of addressing consumers' primary problems, and the iterative nature of innovation in startups. Jeremy stressed the necessity of solving consumers' most pressing problems at the right moment to achieve product-market fit, illustrating this with real-world scenarios where timing dictated success. Additionally, he underscored the iterative principles of the Lean Startup and Kaizen, which prioritize rapid building, measuring, and learning cycles to meet evolving consumer needs while minimizing errors. The discussion also touched on notable examples like Netflix’s shift from DVDs to streaming and the development of reusable rockets by SpaceX to illustrate these principles in action.

(00:00) Jeremy Au: At the end of the day, startups are really about focusing and solving for people's number one problem and so the who, the when, the why is really key because you're solving someone's number two or number three problem.

People don't care, right? Because I got to solve my number one problem. Now, if you have been holding your piss for an hour and you really need to go to the toilet, that's your number one problem. I come up to you and say, Hey, do you want a SAS software that makes you learn faster? You'd be like, no, I need to go to the toilet now.

So people are very problem oriented. And so making sure that you know who the person is, when they are, why they have that problem is very key. So when my CEO was very basically yesterday was like, I have this problem. At this point of time, then I was like, okay, I will solve that problem for you. But in my head, I was like, if a startup was there at that point of time, I could hear him say that problem.

They could have made, a 10, 000 transaction immediately on that point. Now, if that startup talked to him today, they're not gonna get paid, but if they were there present with him, listening to him, maybe through Facebook or Instagram, whatever it is. then that problem would have been solved and the (01:00) money would have been made.

So the who, the when it happens, and why is really key. And we talked about how nobody wants to die, and when you're dying, you want to live forever, right? So when you're young longevity is something that's a bit different, but, we have Brian Johnson, and he's coming to Singapore in two weeks time, biohackers, it will send a video out, and, he has a good slogan called, Don't Die, Met a Guy, very interesting guy but nobody wants to die, right?

I think he has a very good marketing slogan, and he's got a group of people that he has hustling hard for it. So we talked last week about 10x better, we talked about Spotify Tesla, we talked about faster, which is in terms of convenience or perceived inconvenience, and we talked about cheaper, which is about price, right?

So again, for example, the reason why marketers are getting horribly out competed by AI right now is that, two years ago, marketers would have to be like, let's sit down and create a marketing slogan, which all of you did last week. But all of you, within, 20 minutes, were able to generate, use ChatGPT, generate that out.

So the speed of that was so much faster. Now, (02:00) was the quality of your ad really better than what humans could have done? Not really, I don't think so, right? I think we saw the quality was it was very authentic and handmade into a 20 minute time frame. But I'm just saying that a group of writers sitting down together, Would have taken 10 times longer to come up, because you guys took about 10, 50 minutes. So 10 times more would be 150 minutes would be about three hours, so you can imagine, I'm just trying to say here is like the slogans, the quips, the alliteration, the rhymes, what we had it would've taken more time. Soche, GPT as a marketer is 10 times faster than a human marketer.

And of course, one interesting part about this, also cheaper as a result, because when you're replacing a human. It's not only about time and speed, but also because man hours, historically, one hour of effort of my work was done. And a lot of us, historically, have left manufacturing jobs in Singapore.

We've given those manufacturing jobs to other countries in Southeast Asia, like Vietnam, Johor, Indonesia, because we said that robots have replaced Blue collar workers last time in the past when (03:00) you did an hour of machinery work, you got paid one hour, right? And then now with automation in those countries, because 10x faster, but you don't get paid more.

So we gave those jobs away. But guess what? You're now in services. You're now in finance. But the thing is this, if you're an accountant, you're in finance, you're a banker, DBS, you're doing compliance and paperwork checking, now, AI can take that job away. So the same forces that disrupted manufacturing and blue collar made blue collar an unattractive job, the jobs that your parents did and tell you now, please don't join blue collar.

Please don't be an engineer. That same disruption is now happening to the career professions that many of you are thinking are happening today. So it's really important for you to be thoughtful about these products, whether they're 10x better, 10x faster, 10x cheaper. And, we talked about this and I shared this, but not everyone may have a chance to see it.

But again, when we look at Two companies that are very similar, SpaceX and Starlink. Some of you may have heard it, it's Elon Musk. Here's the Steve Jobs of our current generation. But on the left hand side, what we see here is the rockets that have happened from 1960 to 2030, right? And if you look on the y (04:00) axis, basically there, is that you see here what is the price of setting up 1 kilogram.

So 1 kilogram of material, metal or plastic, whatever it is, and the cost of that used to cost around 50, 000, right? The rockets was about 25, 000. Space Shuttle was trying to be , 50.000 because of So many errors and they all blew up in the air as well. And then basically what you see this little cluster on the right is that, again, we talk about how the rockets became reusable, right?

And so these rockets effectively became cheaper. So the cost of it today has gone from 50, 000, 25, 000, 50, 000 to the Falcon 1, which was about 12, 000. And now SpaceX is about 800. And now ship estimate is about 200 per kilogram, right? So that is a 10x cheaper product. The satellite, the rockets are 10 times cheaper to operate, run, maintain, push up payloads.

And then he said to himself, You know what? I want to set up some satellites myself. And so they created Starlink. And Starlink, if you think about it, are just satellites in space that (05:00) provide cellular service. But the thing is, we all know now, is that they provide broadband level type of cell service anywhere in the world.

Bali, Indonesia, Ukraine, any, desert, wherever you are, you get it. And which is obviously, it's expensive, it's about the same price as a cellular telco plant that we have here. But it's 100 percent reliable because it's anywhere in the world you can get it. So it's a 10x better product because from your perspective as a hiker, from your perspective as a remote worker, why am I in Bali dealing with the local telco service that's very intermittent, that is subject to typhoons and so forth, when I can just get a styling dish and get Always on internet coverage, right?

And I think this is an interesting dimension where the same company has created two different products. They used one product, which is a 10x cheaper product, and then they were able to set up these satellites that were much cheaper, and as a result, much easier to make profitable, into the Starlink, which is 10x better coverage for that customer.

I think this is a good contrast of same founder, same (06:00) company, two different products, but two very different value propositions. And this is valuable because all of you will be doing group work today, you guys will be selecting startups, to review and analyze, and you'll be taking on the perspective of that founder to decide whether you are a 10X better product or in whatever form or fashion it is.

And so we talk about unfair advantages, first mover, fast follower, network effects, economy as a scale, IP and patents, regulatory and team, and lots of folks make very good questions. Can IP and patents slow down people? The answer is you can't duplicate some, you know, user patterns on an app anymore because of the patterns that's there.

There are many ways to do it. You can invest in them, you can buy them out, you can out compete them, you can just copy them. So there are many ways to do that. One example that we had is that the USB thumb drive, right?

So what most people don't know is that a USB thumb drive was invented in Singapore and this guy imagine he is about our parents generation and basically at that time Singapore was the blue collar Manufacturing zone for all these consumer electronics similar to how Vietnam (07:00) or China or Shenzhen will look at it today And so at that point of time, he was being asked to make these CD players right digital CD players for the Japanese.

And so they were like city. So because they were popular, right? These digital music players. And so his idea was like, you know what? I think it would be way more interesting if we've removed the products solutions. We don't need ability to play music. We don't need ability to play sound. I just want to have the ability.

I just want to keep the hard drive, but make it easier to stick into laptops, right? The time drives that we all saw. And so he created that and he found a patent for it. And then nobody respected it. So his ex employees two of them were Chinese. They were in China they went back to China, and then they just built another company on top of it.

And then his product was granted patents in the U. S. and in Singapore. And then the Chinese company got the patents approved for China. And that company became a billion dollar company. And now it's not doing so well, but they (08:00) make a lot of money off the patents, the licensing fees. of all the Shenzhen manufacturers, etc.

This guy received some money from the manufacturing licensors, especially from the Western oriented or Japanese oriented manufacturers. But because his patent was not respected, in China as a result, he did not capture most of the outcomes. So the track TREK is that first, Singapore has pioneered the USB thumb drive.

And currently, he's in jail. So his company wasn't doing well as his licensing royalties went down. He unfortunately chose to put together fraudulent accounting statements to hide that drop in revenue over time. And now he's in jail. Interesting story, founder, entrepreneur, pioneer David trying to become a Goliath, but go.

outcompeted because the patterns were not respected and now he's in jail. So but who knows? When he comes outta jail, he might come up with another idea. He actually has come up with multiple ideas over time. He has created a wireless time drive. which is, used in many cameras today for your wireless syncing.

He's a kind of a founder and pioneer and inventor (09:00) persona. So this is something for you to be thoughtful about that, the unfair advantage, a lot of it has to do with this competition. And so we talk about how network effects are really important which is that the more people who use it. The better it is, right?

So we gave the example of memes. The more people who consume the meme, then the more fun the meme is, right? So the network effects of that are there. That's why you say memes are viral, right? We also talked about other stuff that are also viral, but we also talked about how network effects, as a result, help you out compete your competition over time.

And as a result, When you think about network effects, everybody's going to walk around saying, I have network effects, I have network effects, everyone says I have network effects. Does it make sense? But, you have to be very thoughtful about the type of those network effects. And this realization that Uber has Sub city level network effects and even within cities, sub neighborhood effects was not understood until maybe about three years ago in terms of knowledge.

And even today it's still not commonly understood by most founders today. It's not common knowledge. Today all of us, you (10:00) and I sit down together and we say, startups, unicorn, that's common knowledge because that knowledge started coming into existence about 10 years ago, right? So this now is a lexicon where we all know what it means.

But network effects is still in that. middle stage where most people don't really understand it. You know what a word means, but they don't really know how to apply it. And so I'm just saying like there's a lot of money left on the table for you that if you are able to think through those network effects carefully from a first principles basis, there's money for you as a financier, as a founder, as a thinker.

(10:29) Jeremy Au: we have a certain philosophy called Kaizen. So maybe you have heard it. It's called in Chinese, the Chinese characters would be Kaishan. But basically what it is, is that it is the process of which manufacturing became more lean. And so historically, when people make cars and those devices. They were looking at it and doing it in terms of every corner of the workshop, everybody worked on that piece.

But of course, over time, they created the assembly line, which is what Henry Ford did, where people would do (11:00) one thing at a time. So, when a car goes by, I'll add the doorknobs. As the car goes by, I add the wheel. So there's a manufacturing line, and the Japanese improved on it, where they're saying that at any point of time, we can iterate and improve at any point of the manufacturing step.

We will try to improve that step. We don't pretend that it's perfect. We will improve on it. So if you notice that there's an error that's coming down the line, you can pull on a line that rings a bell, and the whole assembly line will stop. So imagine thousands of cars are on a line, tens of thousands of people on a line, you pull the thing, everybody stops.

Why? Because what the Japanese realized was that if the junior worker sees the error, but lets it go, then what happens is that that error happens for tens of thousands of cars, and it's gonna be way more expensive recalling those cars to come back. And so it's better to have a junior person pull that line and raise the alarm and stop work and fix the problem immediately, a small thing or whatever it is, and then go from there.

And we saw that practice at Boeing, where we now have planes that have structural maintenance faults, (12:00) right? So people knew about these issues. They knew that they didn't pass quality control. They knew that they were sloppy in sealing the thing. And then guess what? You're hanging out on your phone, chit chatting, and suddenly the exit door blows open that It wasn't supposed to be an exit door.

It used to be a space for exit door. They put a seal over it. There was no quality control. And suddenly, you're hanging on for dear life while somebody's holding you. And then you go viral because you don't have a shirt. You're shirtless. You're a teenager, university student, who is there, right? And so, Boeing now is in deep trouble because they have multiple of those issues.

Because if you have that issue, if you think about it, Then there are other types of issues that happen. You think that was the only fault that happened on the whole line? So if there's one person that was the first area that got caught, you know, years down the road, Now obviously there are a lot of other faults that have happened all along down the line, right?

And so there's a culture of saying that we should be able to transition and change and transform as much as we can. So what we want to do as a (13:00) result is that, you know, there came a philosophy called the Lean Startup, which is that the Lean Startup is really fundamentally about doing three things.

The first thing is really about building. The second is about measuring. And then the third is learning. So what that means is that when you're doing a learning loop, when you're doing a startup, you should be able to build something as quickly as possible. Then after that, you understand what are the outcomes that were supposed to come from it, you know, because you released it in the wild.

And then you learn from it and decide to do something different, and then you build the next version of it. And so this is the same philosophy of Kaizen, but in a culture of startups, in Kaizen, it was, I need to get a car down this line, and I want it to work, right? So the output of that is perfect cars in that sense.

Less waste, less mistakes. But for the learning loop, for minimum viable product, very much about saying, the output of what we're trying to achieve is the learning of the startup. Some of you (14:00) picked a company for the first time, then you got some feedback and then you learn from it and then you may change your mind.

So the output of startups is not building something, although there's a big part of it. The key value proposition of a startup is that you must be learning faster than everybody else at the frontier of customer requirements and technology. She was asking, how can if Goliath has so many advantages, how can Davids be better?

And I always say, it's not that Davids are releasing products better, but they are learning faster than the organization can do. So the speed of that is really important. And so the Lean Startup is really about saying, how do I incorporate and build better ideas? And so as a result these companies are trying to do minimum viable products.

And what that means is that historically most people think about minimum viable products as saying, okay, let's build the wheels, then we build the chassis, then we build the car. So the minimum viable product is they need to build a whole car. But what we mean by minimum viable product is that what is the minimum requirement to travel?

The first step can be building a skateboard, the second one can be building a (15:00) scooter, step three could be building a side bicycle, step four is building a motorbike, and step five is building a full car. Right? So the minimum viable product is, what is the minimum level product that gets you from point A to point B?

So if this person says, okay, I just want to learn and get an A for my math exam. Step one could be you doing it manually as a human. That would be the fastest to do it because you really can teach a person how to do all level math. Step two could be you can start using some AI tools, paper worksheets that you believe.

Then step three, start incorporating AI tools. Step four is an AI front end, but the humans at the back. And then step five is a full AI agent, right? But at each stage of this product, there is an actual ability to deliver the result of what the customer wants. And so as a result, when we think about startups, we think about it in terms of divergence and convergence.

. It's very much about diamond shape. It's like some of you would notice that you're in different conversations. And at the start, you have the same point, right?

The point that you had was figure out which is the best company out of the 40. So all of you diverge. All of you start (16:00) looking at the different websites and parallel. So all of you are diverging because you're creating in a group of five, five times 40 startups. So you are creating 200 different opinions in that space.

So you are diverging. And then somebody starts saying, we should start filtering. Cut off all the startups that only have male founders. Let's divide this in half, right? Or some of you may say, let's only focus on the AI stuff, right? So there'll be ways for you to converge. You're trying to filter that down.

And so it's really important that when you think about the companies that you're looking at in any product or whatever it is, there'll be a stage for divergence. You're trying to explore as many ideas and some of you will be converging, trying to filter it. And it's important that when people are diverging, you don't try to squeeze people to converge.

And when people are trying to converge, you have to be comfortable with the convergence. And then you have to kind of do this multiple times. Diverging and converging is important, and we see that at startup time.

And problem with a lot of companies when you join a big company, or people complaining about bureaucracy at a very large organization like a public service, is because Every time you are creative, then somebody's like, converge, you know, (17:00) I want to cut it off, it's a bad idea, you know what I mean? So you thinking about that divergence and convergence spectrum all the time.

And so as a result, if you think about it, is that every company, and even for you guys, are thinking about persevering or pivoting. So you may have heard that phrase. So the pivot is to change your mind, okay? And startup founders change their mind all the time. They hear information, they change their mind.

They hear this new piece of technology, they change their mind. They hear a new tool, an AI tool, they use it. So people are pivoting all the time. But also you need to discuss and decide when you need to persevere, what things of it do you want to continue doing. So some of you may choose uncommon, and you may say, you know what, everyone else is wrong.

They don't see the value of what we're doing. We're going to persevere on this company selection. So be it. Some of you pivot, and that's good. You say, hey, I'm uncommon, I'm going to pivot to another company. that I want to work on, not a problem either. The decision, which, where you make a decision between perseverance and pivoting is important.

(17:52) Jeremy Au: And so there are three different types in terms of pivoting and a good way to help understand the customer and market. So there are often companies that have (18:00) chosen to keep the customer but change the product. So what that means is that they like that customer, they understand the customer. Maybe they are that customer, but then they change the product because they realize that the product should be a better version of that.

So for example, for Instagram, it used to be called Bourbon, which is a type of four square. So you could check into locations, take photos of it and then leave reviews. And then what they realized was that people really like the photo sharing piece of it because everybody was a millennial at that point of time.

And so they killed all the other features. They were like, you know what, let's focus entirely on the photo aspect. We don't care about check ins and being a mayor and stuff like that. And today, nobody knows what Foursquare is. Foursquare is dead. It was very popular, but Instagram is still around. So that's one.

If you look at Netflix, Netflix was saying, We want people to watch TV. People want to watch Hollywood. They started out with DVDs a long time ago. And then at some point, the founder was like, You know what? DVDs are old school. Like, we're going to mail the DVDs to them. Right? So, and so, everybody's like, that's crazy.

(19:00) And then he had to create a whole new processing warehouse to mail DVDs, etc. And after that, they were like, you know what? I'm gonna change it again. People still like to watch TV. They want it more convenient. I'm gonna change the product. I'm gonna make it wireless over the internet. Right? And now all of us all watch Netflix shows.

So, the product keeps changing, but they're going after the same customer, and the same, roughly, problem that they have. And lastly, obviously, we look at Rippling. We look at the company here. But basically this is a HR company and basically they're targeting HR professionals that want to automate their life.

So Zenefits was a company that was focusing on providing and automating the benefits package. selection for HR professionals. And today two of those alumni are really famous. Conrad Parker was effectively fired and let go by the board. He will later on found Rippling, which also targeted HR offices again.

It was his revenge startup. He was backed by Y Company, Paul Graham, because everybody hated on him in the media, but then YC and Paul Graham believed in him. And it felt like those accusations were unfounded. And therefore, he built a second company. And that company (20:00) became a billion dollar company called Rippling.

And Zenefits is no more. It's gone. Zenefits was taken over by his COO. Uh, And today, if you watch the All In podcast, I'm blanking right now on his name. It is not Chamath. It is not Jason. It is david Sachs. Who is very famous today, was a COO. He was a part of the team that fired Conrad Parker.

And now he, that COO of Zedafits, who became the CEO for the interim, is now part of the All In podcast, which is the number one tech podcast in the world, right? And so he's a big Twitter personality and so forth. So, but again, you know, you can see how these companies are keeping the consumer, but changing the product.

And so the next version that we have is you can keep the product, but you can change the customer. So for YouTube. It used to be a video based dating site. So people upload videos of themselves, and then people would just like it, and whatever it was. It was for dating purposes. And then basically they said, you know what, we like the product, but we shouldn't (21:00) be targeting people who like to date.

We should be targeting people who just want to watch cat videos, right? So they pivoted, and they kept the product, but they changed the customer. And then if you look at Play Doh, that Play Doh used to be a cleaning party to clean the walls. And then more people disappeared, you know, pretty much effective this industry.

So they're like, you know what? Let's sell this to kids instead, right? And their parents. So, Play Doh is another thing. And if you look at PayPal, it used to be about sending money over PDAs. So sending money from device to device in terms of encryption. But then they decided, you know what? We just need to change our customers instead of servicing this for enterprises, we're doing this for individuals as well.

What challenges do they have when pivoting? When they're pivoting, everybody hates them. your old customers hate you for pivoting, the ones who believed in you. Your new customers think that you're a loser for changing your mind.

And then I think pivoting everybody, you know, is People can be, a lot of founders are egoistic, so they're like, I made a choice, I need to stay on it.

So it's very hard to pivot, it's very hard to realize (22:00) to yourself it's very hard to explain to investors who have invested in you on your previous idea. So, pivoting is very, very hard, but I would say that the actual execution of pivoting is actually easier. So, like, building a new product, if you're a founder, you know how to build product, you know how to hire people.

Like, the execution of the pivot is actually pretty easy, generally. Like, for example, if you're Netflix, Back in the day, you know, he was fighting against Blockbuster, right? So they were both selling DVDs physically back in the day. And then both of them and Netflix eventually became sending it by mail, eventually sending it online.

Blockbuster kept doing its stores and it totally died as a company, right? Called Blockbuster. It's not hard. I mean, if you tell people, like, hiring engineers, they have a lot of money, they both are profitable. It was actually easy to execute. So I think the primary difficulty of pivot, pivoting is changing your mind.

And that's difficult because most people think that pivoting means that you're saying that you're wrong, but it's not. Things change all the time. Like, you know a lot of companies were building therapy companies three years ago, four years ago. Like Human to Human Fractional, like BetterHelp, etc.

Or they're trying to use (23:00) non profits to provide therapy services to youth and people who can't afford to pay. But now, GPT has come out, right? And so you see a lot of therapy startups. Some of them were very quick. Within 30 days, they were like, we're going to use AI from now on, and we're not going to use volunteers or whatever it is.

But of course, the problem is a lot of people saying like, oh, can you use AI therapist? Is that safe or not? Blah, blah, blah. So obviously, you're getting a lot of hate. But then, you see for the companies that are slower to react, a lot of them have closed down because it turns out that you can do it effectively for free and immediately, right?

So I think pivoting, I'm just trying to say here is like. It's okay to change your mind and I think that's the hardest part. It doesn't mean that you're wrong, but, you have to figure out what's the more right answer. If you want a job that tells you that you're right all the time startups and technology is probably not.

Does it make sense? It's not a place of continuous positive affirmation. It's about trying to search for the better answer and truth to it.