How to have a successful AI startups

This week, I had the privilege to receive Greg Coquillo, a distinguished voice in the startup investment space as well as an experienced product manager in the AI space, we dive deeper into the nuances of market signals, startup phases, and the current investment landscape. Greg offers his invaluable expertise, offering insights for entrepreneurs, advisors, and investors in the rapidly evolving world of startups, particularly in AI. We also end up talking about his 2-times LinkedIn top voice awards and discuss how he achieved over 190'000 followers on the platform.

Greg explained the various phases a startup typically undergoes: prototyping, product-market fit, growth, and protector mode. Each phase presents its unique challenges and opportunities, making it crucial for startups to adapt and evolve. For instance, achieving product-market fit is a critical milestone where a startup's product becomes essential to its customers' operations. This phase is followed by growth, where the startup must expand its customer base and fend off competition. Finally, in the protector mode, the focus shifts to retaining major clients by continuously innovating and enhancing the product.

Greg then mentioned the need to understand market signals, which I was extremely keen to dive into. Market signals are indicators that help startups decide whether to pivot or persevere with their current strategy. Something very difficult to know, especially when you need to "kill your baby". Greg highlighted that understanding customer needs and market response is key. For example, if a startup is spending excessively to acquire new customers without seeing a corresponding return on investment, it's a clear signal to reassess and potentially pivot. I obviously kept the best tips for the episode itself ;).

Greg also shared common pitfalls for startups, such as distractions, scaling too fast, and misalignments within the founding team. He emphasized the importance of a clear understanding and articulation of a startup's value proposition, cautioning against the trend of hastily labeling a startup as a 'Gen AI' company without substantiating its core technology or problem-solving approach.

Addressing the current market conditions, Greg expressed that the startup world is always full of opportunities, albeit unpredictable. He finds excitement in startups that can articulate their value even without AI components, indicating a robust foundational problem-solving approach. Greg's perspective is that the market is dynamic, with potential in various sectors, and the key is to identify startups with a clear, unique value proposition.

This podcast episode with Greg Coquillo is packed with insights for anyone interested in the startup ecosystem, especially in the AI domain. Greg's expertise offers a roadmap for navigating the complex phases of startup development and investment. Listen to the full episode on Spotify, Apple Podcasts or YouTube:

Full Transcript:

Greg Coquillo: [00:00:00] How well do you understand the problem space? How well do you know your customer? How do you know it's needed now? How big is the market opportunity? What is the competitive landscape? Why do you think you can solve it better? Do we have the right people put together? So those are the things that you kind of look for as quick signals.

To see, okay, do you have the right team working on the right problems at the right time? Right? So therefore you were able to minimize the risk you're taking with a startup. When you decide to say, yes, I'm going to put my money on it.

Louis-François Bouchard: This is an episode with a great friend of mine, Greg Coquillo. Greg is an experienced product manager in the AI space, but also an investor in the AI startup world. So you can see this episode as your go to for learning about the startup world and also the investment world. Greg offers amazing insights on where he invests in companies or not, and what he looks [00:01:00] for, for a company.

We also discussed the pitfalls and the market signals. To look for Greg is also super active on LinkedIn, where he has built a great following in the eye space of nearly 200,000 people. If you enjoyed this episode, please consider leaving a like in a five star review, wherever you are listening. I hope you enjoy it.

Greg Coquillo: Well, Hey guys. Uh, thanks Louis for, uh, having me to your podcast. I know we've been talking about that for a long time and now I'm here and I appreciate your patience and getting me into your. Awesome platform. I'm Greg Coquillo. I work for one of the FAANGs out there, and I am passionate about technology.

My background is in industrial engineering, and I have been tinkering with data for quite a bit. So in industrial engineering settings, you can think about things such as statistical process controls and other [00:02:00] supply chain management. Forecasting, et cetera. Those are the things that you're probably familiar with if you're in the data space.

Back then, we didn't call these forecasting models, machine learning, et cetera, but that's really what we're doing. I remember before ML, AI was a thing. I was working with third party vendors of specialized camera systems that would take a look at our production lines and send signals to our machines to make adjustments to reduce quality defects, for example.

And today we're thinking about computer vision, more sophisticated. architectures that allows us to scale the reduction of quality defects. So this journey has taken me to tech, to the tech world. And also at the mean, in the meantime, I've been really vocal on social platforms. Most. [00:03:00] Namely, uh, LinkedIn, where I've been pushing out content around technology, AI, machine learning, et cetera, where I was able to gather a nice group of communities who felt, uh, connected with the messages.

And I've been very. Honored to be named, uh, top voice twice by LinkedIn. And, you know, I still continue this journey today and trying to make sense of what's going on in the world at the macro and micro level, uh, especially today, now that the world is trying to figure out what's going on with this thing called Gen AI

so it's been quite the journey. And things are changing so fast that people are really struggling with knowing how to ingest information. But so here I am, Greg Coquillo, who vows to simplify the number of [00:04:00] the amount of information coming in from all avenues to make it simple for folks to consume when I issue a post, for example.

And I also invest on my spare time. I mentor startups. Uh, the founders, um, I advise when I can and I connect them with people who can bring value to them as well. And I work with companies like Madrona, uh, especially Madrona Ventures Labs to mentor or moderate sessions with startup founders and brainstorm ideas and the hard to solve problems to figure out what are the different technologies? What are the different go to markets in other strategies we need to bring to the table to improve people's lives, et cetera. And, uh, I really do enjoy that. And, uh, and I enjoy talking to you [00:05:00] too, Louis. So thanks again.

I know this was a long intro intro, but I wanted to make sure people knew who I am. And by the way, I'm from Haiti. I speak French and Creole, just like Louis. We both speak French and, uh, great to be here. 

Louis-François Bouchard: Yeah. Thanks a lot. And that, that was a great introduction. Very complete. I will definitely talk about startups and I want to first start with, maybe I can start with how did you start to invest in startups?

If you can share what's the first one or just like approximately. When did you get into that and what made you take the leap to give, finally give money or advice with your time? 

Greg Coquillo: Well, first of all, when I was in college, I started, I had a startup and going into the activities of asking people for money is not, it's not new to [00:06:00] me.

This time I'm the one giving money and. It was quite an experience because most people, I was going out there, uh, door to door to ask money to, they wanted to see how much skin in the game I had. So as a founder, uh, you have to show, you have to give the investor confidence about how much time you're willing to spend on this idea that you feel strongly about, uh, to convince them.

So I've been, I've been familiar with that, right? So I was on the receiving end at some point. Although not successful, but I've had some good learnings. On this journey. So I've always wanted to be a startup investor. It was only a matter of time until I felt comfortable with one, my lifestyle to the space I wanted to invest in and [00:07:00] three, the quality of network people network that that would give me if I were to go into it.

So I have to explore all of that to say when is the right time to start acting. To tell you the truth, I can't even remember when I was, it's only recently I started really doing like giving for formally giving startups, uh, seeds, pre seed money, uh, for investments prior to that, it was more of a friend to friend investment in a handshake in hoping that something works, uh, but now it's a little bit more formal.

I take a deeper look into. The startup, but long story short here is curiosity. Again, you know, got me there and you, you walk into an environment and you stay in it long enough. You'll start finding folks [00:08:00] who are, who believe in certain aspects of life. And some of them will resonate and some of them will not.

So when I found folks around me, when I moved to Seattle, who are very passionate about investing, I started to lean in and listening. In learning, and I started to take the leap as well. But of course, it was a measured leap because at the account for the lifestyle, taking care of family, et cetera. What does that do?

So how do you really budget for that, et cetera? And then what do you select? And for that, you have to make calculated risk, understanding what, what works and what doesn't work or why understanding why sort of fail, understanding what makes sort of successful and, or trying to find a framework for sizing the, I guess, the, the risk that you're taking.

Right. And what people don't understand is that when they work for. [00:09:00] A company, or when they go out to go drive to go to work, or when they invest in a company, they're taking risk and these risks, they may not be comparing to each other, but they also have a value, right? Some risks are higher than others.

Yeah. And some risks may be high and they may return a high value. They may be some risks are low. They return a low value. Some risks are low and they return a high value. I mean, you have to understand these. How these play into each other to kind of make a decision, what can paralyze you? And I didn't allow that to paralyze me is not seeing the world as a, with a binary lens.

And in life, you have to find a way to balance everything. There's pros and cons in anything you look into. And at some point [00:10:00] I had to, I felt like it was the right time to pull the trigger. And because I was already in the AI world, I was like, let me look into this AI startup world to see if I can find the right folks.

To guide me in terms of how to venture into that. So that's pretty much how I got done to it. 

Louis-François Bouchard: And why choosing to invest in startups in the AI space rather than going for like a publicly traded company like Google or like some, a big one in the AI space, why investing in something, I guess, much more risky, but with high potential, like, I guess most people like me right now that don't have unlimited money. 

Invest in like anything public that is somewhat safe. So why, if you wanted to go for something AI related, why not, why choose a smaller and more risky company than a bigger one? 

Greg Coquillo: Yeah, [00:11:00] it's a question of risk reward tolerance, not just risk tolerance, but risk reward tolerance, right? As I mentioned before, some big risk may return big reward, and I'm comfortable with that.

Right. Some small risk may return small rewards that I find not being worth it, worth my time. Right. So how do you make a balance between all of these scenarios and, you know, you have to bring in multiple factors in, for example, age, uh, income, market conditions. You know, there are so many things that you have to put in place.

So it's never a one size fits all. It's really about who you are, what you believe in, what's your current condition, what's your current lifestyle, what are you willing to sacrifice. And at the end of the day, what does it give you? Are you doing it for just the money or [00:12:00] doing it for the journey, for the excitement that one day you can look back and say that you've contributed to something big, something that improved, improved people's lives.

It's really how you want to take a look at it. And that determines where you sit, whether you're a risk averse and you want to go the traditional route of investing where you're optimizing your 401k or you're allocating some of your income to some sort of investment account via Ameritrade, something like that, right?

That's known. And by the way, I do that too, right? I go the route of. Investing in a stock market, investing in ETFs, uh, but also with my age, with my passion, with what I believe in, I can size, somewhat size the risk that I take with venturing into AI startups. And that's [00:13:00] pretty much how I take a look at it.

It's never a one or zero type lens, binary lens. It's more of a dial up lens. So basically in a year, you may see that there's so much in, we are living it now. There's so much noise into the AI space. Now, how do I dial it down a bit, which is dial down the bucket, the AI bucket investment and dial up something that's a little bit more stable, whether it's real estate, I'm just taking a few examples now, or traditional enterprise that are already public, et cetera.

And how do I dial it back up when things are more stable and that's going to. What's going to determine that is my risk tolerance. 

Louis-François Bouchard: Another thing I guess that affects you is that you, you have an entrepreneurial mind, like you, you don't only invest your money as we discussed before this podcast, you also [00:14:00] advise or like mentor, you invest your, your time in the companies that you, you want to partner with, to collaborate with.

So that's also something to consider for some people like me right now, uh, time is extremely precious and I, I try to build things so I don't have like much time. And that's why I invest mostly in ETFs right now, just because I don't want to, to manage more things than what I currently, um, I'm doing.

Since you, you invest both your money, your time. So I guess it's a lot of investment. Like if you, if you go with a company or not, how do you size the risk you, you will be taking with them? Or do you have any like lists that you're going through or 

Greg Coquillo: like a framework? 

Louis-François Bouchard: Yeah. Framework or just like you, do you more feel just a person, the entrepreneur, the team, or do you have a framework?

How, how does it work for you? 

Greg Coquillo: Yeah. By the way, it's really difficult to size the risk. You really never [00:15:00] know. I mean, especially the anecdotes that. Are plaguing startups today. I mean, you hear about horror stories. Of our startup fail and you know, 90 percent of them fail, like when you hear stories like that, when, then you say, well, how do I know that I'm looking at a 90 percent startup or a 10 percent startup, right?

A 10 percent startup is kind of like the, the ones that are more viable for success. And I say viable. And I choose that word very carefully because still inside a 10 percent they may be succeeding today. And then 5 years later, who knows they go down, right? So. So the way I look at it is one understanding some of the reasons why startups fail, right?

I was reading an article the other day from CBN sites and I saw, oh wow, this is pretty cut and dry. Some of the five reasons that I can remember thinking about this [00:16:00] now. One is running out of money, two, they may be pushed out by a competitor. Three could be that what they're creating, the market doesn't really want it now.

There's another one, the business model may not be a great fit. And then there's another one that I can remember too is pricing, there may be a flaw in the pricing model that they select. And there are, there are more in that article. I just don't remember what they are. When I saw these reasons, I could only bring it to one thing that's in common, which is you're spending too much money because you're not focused.

On solving that one problem that you vowed to solve a competitor kicks you out. That's because you're not focused on finding the right customer for your problem, [00:17:00] or you're not solving it better than the competitor, right? You're going down as a, as a startup, because. The demand doesn't want it.

The market doesn't want it. Well, that's because you've been building something people don't want. So you didn't spend enough time with the customer to understand their problem. So these top problems, why startups fail lead me to say. Okay. If I'm talking to a founder and listening to their ideas, I want to understand how well do you understand the problem space?

I want to know about that. How well do you understand the problem space? How well do you know your customer, right? Those are the framework that I run it through. How do you know it's needed now? Why should we act on it now? How big is the market opportunity? What are the, what is the competitive landscape?

And why do you think you can solve it better and to answer that question, [00:18:00] that last question. We go into the people aspect of a startup. So looking into the team, do we have the right team members that can go after this problem space from a technology perspective or technical aspect or from a domain experience?

Do we have the right people put together to solve this? The other aspect that are more on the emotional, uh, intelligence space, how much experience have that person had with interacting with empathizing with, with customers? Are they able to motivate a room when they become leaders? Are they able to inspire investors when they get in the room?

Are they able to explain? Things in the simplest way so that investors or customers won't have to [00:19:00] second guess or guess what they're trying to express, you know, so those are the things that you kind of. Look for as quick signals to see, okay, do you have the right team working on the right problems at the right time, right?

So therefore you were able to minimize the risk you're taking with a startup when you decide to say yes I'm going to put my money on it. And then other aspects to is you does this company have traction, right? What are the social proof? Proofs that this, this startup have, right? So social proof can come in multiple ways.

It could be that customers are expressing interest, whether it's a letter of intent that they've signed, or they've already provided data to this startup so they can build a prototype. They made a commitment to that startup, or they're already paying customers. Those are social proof that are very powerful.

That can help an investor minimize their risk. [00:20:00] And you also have the other aspect of social proof where a startup founder can show that, Hey, I spoke to Bill Gates. Bill Gates said, this is a good idea. And here's my proof that I've talked to Bill Gates and take a look at his article where he mentioned my startup.

This is like one of the biggest social proof. That exists myself as an investor, we'll use that as a data point for sizing the risk that I'm taking. So just to say there's no really one size fit all in terms of a framework to size up a risk. And determine go versus no go and investing into a startup, but you do it enough.

You start to tailor multiple frameworks based on the space you're trying to evolve in, right? I mean, you have a different framework for a startup that's in [00:21:00] healthcare versus a startup that's evolving in a less rigorous. Industry like healthcare, like something like, I don't know, gaming, for example, right where data privacy and healthcare is a little bit more rigorous.

Versus data privacy in gaming, right? I'm just taking an example, right? So my framework has to work and evolve with that. Versus a framework for investing in real estate. I have to pull another one. I can't just use one framework for every single situation or use case. So again, life is not binary. We should see it more as a dialed, dialed up experience, dialed up, dialed down experience.

Louis-François Bouchard: You mentioned that you, once you were comfortable enough with artificial intelligence, you started, this is where you started investing. So how much do you need to know about either an industry or like about the [00:22:00] Startup is it's not the startup itself, but the problem they are trying to fix. How much do you need to know about the, yeah, both the industries and their problem to consider that you know enough are comfortable enough to invest.

I guess that's pretty much subjective, but in your case, do you need to be an expert and completely understand their solution, their problem, their, their industry and their, their like ideal. Uh, audience or clients, or do you trust the team and if you understand it somewhat well, and the team seems really experts, you then like not blindly, but you, you trust them, what's your level of expertise that you, you want to have with regards to the company's field?

Greg Coquillo: Yeah. I mean, again, I would say it depends on different conditions. [00:23:00] And how are these conditions coming together? I would say if you're starting to invest in startups, you have, it's probably best you start, you stick to what you know, stick to what you know, in terms of domain expertise. And you start playing in that space, so you can kind of minimize the negative effects of putting your eggs in the same basket.

Right? So you're, you're able to filter out the bad apples in the space that you understand better than in the space that you don't understand. Now, as you grow confidence and you feel like you want to spread your eggs elsewhere. Now this is where who you know and who you trust also matters. And who you trust and that this is not only the, the founders with whom [00:24:00] I believe every investor should, especially if they're, they're involved at the, at an early stage, they should form some sort of relationship with their founders.

It's also who you know, that's interested in the same space you're interested in. So, for example, I don't consider myself an expert in AI, but I surround myself with people whom I call experts and who've been investing longer than I have. So, and whom I trust. So, when they're leaning in to take a chance, I follow their lead sometimes when I need to, even though the bet I take may be smaller, but that's okay.

My point of entry into a world that I'm learning over time. So whom, you know, is really important. And then another thing too, is with regards to the conditions. Market conditions, for example, am I [00:25:00] in a situation where the market conditions allows me to liberate some dry powder, which is a term that, you know, investors say about how much money you have available, how much cash you have available to allocate to investing.

We've seen the past two years, a lot of companies or a lot of investors ran out of dry powder to, to allocate a lot of their money were allocated to. Illiquid assets. So they couldn't liberate, you know, cash to invest. When the market was favorable, right? So if an opportunity comes to me today, although I'm not a fully an expert into it, but I found that my friends and my colleagues, my mentors are all vested into this, I could be unlucky that I don't have enough cash to invest into it.

Right. But if I'm lucky, yeah, I'll lean in a little bit more and take a little bit of risk to see what [00:26:00] happens. Right. And over time, I'll get more confident because I also have to do my homework. I can't just sign a check and not invest my time into it. And one of the vows that I make is one of the commitments that I make is that if I'm putting some money out for a startup, I have to continuously.

Learn about the industry, learn about the space, problem space they're evolving in, uh, from time to time have touch points with these founders to understand how things are going and become someone of value to them as well, right? Knowing that they are a limited team, they may not have. 360 degree view of how things are going.

But when I have information that may be useful to them, I have a trigger for surfacing these things to them, so creating value for them too. And that helps me grow. That helps me grow my confidence for next time. I can lean a little bit less [00:27:00] on to what other people know. I can lean more on to what I know to make potentially bigger bets.

If the opportunity shows up, right? So hopefully that answers your, your question. Yeah,

Louis-François Bouchard: I just want to go back on something you mentioned a bit earlier, but there are, of course, multiple factors that you consider, such as the founding team, the problem and the timing. And so I wonder, what's your opinion on like pivoting and changing problems or wait, well, changing solutions, but even changing problems and more specifically, have you ever trusted a team or a, an entrepreneur enough to invest in their ID, even if it changes or evolves?

Like, have you ever trusted someone enough that you wanted to invest in this person, whatever he builds? 

Greg Coquillo: Yeah, there's a fine balance between really believing in an idea and not letting [00:28:00] go versus knowing when to let go versus knowing when to pivot, when all the market signals come at you. That's why, you know, it's always good to maintain a relationship with the founder, because as a partner, as a friend or partner, you want to be able to size up, you know, the level of commitment, which will change over time.

Right. What I mean by the commitment changes over time from a founder is it's good to have a founder that's. Receptive to feedback, right? Feedback from the market, feedback from customers, companies like Slack exists today because they were able to pivot from an initial idea that lending them into a space that's much more successful, right?

If they didn't let go of the initial idea, you wouldn't be [00:29:00] hearing about Slack today, right? That's an example, but at the same time, you know, it doesn't really. Minimize the founder who has strong conviction about an idea that's rather commendable. However, as an investor, as a friend, as a mentor, as a partner, it's the challenges.

The challenge is how do I help this founder realize that maybe the time is not now for this idea. Here's another perspective. That you should consider to stay afloat until the right time, and then you can act on your vision. Right? So if initially I really believed in an idea and I realized the market conditions have changed in.

My partner, the co founder needs [00:30:00] to evolve with that. I'll try my best to kind of like convince here, here's a better strategy so we can stay afloat during this period. And by the way, this period may be two to three years or four years. Here's how you stay afloat while you fine tune this big vision. And here's how we will kind of deploy that mission, that vision that you've had from 10 years ago.

Right. So when there's a will, there's a way. Right. And the will has to start with the will to be receptive to, to that market feedback because the market will talk to you. The market will let you know. That they're ready to, to use your product to solve their problems. They'll tell you when to be ready.

And if you, if you're one who's not willing to listen, while it might be a great idea, it may not necessarily a great idea for now. And that could put you underwater. And then next thing, you know, 10 years later, somebody else. Who came back to your original idea at the right time is more [00:31:00] successful. And unfortunately, you know, some people take it very hard.

Louis-François Bouchard: And you mentioned market signals to like letting you know to pivot or not. Could you share a bit more about like what are those signals that, that one should be looking for if they should be pivoting or keep pushing? In their current ID. 

Greg Coquillo: Yeah. So market signals to me and, uh, in just to say, you know, startups have different phases that they go through.

And the typical one is what we call prototyping. Then product market fit, then growth mode, then protector mode. I'm just naming them in a way that I can explain. So prototyping is more of a, you're building a proof [00:32:00] of concept. You're building this demo that you want to show your initial customers. Then you move to product market fit where you have.

Well defined list of customers that really see value in the product that you built that without this product, a good amount of their operations will cease to exist or will be significantly impacted. And to me, that's product market fit. Where you're starting to find this kind of like inertia to kind of kickstart, you know, the engine of that startup, right.

Where you're cementing your way into this market space. Right. And then after that, you go into what I call growth mode, where you're now trying to figure out who else in this market space, maybe having the same problem that [00:33:00] we need to bring in with these. vested customers and how do we get them? How do we convince them that the traditional way or the tradition of vendors that they were into are archaic or no longer solving their issues, that I am the one you want to get to.

And that's a very hard place to be. That's a very, very difficult. Place to be, then if you achieve a good growth rate now, as you acquire, as you've acquired your initial customers, you're growing, et cetera, you turn into more of a protector. Now, now you say, okay, I've acquired these big companies, these big fish.

How do I keep them there for the longterm? Now you're starting to think about what are the features you need to launch. To keep them or keep that value [00:34:00] proposition high for them so that they don't leave me to a competitor. And that's another very hard phase to and as companies travels these phases, you want to make sure that they have the right tools to survive.

These phases, right? I may have diverted, uh, my response to another way, another thing, but just if you could remind me of the essence of your question again, I can zone in on the right answer for you. 

Louis-François Bouchard: My question was the market signals to know when to pivot? Yes, 

Greg Coquillo: yes, yes. Absolutely. The market signals I remember now.

So knowing these phases, right? We're going to explore, for example, a customer or startup that has achieved, let's say product [00:35:00] market fit. Well, some of the signals that I can think about now I have product market fit. I'm ready to scale. I'm ready to grow. And I'm, I'm, I'm ready to, and growth means that you have to do, you have to test different things, right?

It's like you're casting your fish. Net or your, uh, fish wire, right? And then you're trying to see, okay, what fish will bite. And every time you have to test that you see your response. And you do more of it, you pivot, right? Sometimes you have to change the bait until they back, right? Some of the market signals is, you know, looking at how much money are you spending, you know, trying to change that fish bait, right?

And, and seeing the return, right? If you're finding yourself constantly spending money to acquire new customers and you're not. Actually getting that return or that return on investment, right? That cost, a cost of [00:36:00] acquisition is significantly high and you're not getting a return over time. Those, those are the first signals that says, Hey.

You know, maybe you're not knocking on the right doors and you need to really go back to the drawing board to understand the customer space, right? Maybe you've honed in on the problem space already, you want to spend more time on the customer space to kind of like knock on the right doors, right? And if you've exhausted all of that data and you still don't have the right response. Now it's a model of, it's only a matter of time that your original customers may start to leave you because their problems too will evolve. Over time. So, you know, I'm just giving you an example of what kind of signals can you get from the market, right? And there are many, right? We could talk about that all day.

There are many, I was just taking an example [00:37:00] of what kind of signals you need to, and this is where, you know, founders will succeed in pivoting. You know, they succeed really because they are not stubborn. They are willing to listen to the market. They're willing to listen to the experts that they've spent time hiring.

Uh, or the advisors that they spent time, you know, onboarding, um, to find the best way to acquire customers. Right. And not drown in the egotistical journey of believing in an idea that most to dare to contrary to their belief. Most folks didn't want.

Louis-François Bouchard: Would you have any common pitfalls in mind for both advisors, investors, and for entrepreneurs? 

Greg Coquillo: Yeah. I mean, there are some common pitfalls, distractions, trying to solve everything at the same time is one that is [00:38:00] so underrated, right?

So, so many founders think that solving it all is the best approach. And I'm going to get another one too, that's more pertinent to today's condition is quickly calling themselves a Gen AI. Startup to me, it's a pitfall, because what that tells me is you're not willing to tell me what's really going on under the hood.

If you cannot express that in a clear way, then I'll question whether you're an AI startup or not. Right. That's another pitfall. Another pitfall is, you know, not aligning as a founding team, right. When there are diverting. Opinions floating around about how to approach a problem, how to solve a problem for customers, that's another pitfall I can think about scaling too fast [00:39:00] is another one is a little bit counterintuitive where when startups get money too much too fast.

And we've seen that right in the AI space where VCs were willing to cut big checks super fast before even startups had a product or working product or working prototype. that, you know, founders get that complacent. They felt that, that, that money they had in the vicinity in the bank was some sort of blanket protection against any conditions, right?

Those are the common pitfalls that I can think about. 

Louis-François Bouchard: Awsome, I a bit related to pitfalls, but also the, the market signals that my last question is to. I want to ask you right now, um, from what we are seeing for the jobs and like what is, I guess, publicly available, the market doesn't seem ideal. [00:40:00] So I'm wondering what are your thoughts on the current market for investing and for startups?

And also where is the potential or where do you think it will be like soon? Is there any industry that is booming or where's the potential in both startups or investments? And is what are the market conditions right now in the startup 

Greg Coquillo: world? A quick response to that is nobody knows, I don't know. And, but also I can say opportunities are everywhere.

Um, the startups that I'm involved in right now that fascinate me are the ones that are able to explain what happens if they remove the AI aspect of things. How will we solve a problem? If they can tell me how they will solve a problem without AI, then I can get very excited about that. That's a [00:41:00] good interview question.

Right? So, and I'm not talking about the entropics of the world, because they are like really AI startups, the open AI of the world, right? Without AI, open AI doesn't exist. Yeah. I'm talking about the companies who are vowing to leverage AI to solve business problems, right? I'm talking about the AI applications or AI enabled applications.

If I remove that aspect, how are you solving that problem? And oftentimes you will see that it may become an engineering problem or architecture problem. And then the AI is an added tool that gives you some sort of competitive advantage. But overall, if you're able to express that end to end user experience to me in a way that convinces me that competitors cannot copy it, then it's more of a compelling [00:42:00] story that convinces me that this is something that's interesting for me to invest in, right? I don't know if I'm answering that question, but you know, it's, that's the way I can see it now. you think about another company, let's say somebody comes to me and say, Hey, I'm building a competitor to entropic, then, you know, it's a different framework that I would use for that.

Right. First of all, how much money do you have? Because these guys were ingested with hundreds of millions of dollars to be where they are. How do you tackle bigger aspects of things right now? You're dealing with a foundational model that needs vast amount of data. So how do you tackle the big issues of today?

Data privacy. Data security and everything like that, bias, et cetera. How do you tackle those hard to solve problems? How will you solve it better than the existing competitors today? The open AIs, the clods, the, the, the interpigs, et cetera. Stability AIs of the world, right? So it's a, it's a different framework to apply [00:43:00] there.

And believe it or not, there are, there will be more foundational models that will come to life. And one thing that I think will happen is over time, these foundational models will get cheaper. To build because we'll achieve some training optimization and even inference optimization. So, uh, it's an interesting space.

So I'm excited. 

Louis-François Bouchard: This is a quick interruption to remind you to leave a like and a five star review. If you are enjoying this episode, it helps a lot both to let me know that the episode was good, but also to make other people discover it. Thank you for listening, and I hope you enjoy the rest of this episode.

You like to Digest what's going on in the AI space, but why did you start doing that first? When did you start doing that on LinkedIn and why LinkedIn and what were you doing? Like your first few posts, [00:44:00] what were they about and why? Yeah. 

Greg Coquillo: I mean, at first LinkedIn, I saw it as an avenue for marketing yourself to find a job.

It did have this social aspect to it, where at first you only connected with folks who worked at the same place as you. And then later on, this is to resume for you, I can say something like curiosity is what got me there. Curious about new ways of solving problems. So back in 2019, for example, I started.

You know, becoming very curious about how to code in Python, for example, and I started sharing my journey on LinkedIn, and this was about the same time where building machine learning experiences were coming up about, [00:45:00] right? So I was learning how to build the simplest experiences in Python, sharing my, my, my journey and at the same time, AI is coming up. So I started reading about it. I started to learn, you know, how Python fits the profile as the go to language for coding machine learning experiences. And, you know, talking about them, talking about the subject from time to time, and gaining traction in terms of folks who were probably on the same journey as I was.

What really helped me is being able to be being vulnerable, being able to share an opinion and not stick to it necessarily, because opinions can change. Opinions should be able to change based on a better perspective that you may receive from someone else. So I didn't [00:46:00] take anybody who contradicted what I said personally.

I actually saw it as an opportunity for me to learn even more. And over time, I got more confident. About what I was sharing. So just to resume for you, curiosity got me there and an alignment with what I've been doing as well in the past, which is manipulating data to generate business value and AI is doing just that.

And, uh, really building some sort of discipline around posting on a regular basis. So now you create a habit of folks waiting for you to share your opinion on something and also feeling, making people feel welcome. So when I post, somebody puts a comment in, I would reply back, et cetera. And over time I was favorable.

I mean, the LinkedIn algorithm favored my, my post [00:47:00] and more and more people starting to see it until I got the recognition. And the other thing that happened in the mean, in, in the meantime, in parallel that I can't deny is that me joining a, a fan company also helped or shall we call them main now. Also helped because I had the opportunity to raise my voice and say, Hey, I would like to play here.

And I was grateful enough to have leaders that agreed that this was a good space for me to evolve in. And I was. lucky enough to be able to tinker in that space. And that also helped me gain confidence. And that also helped me with, you know, not getting approval, but at least getting the attention of the community.

And when you reach these kind of position, when you reach like a company that has high visibility, when that you start working for, you have to be [00:48:00] conscious of what you say. You have to, it's a responsibility to not mislead people. So, yeah. Before I put a post out, although I know it may not be 100 percent correct, and I'm open to feedback, I also have to make sure I do my due diligence so that I don't share things that may be damaging or, or else, right?

Louis-François Bouchard: Are you still posting every day?

Greg Coquillo: I try to, uh, at least I try to do five days a week. These days it's been a really challenge for me to manage. My calendar, um, I, I found myself needing and wanting to spend more time with family and that really is what's important to me, even though it doesn't take me much effort to post anyways on LinkedIn.

The reason is. I spend a good amount of time [00:49:00] curating, not curating, but setting up the information pipeline that comes to me. So basically identifying good article sources and other avenues that brings me information that I can then digest and explain in a. More digestible way to my followers and because I have these Avenues or these articles these quality pipeline of information It takes me a couple minutes to just share an opinion on what I've read And post, right?

So it's probably more taking me more time. Now is when a lot of people are commenting. I have to take some time throughout the day to answer them because I do not like to leave them hanging as they say, right? But regardless, I try to set boundaries in terms of like. What it means to be on LinkedIn, what it means to be at work or what it means to spend time with family and [00:50:00] what it means to spend time with myself as well for my mental health, right?

So I've been trying to lean in more towards, uh, that versus. Uh, getting lost in the LinkedIn verse. So, 

Louis-François Bouchard: and why LinkedIn rather than having a blog or doing videos or a podcast, you seem to be able to talk like very well. And also you have a lot to say, so why not having your own platform or something you could go like, I don't know, more in, not necessarily more in depth, but let's say why not trying another platform.

Greg Coquillo: Yeah, time management, crowdliness, uh, I mean, if you think about YouTube, it's a very crowded space. Twitter as well. I guess it's just luck. There's a luck factor there too. [00:51:00] And then also from a time management perspective, earlier on when I was starting to post, creating videos takes so much energy for some reason.

Yeah. You have to take and retake, take and retake. So I started tinkering with that and I wasn't even doing like five minutes videos. I'm talking about the 30 second videos that was super high energy for me. And I was like, okay, knowing this, I definitely do not want to have a YouTube channel knowing that I'm, I'm working full time.

I have a family, et cetera, and that was a easy decision for me. Right. It's so TLDR here is time management, right? And available time for you to, to, to do those. Like I'm not ruling out ever creating a podcast or having a YouTube channel. It's just that cold start that I'm kind of stuck with right now.

Where it will feel for a long time, uh, lonely in these [00:52:00] spaces. And I felt that on LinkedIn, but I didn't really care for it because I was so passionate about the subject that I felt like some folks eventually will start listening. And they did where, when I look at the value to effort ratio of doing the same on YouTube or Twitter or Tik TOK is not there for me.

It seems like reaching critical mass will take way too long because of so much noise already happening in these platforms. Yeah, I 

Louis-François Bouchard: think I completely understand that's also how I feel just for like, I've, I've started on YouTube and, uh, I guess I also got lucky after, after months and years of, of weekly posts, but, uh, I got lucky that it, it, it worked somewhat well.

And it's true that, for example, I've, I've tried going on Twitter and it's. I [00:53:00] don't know. It's, I don't, well, first I, I'm, I'm extremely bad with social media. Like I don't understand how, how to waste time on there. And I just can't have Twitter open. I don't know what to do. I don't, I just don't like that.

And, and so it's extremely hard to try to post like on, on Twitter. You, you, if you want to grow, you need to post multiple times a day. I think like you need to be very active and basically part of a, of the Twitter community. And that's something. Yeah, incredibly demanding and hard, and it's extremely saturated as well.

I don't know if, like, what Elon Musk is doing is helping or not, but it's, it seems quite impossible to, to start. Not impossible, but Yeah, extremely complicated to start and grow on such a platform. So I definitely make sense to, yeah, I guess just like leverage what you currently have and what you built, which also took a lot of efforts and, uh, and [00:54:00] luck.

And I just wonder, um, was it your idea from start on LinkedIn to build some kind of personal brand or it was more to share? Like to explain things like what, what was your, your ultimate goal? Is it to build like the Greg branding and, and try to, to yeah. Build your personal branding to, so, so that it has rewards or something attached to it, or was it more for other reasons?

Greg Coquillo: No, to answer shortly, uh, it was. More about wanting to share things that are excited and learn in return, learn more in return when people interacted with, with my sharing. So LinkedIn felt more of a natural space for me [00:55:00] to start expressing those things. And with a mixture of discipline and luck, it kind of worked out in terms of.

Reward. I can say yes, because of the community, you kind of build this brand, but really the real reward is having a community that helps you learn from different perspectives. Yeah. Now LinkedIn is different from others. It's not like I'm getting a paycheck from advertisers that are accessing my. Post anything like that, right?

To get more impressions for their products is different, right? You don't get money from having a sizable Community on on LinkedIn you could you could do so on other platforms like like like YouTube, but it's not the same on LinkedIn So the true reward for me is really having a community that can help me be a better [00:56:00] person a better employee Or a better citizen, a better professional when I need help.

That's, that's the true reward for me. I can lean in lean on, on my community to find, for example, a sales expert that I can connect to a startup founder, who's my friend. Right. And therefore I create value for, for both people. Right. And, and to me, this is super, super rewarding when I'm able to connect the dots or make an intro or solve a problem myself, or right.

When someone comes to me and say, Hey, I need help. And I'm able to spend time and help that person that really feels good. And everything else works out right, like, or fits in anything, whatever is going to work, I was going to work out, but it was never about making a conscious decision to improve or [00:57:00] create Greg's brand.

I, that was the last thing in my mind. And it's still the last thing in my mind, like. What does that mean anyway? Do companies really pay attention to that? I don't know. Like we're going to hire Greg because of his brand. Well, is that true? Or is it we're going to hire Greg because he knows a thing or two about the position we're trying to hire for?

So that's a little bit more constrained. And at the end of the day, that's probably the nature of most social media platforms that reward you with a community. When you are among the folks, uh, sharing information that most people care about. So if you look at Facebook, Twitter, YouTube, you'll see the same kind of.

Pattern [00:58:00] with folks who are sharing this, uh, the same around a subject. And then now you become the person people look out for to ingest information. So 

Louis-François Bouchard: now in, in artificial intelligence, mainly there's a whole new field of like curators that didn't exist before. It was just like, we, we followed the news and that's, that's pretty much it.

But now it's, it's kind of a full time job to. Figure out what's important or not from like everything that is released. Every day or every hour. It's just, it's just crazy. Even on LinkedIn, there are like multiple people that I follow to, to see the new research and the new cool things. It's just, yeah, it's just very, it transformed a lot since, since when I started and I didn't even start it long ago.

It was like in, just like you in 2019. [00:59:00] Yeah. It's really crazy how, how much it changed in four, four or five years. 

Greg Coquillo: Lots of, lots of noise. And the funny thing is. I don't know how much effort people are making to ensure that they are ingesting the right thing. And that's a personal effort you have to make is to build that quality pipeline of information.

And you may say, Hey, I know Louis Bouchard. I know when I hear him put out a new YouTube video that I do not have to go out there and confirm that Louis Bouchard has done his job. To make sure that he's reporting something accurate as accurate as possible, right? So. Not a lot of people are maybe, maybe, maybe not a lot of people are making that effort to ensure that their, their pipelines are of high quality, right?

[01:00:00] And because of this increase in content creators with the rise of Gen AI, et cetera, you know, you will find a lot of misinformation and in other negative effects that will further divide people's opinions. And hopefully there are also forces at play that are vowing to reduce that those negative effects.

Louis-François Bouchard: As you know, on, on YouTube, I often share and explain research papers in artificial intelligence and I try to like explain them as simply as possible. I always knew that, of course, when I take a paper that seems interesting, that I see that has potential, like a stable diffusion. I, I saw the paper, it was really cool.

It wasn't a stable diffusion. It was latent diffusion and way before stable diffusion. was created, but I knew that it has potential. I didn't know it was [01:01:00] going to be such a big thing, but I knew that making a video about it should get some like somewhat good reach or whatever. And it did work that way back in the days.

But now, for example, I recently did a video about. A paper that was just released like on the same day that I basically, um, the DALLE 3 paper was released like Monday morning and I saw it as soon as it was released. I was super excited, so I just worked on it all day and released the video like at the end of the day.

And. I was the first one about the first content of explaining the research papers of Dally 2. Pretty sure it's, it's been that way for DALLE 2 and for other papers that I've seen like first. And the videos always performed well. And like, it's obvious just because it's a new thing that is really cool.

And I explained how it works and it's nice to me. But [01:02:00] for this time, I, I published it and it made like almost no, no views. And then I looked on YouTube and I saw like that there were at least hundreds of videos that said DALLE 3 explained or how does DALLE 3 work, but they were all like a month old or like one or two weeks old where there was no information on how it worked, but People explained, like, how it works on ChatGPT, like how to use it and things like that, but it was not at all about like, what's DALLE 3, like, how is it trained and what's the architecture or like, what is it?

And it is just like these, I don't know, keywords or like this, this sentence of like trying people that want to find how DALLE 3 works or how it was trained. It's just like extremely diluted and [01:03:00] so many people just sharing the quick news of like DALLE 3 on ChatGPT that you can use right away. And it's I guess it, it, it's the same on, on LinkedIn and other platforms, but I, I definitely, I saw that, uh, last month or like two months ago on YouTube and it definitely hurts my, my channel, but like, it's, yeah, it's annoying.

It's dangerous also for many people just to, they won't just take that specific example, but if they want to understand more about generative models that, that creates good images, they will just find videos. Of people using ChatGPT to prompt DALLE 3. Like it's not related to what they are looking for, but it's like they are saturating the keywords and all the thumbnails and all the things that make it seems like that they will explain you how it works, but it's, it's not at all this.

Greg Coquillo: Yeah. And that, that's what makes it difficult, [01:04:00] right? How does the algorithm know that Louis's video is what actually goes under the hood to explain how dietary works versus more of a shallow one, although the keywords are perfect for that video to get traction, to get eyeballs. But when you actually watch the video, it's really not what it seems.

It doesn't really go under the hood. It just tells you, here, here's how you use it. Yeah. Which, ChatGPT, et cetera. But it doesn't really explain, oh yeah, here's how it's working. Here's why, in a very layman term, like as you usually do, which are videos, right? So now how do we ensure on YouTube that the algorithm will exactly know?

Who the user is, what the query is from the user saying, Hey, I want to understand Dali. And understand that Lewis's video is the best fit for that, for that [01:05:00] query and really minimize the result of those who don't really, who don't actually go deep into the details, right? Versus what you typically do. And it's a difficult thing, right?

Because so many choices now that folks don't know what to pick. And overall, you have so many content creators that the number of viewership that you have goes down. And I've seen the same thing on LinkedIn, too, where only a few were bold enough to talk about AI. And they got the viewership, which went down significantly this year.

Yeah. Because now so many more people felt comfortable talking. And, uh, some information may or may not be accurate or solving a user problem. And by user problem, I define that as a user who wants to have access to [01:06:00] validated information. Uh, so the algorithm is trying to do its best to favor a content creator.

Uh, but at the expense of individual thought performance. Right. So, 

Louis-François Bouchard: yeah, I assume we will soon reach a point where all the content for videos or posts is, is processed into the algorithm, not just like title and, and, and stats, and it will just understand what the person is looking for and what the video is providing and match the best thing, I guess, Jenny, I will help for that in the near future.

Greg Coquillo: Well, yeah, let's hope, but you know. There are some things that other people play on a lot, like thumbnails, for example, as you mentioned, I may have a good enough title, but if my thumbnail is, I [01:07:00] cat, I cat is an eye capture. Then I may generate more views than you, right? Where I make sure I show up maybe at the mid page of the query result, or maybe at the bottom of the query result, but because the thumbnail is so.

Is an eye catcher that I end up getting more views than you, who has probably spent a fair amount of time perfecting that title, uh, to, to get more eyeballs. Right. So there are so many levers that you need to play with that is, that are hard to control on platforms like YouTube. That is really hard to acquire organic views today.

It's more, it's harder. Versus yesterday where there were less player in the field and that's happening across platforms, not just YouTube. We've seen that on LinkedIn as well. Yeah, 

Louis-François Bouchard: I guess I would recommend [01:08:00] anyone starting like a journey into trying to either make videos or having a blog or Twitter or LinkedIn.

But I will strongly advise to not have any, um, goals in mind or goals in mind or like high expectations just because it's definitely. A lot of work, very difficult, you need to be super lucky and extremely consistent and still you may not get anything out of it. So you definitely need to do that for yourself and to learn and so even like, for example, for a Day 3 video, I don't care if it didn't work well.

It just, it helped me understand the More specifically, in this case, it was a training of the algorithm, which was really cool. And I just enjoyed making it, so it's like completely fine. But yeah, it's, I don't know where, where it's, where it'll go, but I'm not sure about the, the future of like content creation and things like that.

It's a [01:09:00] bit scarier with. Yeah. All that's, that's going on. Yeah. I'd love to talk with you more about artificial intelligence. We didn't even enter the topic, but yeah, we could, we could definitely do another episode on like the just large language models and. All your thoughts on what's happening since ChatGPT is just crazy.

Yeah. Thanks a lot for taking the time for talking with me. And if, please feel free to share anything you'd like to share to the people listening and why you can also tell us why people should follow you on LinkedIn, what you're sharing more precisely these days and yeah, why they should check you out.

Yeah. I mean, 

Greg Coquillo: the only thing I can say is, you know, stay curious, lean in more, say yes more. Uh, and say yes more doesn't mean you say yes to everything and anything. You just say yes more when an opportunity shows up. You [01:10:00] may, you have no idea how that may change your lives forever, right? Lean in more. And, um, you are part of this movement more than you could ever imagine, right?

Just because you're not, yeah. An AI expert does not mean you cannot become one or you cannot contribute. So, and in terms of LinkedIn, I mean, it's, it's very subjective, right? You may find my post useful. You may not find my post useful, but at the end of the day, if you give me the honor to consume my post and give me some feedback, then I can vow to become a little bit better over time, which will then, you know, hopefully serve your needs more.

And serve the people who fit your profile more over time. So that's, that's what I can ask for. And I thank you too, Louis. So for such a great conversation, I had fun. Yeah, 

Louis-François Bouchard: I had a lot of fun as well. And yeah, thanks again [01:11:00] for the hour and a half and all the amazing insights you personally gave me just regarding startups and I'm just getting into that.

So I, I really love talking about this and you, you helped me a lot and I'm sure you help a lot of people listening. So thanks again. I wish you an amazing weekend and I will see you next 

Greg Coquillo: time. I think you too. All right. Talk soon.