Skip to content
The Twenty Minute VCThe Twenty Minute VC

Daniel Khachab: "We Are in the Middle of a Cold War for AI Talent" | E1220

Daniel Khachab is the co-founder and CEO of Choco. Today, Choco’s AI platform facilitates half of all food traded in major cities like New York, Paris, London, and Berlin, cutting food waste and streamlining distribution. Since its founding in 2018, Choco has raised $330 million from Bessemer, Coatue (its first European investment), and Insight, reaching unicorn status within 2.5 years. Previously, Daniel was the youngest Managing Director at Rocket Internet, where he oversaw growth across Latin America, Southeast Asia, Australia, and the Middle East. ----------------------------------------------- Timestamps: (00:00) Intro (01:02) Pivoting from SaaS to AI (05:00) Which Jobs Have Changed Most & Which Will Change Slowest? (06:31) Where Is AI Impact Truly Significant & Where Is It Still Lagging? (08:33) Is SaaS Ending, Marking a Shift to AI-Driven Software? (10:44) Is Europe a Decade Away from Adopting an AI-Driven SaaS Ecosystem? (14:07) Should Foundation Models Build Apps or Enable an App Ecosystem? (16:48) Does AI Lead to Job Loss or New Opportunities? (23:05) About an AI Talent Cold War (25:48) With London’s AI Ecosystem, Why Base in Berlin? (29:21) How Can Europe Strengthen Its AI Position in Chips, Models, and Energy? (34:43) How Does Regulatory Fragmentation Across Countries Impact AI? (36:52) If Ambition Is Key, Why Not Expand to the US or UAE? (38:39) Covid Time (47:37) On Work-Life Balance (48:53) Lesson on Fundraising (56:41) Why Spain is the Worst Market for Choco (58:45) Raise to Compete or Avoid the Frenzy (01:01:54) Quick-Fire Round ----------------------------------------------- From Seed to $1BN in 30 Months: 1. We Killed a $BN SaaS Business to be AI First: Why does Daniel believe that SaaS is dead? What does an AI-first company mean? Why does Daniel believe AI-first companies will win the next 10 years? What foundation models does Daniel and Choco use today? How has the cost of using different models changed? What categories are vulnerable to being attacked with vertical products from the foundation model providers? 2. Europe is F*******: Why and What To Do: Why does Daniel believe Europe is at a massive disadvantage in the next 10 years of AI? Chips: What can Europe do to encourage chip production and manufacturing to take place on European soil? Energy: What can European governments do to encourage energy providers and new forms of renewable energy to innovate to provide the energy AI needs? Talent: Why does Daniel believe AI talent is the hardest problem that Europe faces? What can governments in EU do to resolve this problem? 3. Lessons Scaling to $1BN in 30 Months: Does Daniel regret raising at a $1.1BN valuation? Why did he throw a unicorn party with the round? Why does he regret it so much? What did Daniel spend money on that he wish he had not spent money on? What did Daniel not spend money on that with the benefit of hindsight, they should have spent money on? When your competition raises a lot of funding, does that mean you should also? ----------------------------------------------- Subscribe on Spotify: https://open.spotify.com/show/3j2KMcZTtgTNBKwtZBMHvl?si=85bc9196860e4466 Subscribe on Apple Podcasts: https://podcasts.apple.com/us/podcast/the-twenty-minute-vc-20vc-venture-capital-startup/id958230465 Follow Harry Stebbings on Twitter: https://twitter.com/HarryStebbings Follow Daniel Khachab on Twitter: https://twitter.com/DanielKhachab Follow 20VC on Instagram: https://www.instagram.com/20vchq Follow 20VC on TikTok: https://www.tiktok.com/@20vc_tok Visit our Website: https://www.20vc.com Subscribe to our Newsletter: https://www.thetwentyminutevc.com/contact ----------------------------------------------- #20vc #harrystebbings #danielkhachab #choco #venturecapital #ceo #founder #fundraising #worklifebalance

Daniel KhachabguestHarry Stebbingshost
Oct 28, 20241h 17mWatch on YouTube ↗

EVERY SPOKEN WORD

  1. 0:00 – 1:02

    Intro

    1. DK

      ... 100%. I think SaaS is dead. Great companies, at least from headcount perspective, get smaller and not bigger. I mean, you go in a, in a different country fighting for talent, that to me is Cold War. In Europe today, we're not producing the chips, we're not producing the energy, we don't have the foundational layer models. Now the real downside to me is that it's that one billion mark. Once we come unicorn, then people think, "Oh, we made it." Like, who the fuck wants to be a rainbow-colored pony? Like, I don't want to be a unicorn. We need more long-term commitment as well. Like, we need founders to say, "I'm going to vis- invest 15 to 20 years. I'm going to commit this now, a substantial part of my life, the best years of my life to make this happen."

    2. HS

      Ready to go? Dan, I am so excited for this. Dude, we met at a KOTU dinner with Dan Rose a while ago now, so I'm so excited that we can make this happen in person.

    3. DK

      Very likewise. Thank you for having me. Really looking

  2. 1:02 – 5:00

    Pivoting from SaaS to AI

    1. DK

      forward to this.

    2. HS

      So I want to just dive right in. You were a SaaS company.

    3. DK

      Yeah.

    4. HS

      You were a darling of the, like, traditional SaaS market, raised a lot of venture money. Why did you decide to pivot from SaaS to 100% of your revenue coming from AI?

    5. DK

      I think what kind of gave the spark was just a couple of sleepness, sleepless nights in a row. Because, like, this new technology came out, GPT, and I, like ChatGPT, and I, I played around with it. And then at some point, I, I just couldn't sleep and was literally sitting on my couch, in the dark, staring in- into nothing because I was like, "Man, this thing is gonna learn how to code in no time. And maybe not now, maybe not next year, but maybe '27 and t- maybe in '28, like where's our technological moat is gonna be?" Like, like, it might just be gone, like anyone maybe able to replicate what we have built over years within days. And so I'm like, "Wow, that's gonna be a big shift in company building, and how we gonna position ourselves to win in, in, in such an environment?" And, and while, while I believe that there is some moats, um, I think there were, like, three things that, um, that made us go A- made us go AI first. And, and I think the first thing was, is like, hey, like there's two scenarios that are gonna happen. First scenario is someone is gonna come and replicate our technology, do maybe things in a better way, create mo- more user value, and it's gonna disrupt us. Scenario two is it's actually us doing that. Us disrupting ourselves, um, and hopefully also our, our, our, our competition. And so one of our core values is always play offense, and so we gotta play offense also in that, and we actually have to be that, that driver. I think that's one. The second thing is, you know, if it's true that company building is gonna change so fundamentally because so much value is coming out of the API, then it's also true that there's a lot of skills to be learned. And, um, for every single role within a company, and so we might not know how the future's gonna look like, but the best way to position us for the future is to learn how to build with AI on every single part of the organization. So essentially we're in a race to upskill our teams. And, and, and more so, you know, I w- I told our team at, at some point, you know, "If you were to leave JhoCo tomorrow, and you would not have the skills on how to build with AI, then your skillset will be obsolete." So, so it's also our responsibility as an, as an employer to give our employees the opportunity to learn the most relevant skills, um, of the century.

    6. HS

      So normally, I am, like, pushing CEOs away from selling their company, okay, 'cause they, your promotional shows are never good shows.

    7. DK

      Yeah.

    8. HS

      I'm about to defend your company in the stance that you're not taking. You're like, "Hey, we had to make this shift." I would argue, and this is why I, I want this debate-

    9. DK

      Yeah.

    10. HS

      ... you have proprietary relationships with food distributors.

    11. DK

      Yeah.

    12. HS

      You have a huge amount of data on products, on financials, on invoicing, everything around that. Those are two huge moats that I could come in with a better product. I do not have the historical da- relationships, I do not have the data moats. Uh, technology doesn't matter in that respect. Why does Facebook open source LLaMA? Because the value's in the data that they have.

    13. DK

      100%, I don't think all of the moat is gone, but I think a significant moat is gone. Like, we've still built technology for, for up to seven years, and, and, and that holds for many companies who might have built tech for five years, 10, 15, 20, doesn't matter. And so it's just part of their moat, and that part might vanish, and there might still other be moats, and, and, and we gotta focus on those more for the future, but that might still be gone very quickly. I think it's important to recognize that many jobs will change fundamentally, and, and what these people do every

  3. 5:00 – 6:31

    Which Jobs Have Changed Most & Which Will Change Slowest?

    1. DK

      single day.

    2. HS

      Which jobs have changed most significantly today, and which will be slower to change?

    3. DK

      I think one of the easiest examples to describe is probably, uh, product design because what if, you know, previously, you might've have, might've had, like, a, a UI that was very rich, and let's say you wanted to do payroll, and then you click. You se- you have a search bar, and you search for the, for the right employee, and then you click into that employee, and then you maybe check, you know, like how much bonus should they get. And then you, then you create, like, this PDF, and then you send this PDF kind of to your tax advisor, and then maybe to your bank to, to, to essentially pay out the money. And so in the future that, that just might be a prompt. Yeah, please do payroll, and please let the tax advisor and the bank know automatically, and just report back to me once it's done. And so what UI is there to design besides that chat? W- who are you speaking to? What is the character of that, of that AI, of that agent, if you want, so that you're speaking to? Is that like a very serious structured kind of character? Is it a funny, cheeky character? Um...... are they, um, kind of very concise and precise in their answers? Are they more like conversational and try to take in, do they maybe even flirt with you? Like what kind of character do you want to design that is most appropriate for that particular job? And so a designer would probably have to do that in the future before they've been drawing like great simple user interface. And today it's actually like which character is AI

  4. 6:31 – 8:33

    Where Is AI Impact Truly Significant & Where Is It Still Lagging?

    1. DK

      going to have?

    2. HS

      Other than developers-

    3. DK

      Yeah.

    4. HS

      ... where Copilot is still a long way off and has got a lot of problems-

    5. DK

      Yeah.

    6. HS

      ... hence the rebrand, where are you seeing meaningful, meaningful... I saw in the AI first that 80% reduction in customer service cost.

    7. DK

      Yeah, yeah, yeah.

    8. HS

      But like build with AI, it's not re- it's not really changed much yet.

    9. DK

      Let's say in the future we're actually not... So today most SaaS companies what they do, including what, what we did, is like okay, we talk to the, to the customer, we identify the problems of the customer, and we say, "Look, we've got here a neat solution for you." And then this is how the solution works and these are the features and so on and so forth. But in the future, we are more or less selling an, an employee. And like, you know, here's a very highly competent employee that is spe- that has the spec- spec- spec- specific skills, sorry, to perform in your company. It's not a- never going to take a cigarette break, it's never going to take a day off, and you can hire 0.1 of these employees, you can hire 1,000, you can hire 0.1 in the morning, you can hire 1,000 by noon, and by night it might be 0.1 again.

    10. HS

      Mm-hmm.

    11. DK

      And so how do you demo this, right? And how do, how do you, how do you negotiate a price for it? And so, you know, even in a, in a job that is probably the farthest away from engineering being sales, like there are massive changes. How do you market that employee, right? So, so for marketing, there's, there's, there's going to be massive changes. And then obviously in engineering it's like, yeah, okay, engineering, we got to plug into the APIs, but how do you QA an employee? Because if I put something to LLM, if I put the same question twice, I get a different result. So how do you QA that? Every job will significantly change. Um, and even internal facing ones, it's like let's go back to the payroll example. If I work in HR, in HR today, and many of the things I spend a lot of time with will just be, you know, done in seconds, then h- how do I stay on top of things? How do I not become obsolete in the future? It's like every single job will

  5. 8:33 – 10:44

    Is SaaS Ending, Marking a Shift to AI-Driven Software?

    1. DK

      change.

    2. HS

      So when I summarize this, respectfully, are you saying that we're seeing the end of SaaS and the agentification of all software in business as a fundamental transition?

    3. DK

      100%. I think SaaS is dead. It's, I call it agentification of, of SaaS. Um, in some industries it might happen sooner and in, in some later. But, but naturally, why do we have user interfaces? It's... So, so I like kind of the HR example because I still remember before all of these great HR tools came out, we were managing all of HR on Excel sheets. Not Google Sheets, but Excel sheets. And it was kind of every single employee and when they joined and how many holidays they took and what they earn and yada yada yada. And then this, these great tools came up which made the job much easier. But you still need to learn how to use that software. Whereas in the future, we're just going to prompt it. You don't need to learn anything because you communicate with that machine like you communicate with a human, in voice, in written form, in m- in many different things. And so it's just the adoption curve will just be faster because there's nothing to learn.

    4. HS

      So just so I understand, so we all have... we will still have single sources of data truth, which is like your sales force, which is your repository of sales data, correct? And then we will have like application layer on top of them, which would just be prompt engines? And it'll be the same for HR, the same for payroll. Is that it?

    5. DK

      Could be it. Could also be that, um, kind of in that example, sales force should be the application layer itself. There should not be lay- a layer above sales force. It should be the application layer, and what's underlying might, might just be a database. And so it's like, you know, sales people hate sales force.

    6. HS

      (laughs) Yeah.

    7. DK

      Like go on there, update every single lead, "Oh, I spoke to that person." Upload the presentation that, uh, that, that you have done, change the lead status, who they talked to, or like the points of contacts, whereas like, okay, like here are my notes that I took anyways. Okay, do the rest. So obviously it's, it's, it's a... And you can imagine if you, if you ask (laughs) a sales rep, "Hey, you have two options. Fill the form or copy paste your notes." Like what are they going to choose? And which tool is going to find more adoption? And which tool makes your sales rep more efficient?

  6. 10:44 – 14:07

    Is Europe a Decade Away from Adopting an AI-Driven SaaS Ecosystem?

    1. DK

    2. HS

      So I have so many areas where I want to go in this one. The first one I'll just say is like we- we're in Europe. Um, in Europe, 62% of large enterprises still don't know what Slack is.

    3. DK

      Yeah.

    4. HS

      91% don't know what Notion is.

    5. DK

      Yeah.

    6. HS

      Respectfully, getting them to move to cloud is a challenge in itself.

    7. DK

      Yeah.

    8. HS

      Embracing an agent-based SaaS ecosystem is 10 years out, Daniel.

    9. DK

      So this is where I disagree. I think it's the other way around. So I think AI is the perfect technology for traditional industries. Why? Because the problem in adoption is not that they think, "Oh, digital is unimportant." The problem in adoption is like, "I need to learn something new. I have done this forever. I don't want to change. I'm going to need training. My people are not ready for it." That is the problem in adoption. Versus now it's like, "No, you don't need to learn anything. Do your people know how to use WhatsApp?" "Yes." "Okay, it works like WhatsApp." Type in what you want, it's going to give it back to you. So the adoption curve is going to be way faster. Like AI is the perfect tool for traditional industries. Much more actually than for, for kind of like startups and tech because those people that are tech first, they know how to work with interfaces and stuff like that.

    10. HS

      Totally understand that perspective.

    11. DK

      Yeah.

    12. HS

      How do you respond to the data security compliance, large enterprise, gnarly logistic challenges which prevent the adoption of new technologies like this?

    13. DK

      I don't think everything needs to be an LLM and not everything needs to come off the cloud. I think for many use cases small language models are completely sufficient and many of which can be hosted on pre- even large language models, they can, can be hosted on premise. Nothing's going to leave your doors.... so you can create a product with this as well, um, and it's completely hosted even on your own int- inter- internal service. Um, and so I think you've got to go through the same questions around data security that any SaaS has, uh, h- has to go through so that's maybe, maybe a one-on-one. But then on the adoption curve, you still win.

    14. HS

      I, I speak to many, obviously, AI leaders-

    15. DK

      Yeah.

    16. HS

      ... uh, and they say the single biggest problem right now is actually implementation. It's not the sales cycle. There's budgets and they want to spend on AI, but it's implementation, it's data readiness, it's data cleanliness. Uh, and we have this post-agreement process where it's like, "Ugh. I don't know what to do now."

    17. DK

      Yeah.

    18. HS

      How do you think about that?

    19. DK

      So I do think that actually two interfaces will survive. One is kind of like the interface in which you ask the AI to do something and which it returns you what you want. Um, that can be, you know, how much revenue to, did you make last month? It gives you back a number. That can be, you know, make the payroll, that gives you back the result, and things like this. But, um, AI is not... is not God. Like it makes mistakes. It's just an intelligence, you know? It's, it's also not, not... it also doesn't have superhuman knowledge, and particularly not on your proprietary data. And so you've got to train it. And so that's the second interface that I think we will have in the future, and I think that we need to think in a way that, hey, how do we make it as easy as possible for our users to train the E- AI? Like that's the key. It's not going to be everything's going to work from, from day one and that's not what you should expect and you should not oversell on it. It's, it's not like, "What's the day one accuracy?" It's actually the rate of learning that needs to be as steep as

  7. 14:07 – 16:48

    Should Foundation Models Build Apps or Enable an App Ecosystem?

    1. DK

      possible.

    2. HS

      I'm an investor for a living, alongside the podcast-

    3. DK

      Yeah.

    4. HS

      ... and I have to think about where value accrues in the stack.

    5. DK

      Yeah.

    6. HS

      We mentioned earlier you're using, you know, OpenAI for a lot of customer, um-

    7. DK

      Yeah.

    8. HS

      ... support elements. Where does value accrue and what does it make sense for foundation model companies to build applications for, versus what they just let have in an app ecosystem?

    9. DK

      So I think, for example, what Anthropic has done with computer use very... two or three days ago, um, is, is quite magical actually.

    10. HS

      For those that don't know, can you just explain it?

    11. DK

      Yeah, essentially Anthropic will take over your computer and it can, you know, click for you, look up things for you, fill out forms for you, do everything a human does with a computer. It es- ten- essentially take over your mouse and your keyboard, and then you prompt it, tell it what to do.

    12. HS

      Which is essentially the next generation of RPA.

    13. DK

      Uh, it... more or less, yes. A very, very smart RPA. And so one could argue that's application layer because RPAs exist. One could argue that's foundational layer, you know, what is it? I think another way to think about it is, how are incumbents really doing? But I think... essentially I think they're doing quite well and, um... but the problem is, and which is advantage for the application layer for us, it's starting to get more and more commoditized and the price decreased almost on a monthly level.

    14. HS

      100%.

    15. DK

      Like we literally had to do nothing and our price is now 80% lower than six months ago despite probably having five or six extra transactions

    16. HS

      What do you use?

    17. DK

      Uh, mostly OpenAI.

    18. HS

      Mostly OpenAI?

    19. DK

      Yeah.

    20. HS

      A little bit of Anthropic?

    21. DK

      A little bit of Anthropic, a little bit of Mistral, yeah.

    22. HS

      What... to what extent do you care about price versus utility and functionality?

    23. DK

      100% utility and functionality b- coco- just because it's... we are at that level of, of maturity or non-maturity if it's just about like build value first and later on we think about cost.

    24. HS

      Mm-hmm.

    25. DK

      So we didn't invest anything in cost, but our cost still went down despite volume going up and so, so to that point, should application layer... uh, so- sorry, should foundational layer companies go into application layer? Maybe. (laughs) Right? Because the rest might be commoditized.

    26. HS

      The question is, you can buy OpenAI at 160 or Anthropic at 40. Which one do you buy?

    27. DK

      So I'm not sure if these numbers are right but I recently read that Anthropic is more or less doing 60% of the revenue of, of OpenAI and they do this with less customers so they kind of achieve a higher price point probably. Um, plus the recent release of computer use, I find it quite inspiring so probably do Anthropic. Having said that, I mean, you know, there's been so much shit storm about OpenAI, like let's be honest, like they're the reason why we have this conversation.

  8. 16:48 – 23:05

    Does AI Lead to Job Loss or New Opportunities?

    1. DK

    2. HS

      Wha- 100%.

    3. DK

      Yeah.

    4. HS

      If we go back to how it's impacted the core business-

    5. DK

      Yeah.

    6. HS

      ... I completely understand everything that you've said. Every technology cycle everyone goes, "Oh, we're going to lose our jobs, we're going to lose our jobs."

    7. DK

      Yeah.

    8. HS

      And then every technology cycle we actually just find new things to do.

    9. DK

      Yeah.

    10. HS

      We get better, we get more efficient, more productive. Sad but Klarna's fired 700 people in customer support.

    11. DK

      Mm.

    12. HS

      (laughs) Which way does it go?

    13. DK

      When we look at the Western world-

    14. HS

      Mm.

    15. DK

      ... yeah, when you look at Europe and US, the most significant challenge to business is labor shortage. Like we have millions of open jobs that, that we can't fill and not necessarily in tech. Also in tech, in healthcare, in care for children, um, and, and truck driving. These tend to be like the jobs in which we have the most scarce labor. Even, even in hospitality it's very hard for companies to find people to, to work there so will AI maybe help us to, to reallocate our portfolio of labor to where it's most needed?

    16. HS

      Do you not think we have those... I'm, I'm being deliberately divisive-

    17. DK

      Right.

    18. HS

      ... because they're shit jobs and what I mean by that is not shit but they, they're just... they're not great jobs to do. Truck driving is a bit boring.

    19. DK

      (laughs)

    20. HS

      It is, you know? Yeah, cleaning loos in cinemas and restaurants is boring.

    21. DK

      Yeah.

    22. HS

      Being a hotel maid is a bit shit clearing up after other humans and wiping down loos and... they're all shit jo- humans don't want to do them.

    23. DK

      Yeah but, you know, if, if you're sitting in customer care and everything you get all day is complaints, that much better, and plus these complaints are also very repetitive.... and you, and you're, and you're sitting, like, like, there and just also doing the same 20 prompts every single day.

    24. HS

      This is the point though, which is like, AI will just replace the truly shit jobs.

    25. DK

      Yeah. Exactly. And- and- and so, and so- and so that person, like, we need that person very urgently. And now, you don't need to be, you know, work in hospitality, which is a extremely hard job, but like, we need kindergartners. Like, just in Germany, we're short 50K. 50K. And the government has invested billions over the last... we're still short 50K. So, wouldn't it be great just for general society-

    26. HS

      I'm a believer in Adam Smith's invisible hand, you know, where you have a shortage of 50K, you see salary increases for the kindergarten providers, and suddenly that 50K goes down to 10K.

    27. DK

      And if we would have competent government, that would probably work out. (laughs)

    28. HS

      But why does it not?

    29. DK

      So, obviously it depends on the country but in a German, uh, example, it is, um, mostly more or less public servants in which... that- it's not very privatized.

    30. HS

      Ah.

  9. 23:05 – 25:48

    About an AI Talent Cold War

    1. DK

    2. HS

      Did you do the let goes in one go or did you do them in multiple rounds?

    3. DK

      In- in... so, so the AI-related one in one go, yeah.

    4. HS

      Yeah. Okay, good. When we chatted before, you said about an AI talent cold war.

    5. DK

      Yeah. (laughs)

    6. HS

      And I- I was like, "What the fuck is that?" Can you help me understand that?

    7. DK

      I think the cold war probably has, uh, has two components, and one is infrastructure and two is talent, but we can start with talent, and... so we're in London right now. I've been in London together with some other AI founders sometime this summer for a conference. We got an- an- an- an invitation from someone, you know, let's say fairly high up in the US government to meet up for dinner. And we only AI founders, and then of course it was lo- lovely dinner, it was fantastic opportunity for us, um, but essentially they were making it very attractive for us to move our AI talent over. But we are in the UK here. But the next day, there was a similar event from someone from the British government. And la- no, so maybe two weeks ago there was an- an, um, a conference in Berlin, and same same, UAE and- and Saudi Arabia. So people are really fighting for this talent, they have understood how vital it is.And, um, I mean, you go in a (laughs) in a different country fighting for talent, like, that, like, that to me is cold war and I think it's very interesting. And, and, and-

    8. HS

      What do they want specifically? They want you to build your AI teams in their country?

    9. DK

      Yes, and relocate there.

    10. HS

      Do they offer anything in return?

    11. DK

      Yeah, uh-

    12. HS

      What is their sell?

    13. DK

      It depends what you obviously always get is kind of like golden visa and this k- this, this kind of support. Like, the most extreme offers that you get is, like, people are gonna pl- pay salar- like, government is gonna pay salaries for three to five years for your top AI talent. Like, they're gonna do that.

    14. HS

      Why would you not do that? They're gonna take away the cost base of your AI team.

    15. DK

      Y- I, I think you seriously have to consider it. If it's the best for your company, then, then you have to do it, and if your employees want that.

    16. HS

      Where is your team today?

    17. DK

      Berlin.

    18. HS

      W-

    19. DK

      Like, all of our R&D is in Berlin and then we have sales offices across US and Europe.

    20. HS

      Why? Is it the best in Berlin? Is it most convenient? It's 'cause, it's 'cause you're there? Why is it Berlin?

    21. DK

      So we started Choco in Berlin because we felt we had the best network there. We felt Berlin is a very attractive cities, uh, city for foreigners to come. You don't need to speak any German. It's, it's the most affordable capital in Europe. I think it has one of the highest just quality of life, like living standards, um, in Europe.

  10. 25:48 – 29:21

    With London’s AI Ecosystem, Why Base in Berlin?

    1. DK

    2. HS

      But if it's specifically AI, like, there's no great institution-

    3. DK

      There, there can-

    4. HS

      ... spinning out of, uh, out of Berlin. Like, if you go to London though, we've got goo- we've got DeepMind-

    5. DK

      Yeah.

    6. HS

      We have Facebook. We have Cambridge. We have Oxford. I mean, we have some amazing educational analogs. Like, we're better for AI. Why aren't you here?

    7. DK

      Like, yes, on the foundational layer, I completely agree. But I think on the application layer, there's just so much to learn. Like, it's, it's, it's fairly new and we all start from zero. Like, the race is on since 2023. Foundational layer models, that ha- has been going on for, for a while and you need those institutions that, that, that you just mentioned. But on the application layer, where it's like, "How do I integrate with that API foundational layer? How do I do prompt engineering? How do I design such a product?" Like, those things are new, and you gotta upskill and, and, and we are, we're on that race.

    8. HS

      How many machine learning engineers do you have, broad strokes?

    9. DK

      Uh, 15 to 18, yeah.

    10. HS

      15 to 18?

    11. DK

      Yeah.

    12. HS

      And they're in Münich, in Berlin too?

    13. DK

      In Berlin, yeah. Yeah.

    14. HS

      Okay. Where would you most be persuaded to go to? Who's doing a good job selling? Who's doing a bad job selling?

    15. DK

      When I started to actually entertain the thought was when I realized that in Europe we might be at a disadvantage. As an example, we got access to the advanced voice API from OpenAI f- OpenAI last as a continent. Apple AI or Apple Intelligence, it's gonna be launched last in Europe. And trust me, I tried to work around it. It's fucking hard to work around it if you're a European citizen. And let's continue. Close to 100%, if you go on, on the infrastructure part, close to 100% of all GPUs are built in Taiwan. Obviously for geopolitical reasons, everyone is trying to de-risk Taiwan. So where are these factories gonna be built? I don't see them in Europe. I see, for example, the UAE wanting to support with hundreds of billions to create them there. I see the US, like, like literally, uh, the White House probably one and a half, half months ago, they issued like this paper in which they urged the military to find real estate so they can build this stuff. And, and then th- we, we can continue, like, we're gonna need a lot of energy to it as well for those, for those data centers. People are investing to the, in to, into that energy production in the US, um, in, in the Middle East. I don't see that in Europe. And to me it's like, hey, in Europe today, we're not producing the chips, we're not producing the energy, we don't have the foundational layer models. So in that sense, these countries that don't possess those three elements are not really sovereign countries when it comes to AI. And if we believe that AI is truly as revolutionary of a technology like electricity was when it was first introduced, then we're not sovereign as a country in the first place. And so to your question, where would we go, like, first of all, we'd need to go to a country that has, has AI sovereignity.

    16. HS

      Okay, so first off, I actually interviewed Des Traynor quite recently from Intercom at actually at like a internal Intercom event, and he was like... I'm not gonna do his accent 'cause I suck at an Irish accent.

    17. DK

      (laughs) Yeah.

    18. HS

      Uh, but he was like, essentially, for most of Europe, AI should shut off. Like Fa- he's like Facebook, they just won't entertain putting any of their AI products in Europe because of the regulatory, uh, lack of clarity.

    19. DK

      Yeah.

    20. HS

      And so it's like, "Bup, you don't get AI. Sorry. Bad luck."

    21. DK

      Mm-hmm.

    22. HS

      Is that true?

    23. DK

      I mean, I'm not sure if w- if we're not gonna get it. I think we're just gonna get last. It's gonna just, it's just gonna

  11. 29:21 – 34:43

    How Can Europe Strengthen Its AI Position in Chips, Models, and Energy?

    1. DK

      take longer.

    2. HS

      Okay, so it takes longer and it's hard, and then you said there's foundational models, there's obviously chips, and there's energy. Are those three factors?

    3. DK

      Yeah.

    4. HS

      Okay! So if we just go through them, I'm a European, you're a European, you're in Germany, I'm in London.

    5. DK

      Yeah.

    6. HS

      I, I wanna stay a European. So one by one, chips.

    7. DK

      Yeah.

    8. HS

      What can we do that would make you happy and make us not last?

    9. DK

      You know, very interesting example, the German government wanted to fund Intel to put a chip production plant in Germany, I forgot if it's 10 or 20 billion, which is obviously too little but it's a, but it's a great start, and because obviously the problems that Intel is having, they said, "Okay, no, sorry, we won't do it." So someone is offering them 10, 10, 20, 10 to 20, 20 billion and they say, "No, sorry, we won't do it." First of all, Intel is the wrong company to fund. So the first step you asked me is like put it to the right company, put it to Simon, Taiwan Semiconductor, put it to, to, to, to NVIDIA, people that can actually produce GPO, GPUs, and we need to make it fair at least for them to come and what those companies need is obviously, uh, energy, they need, they need the government support, and they need a lot of clean water, and actually we can provide that in Europe. Um-So I think that's one. We need to, we n- we need, we need to bring them here. Electricity obviously, um, or energy is obviously like a, a way bigger challenge, in particular of the shift, um, to, to renewables. To me, it's an unsolved problem. I also see an ethical problem there because should we, like a, from a German perspective, should we go back to nuclear?

    10. HS

      Do you think we should go back to nuclear?

    11. DK

      I'm, I'm personally not, not, not a big fan of nuclear just because I think it puts a lot of responsibility on future generations for problems that we haven't solved yet. That might be the only way. Um-

    12. HS

      Okay. So we've got chips, like invest in the right winner-

    13. DK

      Yeah.

    14. HS

      ... back them, build them here. Energy, hard problem-

    15. DK

      Yeah.

    16. HS

      ... TBD.

    17. DK

      Yeah.

    18. HS

      Happy to have a TBD. We can't have solutions to all of them.

    19. DK

      Yeah.

    20. HS

      And then foundation models.

    21. DK

      That's the hardest.

    22. HS

      Why is foundation models the hardest?

    23. DK

      Because we have a structural issue here, um, and, and that is we have a talent gap. Like how are we going to bring the right level of talent to bui- not, not to build a foundational model, but to build a competitive foundational model? A model that can compete with OpenAI, that can compete with Anthropic, that can compete with, uh, N- NVIDIA, with X, with LLaMA, um. Well, how do we get this kind of talent and, and where is that founder in Europe that is willing to commit a lifetime to build such a company?

    24. HS

      So I, I, I naively posit, well, we have DeepMind-

    25. DK

      Yeah.

    26. HS

      ... and we have actually huge amounts of LLaMA resources-

    27. DK

      Yeah.

    28. HS

      ... in Paris.

    29. DK

      Yeah.

    30. HS

      And incredible AI centers in Paris.

  12. 34:43 – 36:52

    How Does Regulatory Fragmentation Across Countries Impact AI?

    1. DK

    2. HS

      What about the fragmentation of countries and the different regulation provisions that come with each different, uh, jurisdiction?

    3. DK

      Hey, try to start a fintech in the US. You have, uh, what, 52 different states and some of them need a different license as well.

    4. HS

      I agree.

    5. DK

      Yeah.

    6. HS

      And that's why $15 billion is the biggest with Chime and we have a $45 billion company-

    7. DK

      Yeah.

    8. HS

      ... with Revolut.

    9. DK

      You know what I thought is that... So for example, about Germany, many people complain about the, the notarization of everything and then it's, it's like probably, it's probably 20 to 30 hours of extra work for any funding round that you do. And, and I think it's unnecessary and it's, and it's, and it's a burden. But I think, you know, you know, Mark Zuckerberg, you know, would he have said, "Oh no, sorry, I can't build Facebook here because my lawyer has to do 20 to 30 hours of extra work to go to the notary"? Like, fuck no. That guy would have found a way. Like Elon Musk built in, in Germany, the home of car manufacturing, next to Berlin, next to, right next to the capital, right next to the next to Berlin airport, a factory to produce cars. He made that happen. So I'm like okay, if we have truly great entrepreneurs that yes, okay, there are obstacles and yes, okay, maybe the funding one is gone. Then you have re- then you have, then you have regulatory. But truly great founders will just be stoic about it. They'll be like, "Okay, that's the situation, how to work around it, and let's fucking go." And so I think that's what we need. We just need a pragmatism, that we cannot always, always hold back and look, and look for this kind of excuse. Just, just make it happen. Like it's possible. And to your point, Revolut has proven it, and Adyen has proven it, and Spotify has proven it. We're gonna have more companies prove that it is possible. And so I think we should just stop whining and get pragmatic about it. And then look, I don't think European regulation is great. I think it's horrible. I think governments are utterly incompetent. Like that is the case.... but that should not hold us back for building great companies. And we should not even waste, like, one second complaining about it. Hopefully, it's going to get fixed, and maybe we can, we can, we, we can voice our ideas how to fix it. Um, but it should never, it should never hold us back.

    10. HS

      So would you be long or short on Europe moving forwards?

    11. DK

      So I think s- s- status quo, we have to be short. Um, uh, government

  13. 36:52 – 38:39

    If Ambition Is Key, Why Not Expand to the US or UAE?

    1. DK

      is just... Yeah.

    2. HS

      If you were being as ambitious as you could be for your business-

    3. DK

      Yeah.

    4. HS

      ... why are you not in America or the UAE?

    5. DK

      So we are in America. And in fact, America is our largest market from a revenue perspective, and we also invest most of our money in America. And fairly recently, we're also in the UAE, and we recently opened an office there.

    6. HS

      Do they give you credits for opening offices there?

    7. DK

      They certainly gave us, give us some sort of support, yes, but, uh, n- not in, uh, terms of, of, of money, um, more like in terms of visas and stuff like that and corporation. But so, so how I view is when I, when I look at... Okay, so we got US, we got EU- Europe, and then we got, let's say, Middle East. And so this is our portfolio of markets. And then how I look at it is that, okay, we got US, that's the largest software market on the planet, and, and if, if you want to build globally, you got, you got to win the US, and that's kind of where the dollars come from. Then you have the Middle East, in particular UAE and, and Saudi. And it's like these are just very fast-growing economies, very ambitious economies, very little legacy, which is obviously for SaaS, to your point, implementation is so hard usually because a lot of legacy, but they have very little legacy. And, and they, they want to be the best, right? They want to show the world, "We can do something great," and they actually are on a fantastic trajectory. So, so that's kind of the growth market. And then you have Europe, and Europe is just that kind of, like, almost, like, post-competitive, robust market in, in which you can grow, and if you have the US dollars because you're active in the US, you can likely build a great business out of Europe, but it's kind of like that more or less stagnant kind of thing in the middle. So I think you, you need to think portfolio and you got to diversify.

    8. HS

      Do you live in the US now?

    9. DK

      No. Um, I used to live pre-COVID.

  14. 38:39 – 47:37

    Covid Time

    1. DK

    2. HS

      Talk to me about COVID. COVID you have... I mean, dude, you're a restaurant platform, and I know you'll hate that description, but like-

    3. DK

      Yeah. Yeah.

    4. HS

      ... you still have restaurants at the end of the day.

    5. DK

      Yeah. Yeah.

    6. HS

      What happens?

    7. DK

      Mm-hmm. So COVID hit. We had... We were about to turn two years old. Um, we had... Two years old, so probably we've been like, what, like 18 months live. We're probably up to three, 400 million annualized GMV. Growing very fast, having a good time, and then COVID hits. And I was in New York back then. I'm like, "Hey, I'm not going to do COVID in my shoebox apartment in New York. I'm going to go back to Berlin." And there's a direct flight from Newark to, to Berlin. It's a seven-hour flight. I bought that plane. I still remember, I opened my phone, I checked Looka, and I see our GMV. And it's a Sunday. And I, I land Monday morning, 7:00 AM in Berlin on the runway. I put out my phone, I check Looka. Ninety- 98% of our GMV is just gone. And lockdown hits, and, and our business is just gone. It's... It, it vanished overnight. Lockdowns everywhere, a lot of uncertainty in the market. Um, and, and it was... It was a tough time, and I started smoking. Smoked two pack cigarettes a day, got a beard like this, got, got, got gray hair, was... Lost my humor. Was, was, was horrible. Um, was, was tough, and also because it was such a big break from like, "Hey, no, we're j- we're just, we're just, we're just growing, working hard, product-market fit." To, "Okay, like stop." Not only stop, but go to negative, like lose your business essentially.

    8. HS

      How did you communicate that to the team-

    9. DK

      Yeah.

    10. HS

      ... and to investors?

    11. DK

      I think two things. The first thing I did was (sighs) I Googled best books on crisis management.

    12. HS

      (laughs)

    13. DK

      And I think the best book is, um, uh, it's kind of a biography about Ernest Shackleton, and, and he's that guy who took a boat, and he wanted to be the first one on, on, on, on the South Pole, and he tried three times, failed all the three times, but, but he never lost a man. And it's a fantastic book, and then there's a couple of lessons in there, and one of the lessons is like never lose your humor. Like if you lose h- your humor, you lose your sanity, and you're going to lose your, your, your, your clear thinking. Um, never, never change your values. Like, like, stay, stay true to them. And so I think that was really the, the, the first step, going to the team and say, "Hey guys, look, we don't appreciate the situation, but that's how it... how it is, and we got to make the best out of it. And, um, and you know, we hope everyone has sufficient toilet paper." And because that was... that was a time back then. And then I think me- the second thing is, you know, you know, stay true to your values. And at that point, one of our values really got really part of our culture, and that is always play offense. And so I think, you know, when COVID hit, and when you lose 98% of your GMV, at, at... And it's the very first week of, of it, you know, no one knows what's happening. Like are we all going to die? Is it going to be gone in two weeks? No one knows. It's very easy to play defense and say, "Okay, like look, we're going to hibernate now. We're going to do mass layoffs. We're going to preserve runway. We're going, we're going to preserve cash. We wait until it's all over, and then, then we pump it up again." So, so, so that is like the move... That's a defensive move. And we just say, "No," like, "what do we have to do to play offense?" And that was really formative experience for our company, because there was always one country, uh, in Europe, there was always one state in the US which was not in lockdown. Our, our Spain team moved, moved to Florida, because you can do sales speaking Spanish in Miami. Probably better actually than, than, than English. And Miami became our best city launch to that date. And our France team, they moved first to California and then to Texas, and the German team I think went also to Texas, and then-... and we always had to leave Schengen for 14 days, so we always went to Dominican Republic, waited 14 days and then we entered the US. Um, and so we- we've been growing through- and, and, and we changed in our go-to market from, like, a field sales approach, where on a product optimized for field sales approach, to a product optimized for, like, a tele-sales approach, and then to a product led growth approach. And so we just said, "Okay, like, look, let's be stoic about the situation. That's how it is, and let's just change." And so we didn't do a layoffs. Um, we burned a bunch of cash in that period, 100% sure, um, but hey, you know, it, we stayed true to our values. We, we, we, we, we kept playing offense, we kept our human all positivity and then once the curtain of COVID lifted, like, we were just ahead of the competition, who, who may have chosen to hibernate.

    14. HS

      When was your most recent round?

    15. DK

      Was beginning of 2023.

    16. HS

      Got you. What price was the round?

    17. DK

      1.1 billion.

    18. HS

      1.1 billion. Do you feel that is high now? I think we have a generation of companies that are growing into valuations.

    19. DK

      100%.

    20. HS

      Do you regret raising it, uh, $1.1 billion valuation?

    21. DK

      You know, there's elements of it that I regret, elements of it that I don't regret. I think what I don't regret is that, you know, someone once, once told me if, if you can take the cash, and if your father is not a billionaire, then you should take it. And why is because more cash increase your probability of, of like an outcome, of you achieving, uh, of you achieving your mission and everything, every decision that you do has to be to increase that.

    22. HS

      But do you buy that?

    23. DK

      Yeah.

    24. HS

      I don't buy that. I think it increases distractions.

    25. DK

      Mm-hmm.

    26. HS

      It reduces capital efficiency. Uh, it gives you more runway, which does not increase execution speed.

    27. DK

      So here, here we come to the downsides, right? So I think everything you say is true. Um, I think however we also outlived something like COVID because we, we were doing this strategy. We also kind of came... I don't wanna say smooth sailing. Like, we had our tough patches certainly during the high interest rate area, but also through debt we came because we just took the money. Now, the real downside to me is that it's that one billion mark, because once we become unicorn and people think, "Oh, we've made it," like, like, "We're done. We, we, we're, we're great now," kind of people get less hungry. And then you get applications of people who look for a safe space to work. It's like, "No, fuck it. We're not safe." Like, "We're at high risk. We're gonna work our butts off. We have to take risks. It's gonna be chaos and you're gonna need to thrive on it, and it's gonna be a lot of change." We're not a little changed company now just because we cross... And also, like who the fuck wants to be a rainbow-colored pony?

    28. HS

      (laughs)

    29. DK

      Like, I don't want to be a unicorn. And so just, just that cultural component is... And I remember we, uh, I made the mistake, and we did a unicorn party. I wish we would've never done it. It's like, "Oh, ............................ unicorn started, so, so f-" And then we had, like, flags. And then I realized what it changed in people, and then like two weeks later I told everyone, "Remove those flags in all of our offices and trash 'em." Everyone was like, "Why that? Why this?" Like, "Hey," like, "we're not done." So I think that's definitely what I regret.

    30. HS

      Why did you let that party happen? Like, straightaway there I'm like, oh. Like, you know, we closed the 400 million fund-

  15. 47:37 – 48:53

    On Work-Life Balance

    1. DK

      earlier.

    2. HS

      People need purpose. I totally agree with that.

    3. DK

      100%.

    4. HS

      Do you believe at all in work-life balance?

    5. DK

      I think it's bullshit. Um, you know, what I always tell our people is like, like, like, "We have to be 100% present and 100% intense. So when you're at work, I expect you to be 100% here. Like, not on Instagram, not with your head somewhere else, but like 100% present with all of your intensity. But when you're at home, I expect the same that you give 100% at work, to be 100% at home." And, and I think that's, that's way more important, because, like, work-life balance is like, "Yeah, you know, I'm gonna work every day until, like, you know, 7:00 PM."

    6. HS

      But I would push that. I would say, "Even at home, I don't want you to be 100% at home." (laughs) I'm really sorry.

    7. DK

      Yeah.

    8. HS

      The, by nature of, like, our business, if we are gonna be the best fund in the world-

    9. DK

      Yeah.

    10. HS

      ... we are gonna have to do extraordinary things.

    11. DK

      Yeah, 100%. And, and, and look, I think there's always phases where that's gonna be expected from you. And yet you might get that Sunday morning call-... like that, and you, you gotta be ready for it. But if that Sunday morning call doesn't happen, like, be present because otherwise, you're gonna burn out and your partner's gonna complain-

    12. HS

      100%.

    13. DK

      ... and leave. Yeah.

    14. HS

      100%. Totally agree with that.

    15. DK

      Yeah. Yeah, like, like you need to be ready for war any given moment in time.

    16. HS

      Totally agree with that.

    17. DK

      That's a given.

    18. HS

      How much cash have you raised so far?

    19. DK

      Uh, 330

  16. 48:53 – 56:41

    Lesson on Fundraising

    1. DK

      million.

    2. HS

      What's your single biggest lesson on fundraising?

    3. DK

      You know, fundraising is several step process, and step one is you need to bring a fantastic group of people together. Step two is you need to work on a, on a mission and on a strategy that's, that, that makes sense and those people have to create that. Step three is you gotta work your butt off. Step four is you're gonna look how well you're doing and you put this on an ugly slide that's i- idly just a couple of graphs. And step five is you just hand those slides out. And that's the fundraising process. So I, I... Yes, there's, there's many of things within the fundraising process that, that you can do good and many things that you can do bad. But fundamentally, if the graph is going up and to the right, you're gonna raise money. And if not, it's gonna be fucking hard. So the best way to fundraise, just make your company do well. And then I think, you know, for like, for like first time founders or like when you raise, when you first, when you raise the first couple of rounds is like, don't beg for money. Like, the investor's job is to invest, not to not invest. Like understand... And, and also, what is the psyche of a VC? The psyche of a VC is like, hey, I have a portfolio, and everyone knows, 90% of my portfolio are gonna be write-offs. They- they're gonna fail. That's normal. That's accepted in the venture world. What is not accepted is if you meet the entrepreneur and you decide not to invest, your negative portfolio, if, if you, you miss the deal. So naturally, the, the, the investor is to and has to be, to a certain extent, far more driven. So, you know, in the end of the day, cash is a commodity. The dollar is green. But your company is singular. And yes, you're also reliant on the investor, but like, don't beg. Like, it has to be a conversation eyesight. Don't be arrogant either, but like, meet on eyesight. You have an asset, and other person eventually wants, wants, wants that asset. And, and, and by the way, you cannot choose any investor. You probably can only choose one or two, maybe three per round.

    4. HS

      To your point on upside, the thing founders don't often articulate best is like, how is this a $10 billion company? And I think we're almost trained not to in Europe because it's arrogant or it's too idealistic or it's... But I have to see a billion-dollar revenue business.

    5. DK

      You know, I love that you said idealistic because it ties us back to a previous point. It's like, why Berlin? I think Berlin is a very idealistic city, maybe the most idealistic city among capitals in-

    6. HS

      Wow.

    7. DK

      ... in, uh, in, in, in Europe, and I think a founder, to a certain extent, has to be idealistic and has to put a great vision, because in the end of the day, like what I... Like our vision, for example, is it's enable a sustainable food system, and that's very important for the globe from, from, for many respects. But I'm like, okay, like, you know, we went through COVID. That has been tough, and then there's the high interest rate period, and then there's late, that AI pivot, and, and every company has to go through such things. Yesterday, I watched an early interview of Steve Jobs, and he said they were at 90 days of money left. And so like, what is gonna make you stand in front of your team, not in front of the investor, but in front of your team, and authentically say, "Hey, guys, we need to work extra hard now because it's really, really important that we achieve our mission"? And, and, and if, if that is not like a, a compelling vision, then you're gonna have a hard time doing that authentically. And so I think when you have that, and this is what you have to voice to the investor, and then I, I don't think you need to put a number behind it. It just needs to sound compelling. It, it needs to sound large, need to sound long term. Long term is not a component, like we spoke about, okay, like Europe and regulation and, and extra capital and stuff like that. Like, we need more long term commitment as well. Like, we need founders to say, "I'm gonna vis- invest 15 to 20 years. I'm gonna commit this now, substantial part of my life, the best years of my life, to make this happen. I'm not gonna give the fuck up. I'm not gonna do... I'm, I'm gonna win, and I'm gonna, I'm gonna push through." And I think the investor needs to know this. And you need to... I think this you need to ar- articulate. Like, what is this? Is this a company that we're gonna, you know, try to increase revenues fast and sell it? Is, is it, is it at IPO? You know, leave the company and make holidays on the Bahamas? Or is it like, are we gonna build a generational company? And by the way, that's what we need in Europe, generational companies.

    8. HS

      There's a s- very sad moment for me in a Jensen Huang interview where they said, you know, "Would you do this again?"

    9. DK

      Yeah.

    10. HS

      And he goes, "No. If I knew how hard it was, I wouldn't do it again." Respectfully, Daniel, you know, you're in a great place now with ............................ But with COVID, with the transition, with everything that we've been through, would you do it again?

    11. DK

      Um, I ask myself this question often. Um, there were, there were years when I said, would, I would say, "Fuck no."

    12. HS

      (laughs)

    13. DK

      Um, and there are y- there are years where I said like, "Yeah, and, and, and, and with pleasure." And, and I think that you gotta be realistic about like, it's gonna be tough. And every cell in your body has to tell you, "Start and build that company." And if, if, if one cell in your mind and your body is telling you, "Yeah, maybe," then don't do it. Like, don't. But if every sen- cell tell, tells you to do it, um, then you gotta do it. And I think, you know, people say, you know, some people jump in the pool and they get out and they're still dry. And I think a real founder is, is wet before they even jump into the pool. And like, you gotta sweat. It got... It has to be part of you, um, and, and then I think you might, you might do it again.

    14. HS

      I remember when I was once fundraising, it was, it was tough time fundraising.

    15. DK

      Mm-hmm.

    16. HS

      Um, uh, it's a couple of years ago, and my friend who started Calm, Alex Will, said to me, "Harry, you don't get it, dude. Like, entrepreneurship is about getting punched in the face time and time again-"

    17. DK

      Mm-hmm.

    18. HS

      "... and going, 'Yeah, give me more.'" And you have to enjoy that. Not like be okay with it. You have to, like, actually seek it out-

    19. DK

      Yeah.

    20. HS

      ... like almost this masochism of like, "Yeah, I'll take it."

    21. DK

      Uh, uh, 100%, and I, I think there's, there's, there's many parallels. So when I was younger, it was, uh... I grew up in, in, in the Alps and I was a ski instructor because that was just like the kind of the job that like 16-year-olds would, would do where, where, where I grew up.And, um, it's like- what I saw very early on is, like, the kids that fell all the time, they- they learn the fastest. And like, you know, everyone falls for the first time. That- that's just part of the process. But then it's like, the ones that went around, "Oh, and I don't want to fall again," and blah, blah, they're just not gonna learn. But if you're willing, like, to fall again and again, like, those are the kids that- that make it and that will become great- great- great racers.

    22. HS

      Can I ask a final one? Which is, you said you raised 300 and- 330? Whatever it is.

    23. DK

      Yeah, yeah, yeah.

    24. HS

      What did you spend on that you wish you hadn't spent on first?

    25. DK

      We had a team with- with really great people in there, so it's not the people's fault. Um, it's more like my fault, and it was a special projects team. And it's kind of like, okay, like, here's our core business and here's a special projects team to figure out great things, and once they're great, bring it back to the core and then the core's going to scale it. And I think, today I'm convinced that- that- that's the wrong way to think about company building. Um, actually, this AI would be another example. Like, we could give it back to a special projects, kind of Skunkworks kind of team, you figure out AI and then- then you give it back. So that doesn't work, and the reason is that if you're- if the core of your company doesn't have the skills, the capabilities to innovate, then you have a problem in the first place. Like, your core needs to be able to innovate, you know, to get it there. So the Skunkworks team might- might- might just be an excuse to not being able to do it in your core, that's one too. Let's- let's say your special projects team figures something out and it's fantastic, and then you give it to other people and they say, "Oh, what they did is fantastic. Now can you please like productize it and roll with..." No, like I want that person who figures it out, I want to allow this person to own it, to be the responsible individual for it, to- to scale it, to build it, to get it, with- with that. That is much more motivating, right? And so, um, yeah, I think that understanding that likely if you have a special projects team that's something's not well in your core, um, we should not have spent

  17. 56:41 – 58:45

    Why Spain is the Worst Market for Choco

    1. DK

      money on that.

    2. HS

      Which is the worst market for you?

    3. DK

      Spain.

    4. HS

      Why?

    5. DK

      Highly fragmented, like next level fragmented. To- to give you an example, the average- (clears throat) the average restaurant in the US, in Germany orders from- from four and a half different suppliers. Yeah, so say one for fish and one for meat and one for drinks and so on. France we had nine, much more specialized team of focus on food and so on and so forth. In Spain we had 18, and likely you need a different product for that, right? Because- so the distributors, they're smaller, so does it justify your sales cycle? You need probably more like a self onboarding kind of product, like growth kind of thing. Um, so that's why I think the market is, uh, is- is- is challenging. On the other hand, Spain has many advantages. It's one of the very few economies in Europe that are actually properly growing in GDP from a smaller base of, uh- uh, obviously, um, access to talent is- is okay. Um, people want to move to Barcelona. It's- it's an easy sell.

    6. HS

      What did you not... A final one before we do a quick fire. What did you not spend on that you wish you had spent on?

    7. DK

      At the very... So we- we were lucky, we found product market fit for our first product very early on in our... Probably even actually before we started coding to be fair, um, because we just did a trillion iterations just on design, always with user feedback. So we- we- we product marketed early on. And then, um, and then we went out to the market and it was growing and then, you know, we launched like Berlin and then Germany and then France and the US and- and so on and so forth, and city by city. And it was working well. And I think, you know, something working is actually very rare, and so once you have something that's working, like you got to go all in. Like I wish we would have, you know, spent double the amount of money, have- have tripled the amount of burn probably in the first 24 months of our company and just- and just go big because...

    8. HS

      What- what would that have allowed? A land grab on distributors?

    9. DK

      For sure a land grab. Um, I think more importantly, it would just have created a larger brand within

  18. 58:45 – 1:01:54

    Raise to Compete or Avoid the Frenzy

    1. DK

      the industry.

    2. HS

      When everyone is raising a lot of money around you, the space got very hot very quickly.

    3. DK

      Yeah.

    4. HS

      Some people say, "You've got to raise too. You've got to compete." Others say, "The money will dry up. What matters is actually not getting involved in the frenzy and playing your own game." Which one makes sense?

    5. DK

      You know, I'm- I'm not sure if it's an- if it's an either or. Like I- and I'm not saying you should like just triple your burn. I'm not saying you should triple your burn when something is working really well, right? Within- within the constraints of- of good economics and- and capital efficiency. But once you figure that complete thing out, not only the product market fit, but also like a go to market of good economics, then just go as fast as possible because that's when you actually will attract copycats. It needs to be the- the- the complete thing. And then I think we can also answer this question in isolation, like we also need to look at the competitor situation. And the competitive situation-

    6. HS

      Do you think you had good competitors?

    7. DK

      Yeah, I think- I think we had good competitors. I think we still have good competitors.

    8. HS

      Which competitor did you most respect?

    9. DK

      So early on, certainly a- a company, it was called Reki.

    10. HS

      Mm-hmm.

    11. DK

      And- and I just think they had a fantastic product for restaurants, and probably still have. Like we were directly competing in Paris, in Chicago, New York, Berlin, everywhere. It was like a multi-frontier war if you want. So I think we just sat down and- and- and- and- and- and analyzed the strengths and weaknesses and- and just- just made a plan on how- how- how to win and execute on it. And I- I think actually it's fantastic that we had a good compe- and actually a competitor that at that point was probably already 10 times larger than us. We had maybe raised two million, they had raised 20 million. Um, that was very fertile ground for us to start.

    12. HS

      I always say be grateful for your competitors. (laughs)

    13. DK

      100%. 100%. And it's- it's, um, it- it makes you better but it's also like it's the simplest motivational tool. You know, when you're second, when you're- when you're third, like what's the goal? Like here's the number one, the goal is get a bit above the number one. It's as simple as that. Now once you're the number one, okay, what's the goal now? Staying number one is a bit defensive, right?... so like kind of like, you know you need to go from like winning to like dominating, how do you define it? And suddenly, things become a bit unclear. Like, you're essentially always pushing forward and excellent stuff like that. But-

    14. HS

      You also need to create a common enemy.

    15. DK

      Yes. To a certain extent, right? Because I think a common enemy, an extrinsic motivation, if you want so, works. But what, what I still believe what's, what's more, what works better is an intrinsic motivation. It's like, like, "We are here to achieve our mission. That's why we're here." You know, and by the way, lions don't lose sleep over sheep. Like, we need to focus on what we are doing and not focus on winning against competition but like at- at achieving our mission and doing it as fast as possible. And so I think that intrinsic motivation still has to be number one. And I always, you know, I'd rather think like the kind of the extrinsic one, like the competition, the market. That, that has to be like a far two.

    16. HS

      Lions don't lose sleep over sheep.

    17. DK

      Yeah.

    18. HS

      I love that.

    19. DK

      Yeah.

    20. HS

      I haven't heard that before.

  19. 1:01:54 – 1:15:03

    Quick-Fire Round

    1. HS

      Uh, listen, I love doing this. I want to do a quick-fire with you.

    2. DK

      Yeah.

    3. HS

      So I say a short statement and you give me your immediate thoughts.

    4. DK

      Okay.

    5. HS

      So what do you believe that most around you disbelieve?

    6. DK

      I do believe only AI-first companies will win. And by AI-first, I mean that a majority of your total revenue or of your new revenue, which will lead to a majority of total revenue, has to come from an AI product. Why? Because you're gonna be more capital-efficient, it can provide more- more- more- more- more user values, just fast adoption curve. On top of that, kind of using AI internally, so revenue's kind of more ............................ product, but using it internally to get more productive, to automate things. It's just gonna be the required status quo. A company with a social media account is not a so- won't become a social media company, it's just a requirement of an Instagram account just t- to hold the status quo of a company. And we, companies that automate ............................. with AI, we don't call them AI-enabled, we'll just call them a company in the future.

    7. HS

      I know your revenue growth has scaled immensely.

    8. DK

      Yeah.

    9. HS

      If I'm allowed to ask, what's your revenue?

    10. DK

      We're in a very competitive situation, so I don't- don't think we like talking about it. But, um, I can tell you our GMV since this AI transition grew like tremendously. I think it took us, with kind of like the two years of COVID break, it took us probably five years to get to one and a half billion-ish. And it took us then less than a y- yet another billion.

    11. HS

      So you're at two and a half billion in GMV?

    12. DK

      Mm-hmm.

    13. HS

      What's the closest competitor in terms of size?

    14. DK

      Frankly, I don't know. Frankly, I don't know. I assume half, maybe less than half.

    15. HS

      That is insane though, when you look back to that flight, and you look back to the thinking, reminding you of that look of, "Whew," yeah, yeah.

    16. DK

      Yeah.

    17. HS

      But like 300 and it's like, that was, that was the great, and then it got shot. And now to be at two and a half? Whew.

    18. DK

      Yeah, yeah, it's, uh, it's been wild.

    19. HS

      (laughs) It's good to have the really shit times 'cause it keeps you level on what- what truly shit can be.

    20. DK

      Yeah. And I- I- and I don't wanna miss it. Like, you know, I- I always tell- tell our people kind of like, "Hey, you know, r- running a company is like, you know, we need to cross a jungle, okay? And then- then we're here, and then there's the jungle, we need to go to the other side, and we have different jobs. And some people need to cook, and some people need to carry, and some people ............................ build tents." And sometimes, you know, it's fantastic, you're like on the top of a mountain and the sun is rising, the birds are chirping, it's just beautiful. And sometimes, it's just raining for three days, it's cold, we are all wet, we're hungry, don't find food. And right then, middle of the night, we have to cross a river full of piranhas and crocodiles. And we're like, "Fuck." But, when we're all old and gray and we look back, the cooler story is actually like, "Hey mate, remember when we used to have to cross that river in the middle of the night full of piranhas and crocodiles?"

    21. HS

      (laughs)

    22. DK

      Like- like that's actually what makes it, what makes a great experience.

    23. HS

      I- I- I totally agree with you.

    24. DK

      Yeah.

    25. HS

      You can be CEO of another company for a day, which company would you like to be CEO of?

    26. DK

      One on the inspiring side, and one off the, "Okay, what the heck is actually going on?" side. And I think, I mean, Pixar, um, just because of like the- the talent density, and kind of like this- this- this creative work, which is very far from what we do, we're like more like B2B and, you know, hardcore efficiency. Like, I think that'll be a fantastic environment to experience that. I think kind of on the, "What the fuck is going on?" side is like the German rail system. Like- like, Deutsche Bahn is like, "What the heck are those guys doing every single day, 300,000 people?"

    27. HS

      Why, what the heck are they doing?

    28. DK

      Uh, (laughs) couple of examples. So, last year, 2023, there was like this big bonus. Okay, like bonus payments to the executives, but to every single employee at Deutsche Bahn if they were to get more on time, more punctual. And so they did not only, like, miss that goal of getting more punctual, like, not like only got half of more punctual. No, no, they went negative. 9% negative. 100% of the bonus was paid out. So like, okay, like, "What the heck is going on in that..." And it's getting worse by the day. Like, "What are they doing?" Like, French, and I believe even Italian railway system, doesn't let German trains anymore enter their countries because they're just too fucking late all the time. Like, "What's up? Sorry guys." Eh, e- and, so how dysfunctional must a company be? I'll give you an example.

    29. HS

      Also, what precedent do you set that you miss the goal-

    30. DK

      Yeah.

Episode duration: 1:17:00

Install uListen for AI-powered chat & search across the full episode — Get Full Transcript

Transcript of episode pGU78Q5tJgA

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome