20VC’s Codorniou Backs Young European Tech to Outpace Global Rivals

20VC合伙人Codorniou看好欧洲年轻科技力量超越全球对手 欧洲正迎来科技领域的向上突破，这主要受ARR快速增长及海外人才回流推动，但与美国市场相比，本土企业采纳仍是一大关键瓶颈。 Julien Codorniou阐述了欧洲科技行业正在经历的“向上突破”，这一趋势由回流的精英人才、快速扩张的AI企业以及面向年轻创始人的Project Europe等结构性倡议共同推动。以Kelkoo的人才外溢网络以及Pigment在企业级市场的迅猛增长为例，他强调欧洲正形成日益紧密的自增强型科技生态。Codorniou还指出超高效“单人独角兽”的兴起、面向一线员工且潜力巨大的SaaS市场，以及诸如French Tech Visa等前瞻性监管政策如何赋予欧洲相较于美英两国的独特竞争优势。 欧洲科技领域正经历一场“向上突破”，其特征是高素质且经验丰富的创始人纷纷入局种子轮，初创企业以空前速度扩张。 (0:01) 在强大的人才回流网络与成熟融资价值链的支持下，DeepL、ElevenLabs、Lovable和Nabla等欧洲初创企业在短短12至24个月内，ARR就从零增长到1000万至1500万美元。 (0:01) Project Europe的联合创立旨在为25岁以下的创始人提供早期资金，并让他们能够对接150位成功的科技创业者，在不同地区树立本土榜样。 (0:20) Kelkoo人才外溢网络的长期影响表明，单个重大的欧洲成功案例能够在数十年间为整个初创生态播下种子并持续赋能。 (4:40) 法国拥有来自顶尖学府无可匹敌的工程人才，但历史上在分发与营销方面举步维艰，这一差距正在现代软件分发模式下逐步缩小。 (6:41)

Julien Codorniou argues that European tech has entered an "upward breakout" driven by elite technical talent, returning diaspora, and hyper-lean AI architectures. However, he contends that capitalizing fully on this shift requires overcoming risk-averse enterprise procurement habits and fragmented cross-border equity regulations. European Tech Breakout and Generational Diasporas (0:01) Codorniou observes that European startups are scaling at unprecedented speeds, citing companies like DeepL, ElevenLabs, Lovable, and Nabla reaching $10 million to $15 million in ARR within 12 to 24 months. He attributes this momentum to an increasingly mature value chain and an experienced diaspora returning from companies like Stripe and OpenAI to launch ventures such as Dust in France. He points to historical precedent: the alumni of early French comparison shopping engine Kelkoo later seeded the ecosystem by funding or founding major companies including Criteo, Pigment, and Teads. To expand the talent pipeline, Codorniou co-founded Project Europe, which invests €200,000 for 6.66% equity in founders under 25 and connects them with 150 European tech mentors. While France produces elite technical talent from institutions like École Normale Supérieure and Polytechnique, Codorniou emphasizes that the region still needs better localized storytelling and visible role models beyond the English-speaking sphere. Lean AI Architectures and Defensible Communities (10:45) Characterizing current software dynamics as the "Flappy Bird era of SaaS," Codorniou notes that AI-native startups with only 25 people can rapidly reach $40 million to $100 million in ARR, citing Lovable's rapid growth to $50 million ARR. However, because basic software wrappers face instant replication, he argues that defensibility hinges on cultivating high-trust brands and devoted user communities. He also aligns with predictions of "one-employee unicorns," forecasting that profitable global software businesses run by one to three people will migrate from gaming into B2B SaaS. Rejecting dystopian predictions that AI will replace developers entirely, Codorniou compares autonomous agents to assistive copilots—drawing a Star Wars analogy where human operators retain ultimate decision-making authority. He points to Cognition's internal workflow as an operational template, where 15 human engineers orchestrate roughly 15 Devin AI agents each to achieve a fifteen-fold productivity gain. Frontline Workers as Software's Untapped Market (14:13) Disputing Microsoft CEO Satya Nadella's assertion that SaaS is dead, Codorniou argues that software adoption is entering a major expansion wave focused on frontline workers. Drawing on his experience launching Workplace at Facebook—where adoption was driven by deskless employers like Walmart, Starbucks, Club Med, and Danone rather than tech firms—he identifies a global market of 2 to 2.5 billion blue-collar workers who lack corporate email addresses or desks. Codorniou asserts that generative and mobile-first AI uniquely addresses this segment by bypassing expensive per-seat software models. This allows enterprises to deploy automated workflows for shift scheduling, inventory tracking, and paid time off management directly to frontline employees. Enterprise Procurement Hesitancy and Regulatory Friction (15:38) Codorniou identifies corporate conservatism as a major structural hurdle for European software. While US Fortune 500 enterprises treat software purchases as investments aimed at generating returns, European corporations often treat software strictly as a cost to be minimized, deferring to legacy vendors under the logic that "you don't get fired for going IBM." To illustrate, he notes that French planning software startup Pigment closed a global contract with Coca-Cola's US headquarters in six months, but spent over two years attempting to secure a contract with a corporate buyer in France. In contrast, he notes that French SMBs adopt local software much faster due to operational pragmatism. On policy, Codorniou criticizes Europe's fragmented employee stock option frameworks as an administrative nightmare that hinders cross-border talent compensation. Conversely, he highlights the French Tech Visa as an effective immigration pathway that provides continental Europe a recruitment edge over the more restrictive visa environments in the US and UK. While Codorniou sees European tech benefiting from superior technical talent, capital access, and hyper-lean AI models, the ecosystem's ultimate trajectory remains constrained by enterprise procurement inertia and fragmented intra-European compensation rules.

20VC普通合伙人Julien Codorniou指出，欧洲科技生态正迎来结构性的向上突破，人才回流与AI原生范式正驱动初创企业以创纪录的速度实现规模化；然而，若要将这一势头转化为长久的全球竞争力，欧洲必须克服本土大型企业保守的采购倾向以及各成员国碎片化的期权监管障碍。 欧洲科技的向上突破与人才网络飞轮 (0:01) Julien Codorniou认为欧洲科技行业正在进入强劲的向上突破期，其根本支撑在于创始人质量与生态成熟度的显著提升。当前进入种子轮与种子前轮的创业者很多曾在Stripe或Facebook等公司历练多年，或是拥有多次创业经验的连续创业者；与此同时，由OpenAI和Stripe等顶尖机构离职员工回国创立Dust为代表的人才回流潮，为本地生态注入了关键动能。在充沛资本与涵盖天使投资人、孵化器至成长期基金的本地价值链支持下，DeepL、ElevenLabs、Lovable以及Nabla等欧洲企业，仅用12至24个月就将年度经常性收入从零提升至1000万至1500万美元。包括Sequoia、IVP和NEA在内的美国顶级风险投资机构在欧洲设立实体办公室，进一步确认了该市场的战略价值。 这种爆发式增长源自欧洲历经数十年沉淀的校友网络飞轮效应。以1999年前后成立、在互联网泡沫后被雅虎以5亿美元收购的法国比价引擎Kelkoo为例，其联合创始人Dominique Vidal加入Index Ventures后投资了Criteo并促成其上市，而Criteo的创始人后来创立了目前高速增长的Pigment；另一位联合创始人Pierre Chappaz则创立了Teads并成功高价出售。这种成功创业者转型为投资人、启动新企业并反哺后辈的循环机制已在欧洲确立。为加速这一传承，Codorniou联合创办了Project Europe，为25岁以下的年轻创业者提供20万欧元（换取6.66%股权），并汇聚150位本土知名企业家担任榜样导师，致力于消除非英语地区由于榜样叙事不足而形成的创业鸿沟。 从单人独角兽到AI时代的软件护城河 (6:41) 在区域禀赋方面，Codorniou强调法国在培养前0.1%的顶尖技术人才上极具优势，但历史上长期面临工程研发强大而分销和营销相对较弱的短板。随着产品驱动增长与AI工具大幅简化分销流程，优秀的软件设计已成为最关键的起点。这一转变契合了Sam Altman提出的“单人独角兽”构想，即仅由1至3人组成的高利润轻量级团队正在从游戏领域扩展到SaaS行业。这类高效团队也促使欧洲风投生态从传统银行家主导转向更具实操共情能力的运营型投资人。 然而，当前软件生态也进入了高增长与高仿冒并存的“Flappy Bird时代”，25人左右的AI原生团队虽能迅速达到4000万甚至1亿美元的年经常性收入（如Lovable迅速突破5000万美元），却随时面临被迅速克隆的风险。Codorniou认为，在这一环境下构筑持久壁垒的核心在于建立深入人心的品牌信任度与高粘性的极客社群。关于AI对开发者的影响，他反对机器完全替代人类的机器主宰论，更倾向于将AI视作《星球大战》中协助人类消除盲区、提供决策辅助的机器人搭档，强调最终由人类掌握决策权；如Cognition的15名工程师每人实时调用约15个Devin智能体协同作业，展现了AI将人类生产力放大15倍的赋能形态。 一线SaaS蓝海与欧洲本土企业采购滞后 (14:13) 针对微软Satya Nadella关于“SaaS已死”的观点，Codorniou提出了截然相反的判断，认为服务一线工人的SaaS革命才刚刚起步。他在Facebook负责企业协作平台Workplace时发现，其最早且规模最大的客户并非科技公司，而是沃尔玛、星巴克、喜力、地中海俱乐部和达能等拥有庞大无工位员工的传统企业。全球有20亿至2.5亿没有公司邮箱和固定工位的蓝领及一线员工，代表着长期被传统企业软件忽视的蓝海市场；AI的介入降低了昂贵的单席位软件采购门槛，使其能够以移动端优先的形式高效解决排班、库存管理与请假追踪等刚性需求。 与此形成鲜明对比的是欧洲本土大型企业在软件采购上的保守文化。美国财富500强企业普遍将软件视为带来回报的投资，而欧洲大型企业往往将其视作一项需要极力压减的成本，其决策者受制于“选IBM绝不会被解雇”的避险心态，对采购本地初创软件持迟疑态度。法国规划软件Pigment仅用6个月便与美国可口可乐全球总部敲定合同，但耗时两年多仍未能打入距离其总部仅十分钟车程的法国本地大型企业。不过，法国中小企业展现了高度的务实性，更关注实际效用而非品牌标签，因而积极采购本地软件。Codorniou主张，要打破大企业的采购壁垒，更需要整个生态在文化层面对勇于采购本土软件的企业决策者予以认可和鼓励。 长期试错文化与欧洲监管的机遇挑战 (18:16) 从组织管理的角度，Codorniou借由Meta的发展历程论证了创始人控制权与长期主义耐心的价值。Meta在2012年启动移动端转型时曾遭受资本市场的大规模做空与质疑，内部团队经过数年扎实推进，至2014至2015年才彻底确立移动端主导地位；同样，尽管Meta在元宇宙上的巨额资本支出在直观收益上并不理想，但其积累的算力与硬件基础设施却直接成为其当下布局尖端AI的核心支撑。Codorniou回顾了自己在2012年主导在Facebook推出真钱博彩游戏、首月仅获约200欧元收入便被果断叫停的经历，说明允许快速试错且包容公开失败的心理安全感是企业保持创新的基石。 在政策与区域生态层面，Codorniou呼吁改革欧洲高度碎片化且复杂的员工股票期权法规，指出法德等国不统一的监管体系是初创公司有效绑定员工与股东利益的主要障碍。与此同时，他高度评价法国科技签证政策为引进国际技术人才提供了极佳通路，在英美移民门槛收紧的背景下构成了欧洲的独特优势。在生态布局上，他将法国南部的蒙彼利埃列为2025年被严重低估的创业热土，并强调Xavier Niel通过投资以及创办Station F和编程学校所建立的基础设施，是法国过去15年乃至当前繁荣生态的根本基石。 Codorniou的观察描绘了欧洲科技在工程研发水准、人才回流与AI原生创业浪潮驱动下的爆发潜力，但也明确指出了其面临的结构性挑战：本土大型企业根深蒂固的避险采购文化，以及碎片化的泛欧员工股权激励制度，依然制约着本地初创企业的发展步伐。欧洲能否在接下来的技术周期中诞生具备全球统治力的科技巨头，不仅取决于技术团队的效率与社区粘性，更取决于欧洲内部能否在企业采购与制度协同上实现真正的开放与融合。

[6:41] Julien Codorniou: When you see a French entrepreneur pitching, usually you don't tell yourself you've met [Cheryl] Sandberg or Marc Benioff [10:45] Julien Codorniou: I think of AI as a robot on your side that can help you to remove blind spots, to make better, faster decisions, but that cannot replace you... It's Luke Skywalker who makes the decisions and who ends up shooting the Death Star. [14:13] Satya Nadella: SaaS is dead [15:38] Julien Codorniou: you don't get fired for going IBM [18:16] Julien Codorniou: license to try [22:20] Julien Codorniou: Paris for vacation, London for work.

[0:01] Julian, thank you so much for being [0:02] here. My pleasure. Thanks for having me. [0:04] Um, you need little introduction, but [0:07] Julian recently joined 20VC as a general [0:10] partner and he has a wealth of [0:12] experience in tech, both as an investor [0:14] and of course also as an operator. Um, [0:17] you were at Microsoft, you were at [0:19] Facebook, um, you have been an investor [0:21] in some of the most exciting French AI [0:24] startups including Knobless. So very [0:26] excited to get your thoughts today on [0:28] the tech ecosystem. Awesome. Um, let's [0:30] dive in. You have recently argued that [0:32] Europe is moving faster than many [0:35] pundits admit. Yeah. Uh, and you know, [0:38] cheerleading aside, I'm curious one [0:41] metric or data point that you look at [0:43] that gives you conviction that Europe [0:45] has generally entered this kind of [0:47] upward breakout. It's not so much one [0:49] metric. It's a series of anecdotes um, [0:51] and things we are starting to see in [0:53] Europe and in France in particular that [0:56] we never saw before. So we just for us [0:58] at 20VC I work with the the seed and [1:00] preed funds. So we try to invest as [1:02] early as possible in amazing funders and [1:05] the caliber and the quality of the [1:06] people we see is just at an all-time [1:08] high. People who've worked at Stripe for [1:10] a few years, people who worked at [1:12] Facebook, people who had one or two [1:13] exceeds before, people who've seen the [1:15] movie doing their second, sometimes [1:17] third company, which is great. But also [1:20] we see you know companies like deeper 11 [1:22] labs lovable Nabla going from zero to 10 [1:25] from 10 to 15 in just 12 months 18 [1:28] months two years we did not have that. [1:30] So that diaspora of talent access to [1:33] capital that is almost unlimited plus [1:35] the experience and the whole value chain [1:37] from angel incubator seed funds to [1:39] growth fund all all in Europe right now. [1:41] It's it's a reason to to be hopeful. [1:43] Truly remarkable. I got to have lunch [1:45] today with about 60 French founders uh [1:48] including some that you have invested in [1:50] and just before this was meeting with [1:51] dust which is uh a couple of uh stripes [1:56] uh are are are in their founder group. [1:58] So really and open AI but that's a very [2:00] good example y one from stripe one from [2:02] openai coming back to France to start a [2:04] business there like it's it's a dream [2:06] come true for an investor. Yeah. Let's [2:08] talk about um some of the the young [2:10] founders. So, Project Europe uh is [2:13] backing founders under 25 uh with I [2:17] think it's 200,000 euro and 6.66% [2:20] equity. I'm curious what problem are you [2:23] trying to solve with that? It's it's [2:25] both for inspiration. We want people to [2:27] uh be exposed to amazing funders which [2:29] is what we do with project Europe. We [2:31] funded it ourselves with our friends [2:33] from point 9 and adjacent but we also [2:35] brought [2:36] 150 entrepreneurs people who really made [2:38] it big in tech in Europe and we connect [2:40] them with students people who are [2:42] between 16 and 25 who do want to start a [2:45] business but they need a role model [2:47] sometimes they don't know where to start [2:49] sometimes they live in countries where [2:50] the ecosystem of angels or VC is not [2:53] very sophisticated so we're just trying [2:55] to fill that gap so it's about the money [2:57] a bit but more importantly the entourage [3:00] the inspir ation that we provide with [3:01] access to these funders and you would [3:03] not believe the quality of the stuff we [3:05] see like we we see people in [3:07] neuroscience, cyber security, robotics, [3:10] AI, of course, kids who are 18 years old [3:12] who want to change the world obsessed by [3:14] the problem they're going after, but [3:16] they just don't know where to find some [3:17] money. So that that's that's why we [3:19] created it's you know it's [3:20] hyperconistent. I did some back channels [3:22] on you and um and the founders that work [3:25] with you say how much they value that [3:26] you're truly an operator, right? that [3:28] the steers you give them are deeply [3:30] rooted in decades of experience in tech. [3:33] So, I love that you are scaling that [3:35] operator view and and bringing along [3:37] that entourage. It's a very nice way to [3:39] say I'm old. [3:41] Decades in a good way. Um, you talked [3:43] about, you know, teens able to do [3:45] amazing things. Certainly, we all know [3:46] Bill Gates and Zuck and actually even um [3:49] Patrick and John. Why do you think there [3:51] are fewer teen founders today? I think [3:54] they they need role models. They need [3:56] inspiration. I think everybody realizes, [3:58] especially with what's happening with AI [4:00] and the whole, you know, new world of [4:02] software we live in, that they can start [4:04] something, they can build something [4:06] amazing from scratch with AI in just a [4:08] few few hours or a few days and make a [4:10] living from it. But I think they just [4:12] need role models and we don't share [4:13] these stories enough. I'm sure many [4:15] people don't know that the the founders [4:17] of Stripe started a company where they [4:18] were, you know, they're very young. Same [4:20] thing for Facebook. So, uh, the [4:22] storytelling in Europe, I would I would [4:24] say in France in particular, needs to be [4:26] pushed, I think, a bit, um, a bit more. [4:28] People don't know these stories or [4:31] sometimes it's in English only. So, if [4:32] you are a kid in Ukraine, maybe you [4:34] don't know about these these guys and [4:35] how they made it and how you also could [4:37] make it. I want to come back to Ukraine [4:40] in a bit. In your 2005 Kelku book, you [4:44] chronicled a very different era of [4:45] French tech. And I'm curious if you were [4:48] to write a bit of an epilogue, maybe a [4:50] couple sentence, one paragraph epilogue [4:52] today, what would you add? [4:55] I would I thought about it. I thought [4:57] about writing uh the sequel to uh to uh [5:00] to the book. And actually, what's [5:01] fascinating is to see what happened to [5:03] the people who built or contributed to [5:06] the the building of Kelku. Kelku was a [5:08] for for people who were there in 1999 or [5:11] 2000, it was a French company, a shopbot [5:13] that got acquired by Yahoo for half a [5:15] billion dollars. It was a very [5:17] profitable company. They acquired a [5:19] bunch of companies across Europe. And [5:20] they it was the first big internet [5:22] acquisition after the the nuclear winter [5:24] of the the internet bubble who exploded [5:27] in 2000. And so I wanted to tell that [5:29] story. If you look what happened to the [5:31] main um people who worked at Kelku, who [5:34] created Kelku, uh you see them in some [5:37] of the most amazing companies that we [5:38] have today. For example, Dominic Videl [5:41] was the co-founder and CFO of Kelku. [5:43] Then he became a VC at index. First [5:45] company he backs at index is CTO. He [5:48] helps CTO to pivot to the advertising [5:50] model. They go public and the founder of [5:52] CTO after KTO becomes the founder of [5:54] Pigmont which is one of the fastest [5:56] growing company we have there and in [5:57] Europe. You see the impact of having one [6:00] success story of having that diaspora [6:02] many many years later. Most of them [6:04] became angels. Some of them started [6:06] something again like Papaz after Kel [6:08] started TIDS that he sold for a fortune [6:10] to I think [6:13] Ali Altis and Patrick. So it shows [6:17] something that you you've had in the [6:18] Silicon Valley forever for many [6:21] generation. But it shows the impact of [6:22] having one success story that creates a [6:25] generation of talents who can invest, [6:27] who can start something new and help the [6:29] rest of the ecosystem. So that's what we [6:31] need more of, but this is also what we [6:33] start to see more and more in Europe. So [6:35] not an epilog, but a full sequel. Full [6:36] sequel. Full sequel. We're ready for it. [6:38] Maybe a movie. Yeah, we'll tell Ashley [6:41] Vance. Um, from your vantage point as [6:43] both an investor and also an operator, [6:46] where do you think the French ecosystem [6:49] really outperforms broader Europe? And [6:52] where is the hype perhaps a bit ahead of [6:54] the substance? I think we're we're [6:57] probably the best. I'm going to be [6:59] humble there. Full stop. He just leaves [7:01] it there. No, we're we're amazing at at [7:03] training the top 01% of technical [7:06] talents. Whether it's people who go to [7:08] eomal superior poly technique or people [7:10] who just learn on their own, you know, [7:12] hacker style uh how how to how to uh [7:15] develop software. I think we we have [7:18] extraordinary developers. What we did [7:20] not have back then was the ability to [7:22] distribute and market this product. So [7:24] that's why you see a lot of great French [7:26] companies with amazing products poorly [7:29] distributed. I think with the new world [7:31] of software we live in product growth AI [7:34] you know stripe and all of that it's [7:35] easier than ever to distribute software [7:37] everywhere but I think it's still a [7:39] skill set we need to develop when you [7:40] see a French entrepreneur pitching [7:43] usually you don't tell yourself you've [7:44] met Cherberg or Mark Benov you've met [7:47] fantastic scientists or developers but [7:49] we need that that ability to show up to [7:52] excite people and to sell a vision to [7:54] sell you know the the the big picture to [7:57] people who you know need to be convinced [7:59] But that's a skill set we we really [8:01] never had. I think Bernalo was the first [8:03] to do it properly when he when he [8:06] started enlisted business object. But I [8:09] think we are working on it. But it gives [8:10] me hope because we have great products. [8:12] We just have to be good at distributing [8:13] them. I I would not have to want I would [8:16] not want to have the uh the opposite [8:18] problem. Yeah. Speaking of being great [8:20] at distribution and many things, Sam [8:23] Alman has talked about kind of the one [8:25] employee unicorn, right? sort of there's [8:28] enough infrastructure and functionality [8:30] including with AI lifting all boats that [8:33] one employee can actually build a giant [8:35] company. If this actually happens, how [8:38] does venture capital need to reinvent [8:40] itself? [8:42] I think that's why I do seed and precede [8:43] because I want to get as fast as as [8:45] possible into these companies who might [8:47] never raise money ever again. And we've [8:49] seen that in gaming. We we know many [8:51] gaming companies that have been [8:52] completely self-funded, bootstrapped and [8:54] who became, you know, unicorns and more. [8:57] So that model that that model that we [9:00] saw in gaming is happening in SAS right [9:01] now. I think we'll have to adjust. Some [9:03] of them will want to accelerate maybe [9:05] the growth. We'll want to open offices. [9:07] We want to go global. We want to and [9:09] we'll need more developers. But I do [9:11] believe that we'll start to see that [9:12] more and more. maybe not, you know, [9:14] billion-dollar businesses, but fantastic [9:16] businesses, profitable businesses, [9:18] global businesses with a team of of one [9:20] or two or three people. Thanks to AI, [9:23] but also thanks to the fact that [9:25] building software cost nothing. [9:26] Distributing software cost nothing. You [9:28] know, Facebook ads and, you know, [9:29] Instagram ads and Tik Tok ads really [9:32] work. Uh, and monetizing software has [9:34] never been easier. So, it's an [9:36] opportunity for people who can build [9:37] great products and who can use the [9:39] latest tricks including AI to distribute [9:41] them and and monetize them. That gives [9:43] me a lot of hope. Yeah. Do you think [9:45] this world will benefit more operator [9:48] VCs and that dumb money will have a [9:50] harder time if that's my bet. [9:53] It needs to work. I don't know. I mean, [9:54] if you look if you look in the US, the [9:56] best VCs are former operators. uh if you [9:58] look at you know the the the benchmark [10:00] the Kleiner Perkins the USV um Sequuia [10:04] you see a bunch of people who worked at [10:06] great companies or started grand [10:07] companies it's a bit different in Europe [10:09] to be fair the best VCs in Europe uh are [10:12] born VCs that's what they've done most [10:14] of their lives or they are banker so I [10:16] think that trend is changing I think [10:17] you'll see more and more operator but [10:19] more importantly entrepreneurs becoming [10:21] investors and I think it helps to [10:23] connect with founders it helps to try [10:25] and you know modestly be be helpful [10:28] But it's it's also changing. We also see [10:30] all of that smart money from the US [10:32] coming to Europe. I mean Sequua had [10:34] nothing in Europe 5 years ago. Now they [10:35] have an amazing office. We saw IVP [10:37] coming in as well. NEA as well. So [10:40] people realize that you can find amazing [10:42] companies and talents in Europe. You've [10:45] written about us entering the Flappy [10:48] Bird era of SAS. Yeah. Um so you know AI [10:51] native teams can hit 40 million or even [10:54] 100 million in ARR with 25 people teams. [10:58] Um but then they can face these kind of [11:01] instaclones. And so I'm curious what are [11:03] the durable moes that founders should be [11:05] focused on building um given that any [11:08] clever rapper can potentially be [11:10] replicated overnight. I mean this is [11:12] something that Stripe has done extremely [11:13] well which is building a brand a brand [11:15] that people love a brand that people [11:17] trust. of course a great product uh that [11:20] does the job better faster than anything [11:22] else and when you have that you know if [11:24] you can build a community around it [11:26] almost a cult among developers and users [11:28] this is I think how you can build a mode [11:31] and again it's not a skill set we you [11:33] know we usually have especially in [11:34] Europe but I think it's changing look at [11:36] lovable do you guys know lovable they [11:39] went from zero to $50 million in no time [11:42] everybody knows the brand they have 50 [11:44] competitors but people keep going back [11:45] to lovable because they trust them they [11:47] do a great job stimulating, fostering, [11:49] managing that community of developers. [11:51] And so that's why they they they they [11:53] keep winning. But when you see these [11:56] companies going from 0 to 10 in 24 [11:58] hours, it also means that someone else [12:00] could come and do something better, [12:01] cheaper, faster in no time. So the brand [12:04] the the the the quality of the [12:06] documentation of the or the support if [12:08] you're in a platform company can make a [12:10] a massive different a massive [12:12] difference. There's a big premium to [12:13] being able to build that community and [12:15] that cult. Whoever named lovable was [12:18] brilliant. If you want to be loved, just [12:20] call yourself lovable. Like, full stop. [12:22] Um, so AI systems are becoming [12:25] semi-autonomous. I'm curious what you [12:28] think the most valuable human skill that [12:30] developers need to have and hone is in [12:34] order for them to stay indispensable. I [12:35] would say it's that it's being able to [12:37] uh build the community, create that [12:39] developer love in the ecosystem, [12:41] especially if you're building a a [12:43] platform or if you want to build a [12:44] platform company. But not just build a [12:47] great product but also build a great [12:49] brand and a great community of people [12:51] who love it, who use it, who pay for it, [12:53] who recommend it to their friends. This [12:55] is when when you when you can do that [12:57] all at the same time, which of course is [12:59] extremely hard. I think you can uh you [13:01] can start to dream very big. But it's [13:03] it's not easy when you come from a [13:04] developer background. Do you think the [13:06] machine will manage the developer or [13:10] will it stay the other way around? No, [13:12] I'm I'm I'm you know when AI when people [13:15] started talking about AI, people talked [13:17] about the Terminator Skynet scenario [13:19] where the machine replaces the human. [13:21] The what I the movie I prefer to think [13:23] of is Star Wars and Air2-D2. So Air2-D2 [13:26] as you know is helping Luke to make the [13:28] right decision is the robot is sharing [13:30] important information with Luke but at [13:32] the end of the day it's Luke Skywalker [13:34] who makes the decisions and who ends up [13:36] you know shooting the Death Star. So I [13:38] think of AI as a robot on your side that [13:40] can help you to to remove blind spots to [13:42] make better faster decisions but that [13:44] cannot replace you. That's my you know [13:46] optimistic vision of the future of AI. I [13:49] was talking to Scott Woo the co-founder [13:52] of cognition uh about how they use Devon [13:55] their agent internally and he told me [13:57] every engineer there aren't that many [14:00] engineers there 15 or something. Every [14:02] engineer at cognition has 15 devans [14:06] working for them at the average moment [14:08] in time. Yeah, just if all of us could [14:09] be 15x more productive, that would be [14:11] would love to quite something. Um, Satia [14:13] Nadella recently proclaimed that SAS is [14:16] dead. Um, you've argued instead that [14:19] we're heading towards a new SAS [14:21] revolution, in particular for frontline [14:24] workers who've historically been largely [14:27] underserved. Um, how should those of us [14:29] in this room reconcile those two [14:31] perspectives? Well, I I've been obsessed [14:34] by Frontline SAS for a while. When I was [14:36] at Facebook, I I created a business [14:38] called Workplace that is like Facebook [14:40] for work or it's an internal [14:41] communication platform that we thought [14:43] only tech companies would use. The [14:45] surprise we had is that the people who [14:46] are buying workplace were companies like [14:48] Walmart, Starbucks, Hineken, Club Med, [14:50] Danon, companies with a lot of frontline [14:52] employees and a lot of knowledge [14:53] workers. And so I discovered by accident [14:56] completely that market which is the blue [14:59] collar frontline SAS market. So people [15:01] who go to work every day, they don't [15:03] have a desk, they don't have a PC, they [15:05] don't have an email. And yet, just like [15:07] you and me, they want to use software, [15:08] they want to be augmented by AI. And so [15:11] I think, you know, SAS is just getting [15:13] started for these people. Again, they [15:15] don't have an email because it's too [15:16] expensive, but AI can do amazing things [15:18] for them. Think, you know, shift [15:20] management, inventory management, PTO [15:22] management, and all of that. So I think [15:24] there is that blue ocean market of two [15:27] 2.5 billion people that have yet to be [15:29] augmented and revolutionized by [15:31] software. So I I don't think he means [15:33] SAS is dead especially because he runs [15:35] the biggest SAS company in the world. [15:37] But I think there will be many [15:38] opportunities in the most unexpected [15:40] places and the most unexpected markets [15:42] for people who want to start businesses [15:44] or invest in these businesses. [15:46] Um, you talked a little bit earlier [15:48] about some French founders having a hard [15:50] time getting the full distribution and [15:52] reach and even many large European [15:56] corporates still hesitate to buy local [15:58] software until it's sort of quote [16:00] impossible to get fired for I think were [16:02] the words you used. So what are the [16:04] concrete changes in in policy in [16:07] procurement in incentives that need to [16:11] change in order to flip that mindset? [16:14] It's hard. I'm not sure a policy change [16:16] would do would do much. Um, but to give [16:18] you a real example, I was discussing [16:20] with Elonor Krespo. She's the CEO of [16:22] Pigmont. She sells big sophisticated [16:24] complicated software competing against [16:27] Oracle, Annaplan, uh, even workday. And [16:31] so to give you an example, it took her [16:33] six months to sign Coca-Cola, Coca-Cola [16:35] Inc. like at the HQ. And I think she's [16:38] been trying to sell to Katar company for [16:40] more than two years. Hm. And it should [16:43] not be this way. You know, it's a French [16:45] company. The product is developed and [16:47] it's 10 10 minutes from here. If you buy [16:50] it, you have access to the people [16:51] building it. You have access to the co [16:53] if you want to give feedback. You have [16:54] no reason to buy software built, you [16:57] know, outside of outside of the country [16:58] if you want the the best and the [17:01] brightest thing. But people still do it [17:03] because you don't want to be fired for [17:04] betting on a startup. And I think in the [17:06] US for example Fortune 500 companies [17:09] they think of software as um as an [17:11] investment and there will be a return on [17:13] investment. I think in Europe and in [17:15] France in particular big companies tend [17:17] to think of it as a cost and so you have [17:20] to reduce the cost to the minimum. You [17:22] you don't think there will be a return [17:23] on investment and so that needs to [17:25] change and I think if anything we need [17:26] to highlight and encourage people who do [17:29] make this bet to buy local software and [17:32] uh and make them look good. [17:33] Surprisingly, French SMBs buy more local [17:36] software than the big companies. [17:38] That's fascinating. Because they just [17:39] want something that works. They don't [17:41] care, you know, where it's coming from. [17:43] They just want something that works. And [17:44] so if you have a local provider doing a [17:46] good job at reaching out the longtail of [17:48] SMBs with a great product, that's [17:50] enough. They give people a chance. Big [17:52] companies in France, again, you you [17:55] don't get fired for for going IBM. And [17:57] that's that's still a problem. The [17:59] reputational bit is super interesting [18:00] because in the US I think in some cases [18:02] it's actually the reverse which is if [18:04] you're at a large enterprise and you're [18:05] not adopting cutting edge technology and [18:07] you're say sitting in the IT [18:09] organization you risk being stamped as [18:11] sort of a lite right um and so that [18:14] cultural shift is is striking um [18:16] everyone loves to hate meta but they've [18:19] been particularly impressive in their [18:21] ability to be kind of constantly at the [18:23] frontier of tech and in some cases it's [18:25] worked out like mobile and AI and in [18:27] other cases is less so like Libra and [18:30] Metaverse, but you've lived through [18:32] those pivots um from the inside. And so [18:34] I'm curious, what are the lessons from [18:36] Meta that you think established [18:39] companies need to remember? [18:42] It's hard because Meta is quite quite a [18:44] unique company just because of the way [18:45] it is run by the the founder who's still [18:48] there who controls everything. And as [18:50] you've seen, sometimes he make he makes [18:52] big bets. That can work. Mobile is a [18:55] good example, but I remember when when [18:57] we when Facebook pivoted to mobile in [18:59] 2012, uh people were still shorting [19:01] Facebook, people did not understand that [19:03] Facebook was doing it. So, we had the [19:04] hardest time telling the market. We know [19:06] mobile is coming. We're doing all the [19:08] right things. It will take time. And [19:10] then after the IPO, every three months [19:12] at the Onyx call, you would see the [19:14] market share of mobile versus web going [19:16] up. And so, and next thing you know, in [19:18] 2014, 2015, Facebook was a mobile first [19:20] almost mobile only now company. Um, but [19:23] sometimes they get it wrong. like I [19:24] don't think the metaverse was a was a a [19:27] great investment uh but some of the [19:29] investments they've done in a metaverse [19:30] help Meta to build the AI infrastructure [19:32] that they have. So having a founder that [19:34] is making build big long-term goals uh [19:38] with the right people internally I think [19:40] can make a difference but [19:42] you know that very well but Meta has not [19:44] always been right uh in the choices [19:46] they've made but they've been resilient [19:48] and when Zuck goes big it really goes [19:51] big like he's doing right now in AI and [19:52] so it's very inspiring to watch and even [19:54] when it doesn't work you learn from it [19:57] I've launched more stuff that failed [19:59] very publicly at Facebook than stuff [20:01] that worked but the conviction is [20:03] striking. Yeah. Yeah. Yeah. I can give [20:05] you a story if you want to. Yeah. I want [20:06] to hear the story. In 2012, I thought [20:08] the future of the Facebook platform was [20:10] gambling on the Facebook platform. So, I [20:13] convinced Cheril and the gang to [20:15] basically let me launch Bingo, real [20:17] money gaming on the Facebook platform on [20:19] the web. We launched, made a lot of [20:22] noise. There is an article of me in BF [20:24] talking about the future of gaming on [20:26] Facebook being bingo and slot machines. [20:29] and and after a month we made like €200 [20:32] euros and we killed it. But I had the [20:35] license to try I had the license to try. [20:38] It was [20:40] big very rapid failure. Thankfully it [20:42] was it was not too long. Uh and we [20:45] learned from it and so when you have a [20:47] license to try new stuff like that uh [20:49] you feel great as an employee and I [20:50] think as a shareholder as well. Will you [20:53] entertain a quick rapidfire closing? [20:55] Let's do it. Okay. Most underrated city [20:58] to start a company in 2025. Mier, [21:04] I was an exchange student there for a [21:05] hot second. Okay, what's the company? [21:06] Sorry, it's supposed to be rapid fire, [21:07] but I got to know. M it's in the it's [21:09] home. It's in the south. It's sunny. [21:11] You're very close to the Mediterranean [21:13] beach and you have some great companies [21:15] there like Septio or Sw. Amazing. And [21:17] it's three hours and 30 minutes from [21:19] Paris. Perfect. Good airport as well. [21:22] You you wanted some some surprises. I [21:24] love it. I love it. one European [21:26] regulation that you'd scrap tomorrow. Uh [21:29] probably regulations around, you know, [21:31] stock options and in in Europe, it's a [21:34] mess. It changes, you know, depends if [21:36] you're in France or in Germany. It's [21:38] it's a nightmare. We have to make it [21:39] very easy to align the interest of the [21:41] employees with the interest of the [21:42] shareholders. One French regulation that [21:45] you would scale globally, I would say [21:47] the tech visa and the uh immigration [21:49] policy. It's very easy to bring top [21:51] talents in France. It's not that easy in [21:53] the UK for example and in the US you [21:55] know we don't even know if it will be [21:57] possible to go and work there anymore [21:58] but that's an opportunity for Europe. [21:59] That's a massive opportunity for Europe. [22:02] Uh your favorite under the radar side [22:04] project. Uh it's it's a thing I it's a [22:07] side project I'm doing at 20VC. It's a [22:10] company I'm incubating uh that I will [22:13] talk about in a in a few months. But [22:14] that that's keeping me busy. So we're [22:16] not just investing in companies, we're [22:17] also incubating projects now. [22:19] Fascinating. Um, okay. This is going to [22:21] be a start bench cell. Yeah. Yeah. Okay. [22:24] But tech edition. Go ahead. All right. [22:26] Steve Balmer, Mark Zuckerberg, Savaniel. [22:31] It's impossible to that question. Okay. [22:34] I love I love them all. [22:36] They're very different. I mean, they're [22:37] very lovable. If you start a company, [22:39] you need the three the three of them. [22:41] You have the the sales guy, obviously. [22:43] You know who it is. You have the CTO and [22:45] the CPO. Uh, and one of them could be [22:48] the CEO. But no, I worked for for all of [22:51] them. I worked with Xavian at some [22:52] boards. But uh all of the things we're [22:54] discussing right now, the fact that we [22:56] can do this and that there is such a [22:57] vibrant ecosystem in France is because [22:59] of everything Zean has done, you know, [23:01] for the last 15 years. You see his DNA's [23:04] fingerprints on every company that is [23:06] doing very well in France right now. And [23:08] he built not just, you know, the the the [23:10] funds to to and the the st to do it, but [23:14] just the infrastructure to uh to [23:16] accelerate and to empower that new [23:18] generation. So I would say I would [23:20] choose Xavier. [23:22] Last one. And I know you live between [23:24] the two. Paris or London? Uh [23:27] Paris for vacation, London for work. [23:29] Wonderful. Well, thank you Julian for [23:32] such sharp and grounded insights and for [23:34] reminding us that software is a bet on [23:36] optimism and conviction. So, um, thank [23:39] you so much for joining us today. Thank [23:40] you. Thank you.