Key Moments
⚡️ Competing with ChatGPT and Sierra, building a $10M ARR company — Yasser Elsaid, Founder, Chatbase
Latent Space PodcastKey Moments
Chatbase achieved $10M ARR by focusing on product-led growth and customer service AI, proving bootstrapped startups can compete with VC-backed giants.
Key Insights
Chatbase reached $1 million ARR in just 117 days by leveraging early AI hype and focusing on 'doing the business stuff' like sales calls and demos.
The company's growth strategy is primarily Product-Led Growth (PLG) with a strong self-serve focus, mirroring successful companies like Stripe and Vercel.
Yasser Elsaid believes that for B2B startups, focusing on traditional sales outreach like demos, sales calls, and cold emailing is more effective than chasing viral trends on platforms like TikTok.
Chatbase processes close to 100 billion tokens monthly across various models including OpenAI, Anthropic, and Gemini, indicating significant adoption.
The company is evolving its AI agents from basic customer support to 'chief customer officers' that handle sales, onboarding, and provide business improvement insights.
Open-source models saw an initial hype spike but usage has largely returned to major providers like OpenAI, Anthropic, and Google, due to switching costs for perfected harnesses.
From Student Project to AI Leader
Yasser Elsaid, founder of Chatbase, shares his journey from a computer science student focused on making money online to building a $10 million ARR company. Initially working on side projects for students, Elsaid pivoted when he realized the potential of augmenting GPT-3 with custom data, a concept that predated the widespread adoption of Retrieval-Augmented Generation (RAG). He saw an opportunity to create a product that could provide context-aware AI responses, a significant departure from the base model's limitations. This insight led him to drop other projects, take a break from school, and focus entirely on developing the first version of Chatbase, which he launched to just 16 Twitter followers. This early, focused effort was driven by a sense of urgency, a belief that speed was critical in the nascent AI space.
Rapid Growth Through 'Doing Business Stuff'
The initial launch of Chatbase went viral, but Elsaid emphasizes that viral moments alone are not enough. While many AI demos gained traction, they often lacked stickiness. Chatbase's success in reaching $1 million ARR in just 117 days is attributed to a deliberate focus on fundamental business activities. Elsaid proactively engaged in user calls, demos, and sales, differentiating Chatbase from many contemporaries who focused solely on product demos. This hands-on approach, coupled with leveraging the AI hype on platforms like LinkedIn where the audience was receptive to new products, proved crucial. For instance, a single impactful post about Chatbase could demonstrably increase Stripe MRR by several thousand dollars. This period highlighted the importance of converting initial interest into tangible business value through direct customer engagement.
The Power of Product-Led Growth
Chatbase's growth strategy is heavily rooted in Product-Led Growth (PLG), with a significant emphasis on self-serve capabilities. This approach mirrors that of successful companies like Stripe and Vercel, where users can easily sign up, experiment with the product, and derive value independently. Elsaid contrasts this with sales-led growth, which often requires a larger upfront investment in sales teams and can lead to assumptions about target markets that might miss niche yet lucrative opportunities, such as theme parks which discovered Chatbase organically. The PLG model forces the creation of an intuitive and powerful product that appeals to both small startups and large enterprises. While this slower initial growth is a trade-off, Elsaid views it as a more sustainable and advantageous long-term strategy, fostering a product that is appreciated even by larger, more discerning clients.
Navigating the AI Model Landscape
Chatbase processes a massive volume of data, with monthly token usage approaching 100 billion, utilizing a mix of models from OpenAI, Anthropic, and Google. Elsaid notes that while OpenAI still holds a significant share, Anthropic and Google are growing due to the company's policy of allowing customers to choose their preferred models. He explains that while the 'harness' or the framework around the model needs adjustments for each provider (e.g., instruction tuning), the core intelligence for applications like customer service has been present for a while. The key challenge lies in optimizing this harness to leverage the model's capabilities effectively. Interestingly, while open-source models initially generated hype, their usage has largely stabilized, with many businesses returning to established providers due to the significant switching costs involved in perfecting a harness for a specific model.
Evolving from Support to 'Chief Customer Officers'
The company is strategically evolving its AI agents from simple customer support tools to comprehensive 'chief customer officers.' This involves leveraging the AI's deep knowledge of a company's documentation, internal notes, and customer interactions to act as a brand ambassador across support, sales, and onboarding. The vision is for these agents to manage the entire customer experience conversationally, not just reactively addressing issues. By consolidating context from all business operations and customer touchpoints, the AI can offer personalized guidance throughout the customer journey. Furthermore, these agents will surface insights derived from countless customer conversations to help business owners improve their products and strategies, effectively becoming a CCO that understands the entire business landscape.
The Future of Customer Service and Integration
Chatbase aims to eventually replace systems like Zendesk, acknowledging the inherent desire to integrate while simultaneously competing. They are currently acting as an agentic layer on top of existing systems like Zendesk, seamlessly escalating complex issues to human agents. This phased approach mitigates the high switching costs for enterprises and builds trust. Over time, as clients appreciate Chatbase's capabilities, they are more likely to migrate entirely. Elsaid also highlights the importance of 'warm outbound' – reaching out to customers who have already shown interest, which complements the PLG motion. He believes AI agents can significantly enhance personalization, building context with each interaction to guide users more effectively and drive business value, making them indispensable in coding and customer service.
Bootstrapping to Generational Ambition
Despite being bootstrapped, Chatbase has achieved a financial position that allows it to operate with the aggressiveness of a venture-backed company, aiming to build a 'generational company.' Elsaid emphasizes that while bootstrapping can foster efficiency and creativity, it doesn't preclude ambition. The company is now in a phase where it can afford to be more aggressive in hiring and scaling, driven by a strong conviction in its growth strategy. He also discusses the evolving nature of hiring, seeking individuals with an open-minded, results-oriented approach to AI and product development, rather than those solely focused on the craft of coding. This adaptability is critical in an era where models and workflows change rapidly. For employees, Chatbase offers equity that holds substantial current value due to the lack of a high, potentially unattainable, valuation, making it attractive to top talent.
Optimizing Growth and Customer Relationships
Chatbase's go-to-market strategy centers on finding the 'equation where if you put $1 in, you get more out,' and then scaling that effectively. This involves a focus on executable B2B tactics like demos, sales calls, and outbound efforts, rather than solely relying on experimental growth hacks. Content plays a vital role, not just as a standalone piece, but as a support to other go-to-market strategies, enhancing brand recognition for paid ads and outbound efforts. User-Generated Content (UGC) is seen as the new influencer marketing, with employees encouraged to contribute consistently to build personal brands and company visibility. The company also values strong customer relationships, treating them as 'friends' and recognizing that word-of-mouth remains a significant growth driver. Pricing is viewed as a key lever, with experimentation encouraged, including outcome-based pricing for select larger clients, though usage-based pricing remains dominant due to predictability and perceived value by customers.
Mentioned in This Episode
●Software & Apps
●Companies
●Organizations
●People Referenced
Common Questions
Yasser Elsaid was studying computer science and working on side projects. He realized the potential of adding custom data to large language models like GPT-3, which was the foundation for Chatbase.
Mentioned in this video
Cited as an example of a company with a strong product-led growth (PLG) model, emphasizing easy-to-use tools that are powerful enough for large companies.
A client using Chatbase's services, illustrating the global reach and diverse applications of their AI solutions.
A customer service platform that Chatbase aims to replace or integrate with, acting as a strategic competitor and potential partner.
A financial services and software company, cited as an example of a company with a strong product-led growth (PLG) model and a major beneficiary of early AI adoption.
Mentioned as an example of an established company that, like Chuck-E-Cheese, is embracing AI nativity, highlighting the broad appeal of Chatbase's solutions.
A company founded by Yasser Elsaid, offering an AI chatbot for businesses that can be trained on custom data. It reached $10M ARR within 3 years as a bootstrapped company.
A company mentioned in relation to their potential or existing use of AI for customer service, highlighting the industry trend.
A notable customer of Chatbase, discovered through the self-serve platform, highlighting the diverse range of industries using their service.
An IDE for AI development used by the Chatbase team, with Yasser Elsaid also using its app for coding tasks. It's also mentioned for its AI agent capabilities.
The platform where Chatbase keeps its 'Go-To-Market Manifesto' document, detailing their growth strategies.
An early large language model from OpenAI that Yasser Elsaid experimented with, realizing the potential to add custom data.
An OpenAI model that predated ChatGPT's launch, which Yasser Elsaid used to build early versions of Chatbase.
The conversational AI model that launched, sparking a wave of AI-driven product development. Chatbase was built before its launch.
The API Yasser Elsaid used to build Chatbase, enabling the integration of custom data into large language models.
A framework used to build applications powered by language models, which was part of Chatbase's initial tech stack.
Mentioned in the context of offering different tiers for open models and fine-tuning, and the details regarding annual vs. monthly pricing.
An AI model used by Yasser Elsaid for research and deciding on new tools, and later as a tool that can be prompt-injected for biased results.
A mobile app company focused on reducing screen time, using Chatbase as their AI agent and featured in Chatbase's 'Around the World' customer story series.
An AI agent mentioned as being used by Yasser Elsaid, alongside Cursor, for exploring the background agent space.
An AI coding assistant that Yasser Elsaid uses, often alongside Codex, appreciating its app interface.
More from Latent Space
View all 211 summaries
92 min🔬Top Black Holes Physicist: GPT5 can do Vibe Physics, here's what I found
75 minCI/CD Breaks at AI Speed: Tangle, Graphite Stacks, Pro-Model PR Review — Mikhail Parakhin, Shopify
86 min🔬 Training Transformers to solve 95% failure rate of Cancer Trials — Ron Alfa & Daniel Bear, Noetik
49 min⚡️ How to turn Documents into Knowledge: Graphs in Modern AI — Emil Eifrem, CEO Neo4J
Ask anything from this episode.
Save it, chat with it, and connect it to Claude or ChatGPT. Get cited answers from the actual content — and build your own knowledge base of every podcast and video you care about.
Get Started Free