SaaStr Annual 2024: Building an AI Centered SaaS Bottoms Up Start-up with Otter.ai

In this session Otter.ai Co-founder and CEO Sam Liang, will discuss the strategy, decisions, and frameworks used to build an AI-powered bottom-up SaaS startup. He will explore the key decision of building your AI components in-house versus buying or licensing third party solutions - dissecting considerations that go into deciding to build versus buy and including personal anecdotes on how he came to these decisions and implemented them successfully.

1. How to Build a Bottom-Up AI SaaS Startup Sam Liang CEO & Co-Founder Otter.ai

2. Introduction Founded: 2016 with co-founder Yun Fu (Otter CTO) - with engineering background Problem to solve: capture and share lost knowledge from meetings Purpose: make conversations more valuable Product: Built the first AI meeting assistant,using in-house AI Collaborative GenAI So far: Processed +1 billion meetings +18 million users $63M Series B funding - by early investors in Google, DeepMind, Facebook, and Tesla.

3. Agenda Determining the Startup Vision and Market Go-to Market Strategies AI Technology and Infrastructure Key Takeaways Q&A

4. Determining the Startup Vision and Market

5. Identifying the Right Market Huge Market versus Niche Market: Have a bold vision and think differently Appeal to professional consumers and knowledge workers in enterprises Horizontal versus Vertical Approach Build a horizontal product that targets end users first, get bottom up adoption Identify the cluster of users in verticals: sales, marketing, recruiting, project management, user research, etc. Build vertical solutions on top of the horizontal platform

6. Problem Selection and Market Demand Build a product you use every day: Our own team and myself use the product every single day. We all debug the products, identify new needs and pain, and figure out new features. Need to build a product that Gen AI can provide real value. Defensibility Technical moat: can’t just be a thin layer calling OpenAI API High frequency usage, generate big data Feedback loop: learn from users usage, feed that back to the AI models, so that it can improve itself.

7. Go-to-Market Strategies

8. Viral Growth and Media Coverage Leveraging virality and media for user awareness Best to have intrinsic viral loop: sharing meeting notes with colleagues invites more users to sign up for the app Media reporters and Conferences, like TechCrunch Disrupt, adopted Otter early, generated free PR Building a user base with low marketing budget Achieving media coverage through constant innovative AI product launches Disruptive use cases generated lots of debate and buzz

9. Pricing Strategy Disruptive pricing Price it low, carpet bomb the market with a freemium model, get market share fast Exploring new pricing models Value-based pricing for AI products Experiment with different pricing strategies

10. Scaling and Enterprise Penetration Path to Enterprise Adoption Strategies for transitioning from individual users to enterprise contracts. Example: Aggregating individual enterprise users into enterprise contracts. Enterprise administrative control Security compliance: e.g. SOC2 Type II compliance Support each department with the vertical solutions: Sales, marketing, recruiting, product management, user research, etc. Empower them to collaborate on the same horizontal product.

11. AI Technology and Infrastructure

12. Build Versus Buy The public LLMs are generic models based on public data on the Internet It doesn’t have proprietary enterprise knowledge. For us, it does not perform well for multi-speaker conversational data Build unique models for your product specific use cases. Reduce cost of using 3rd party APIs: Allow scalability Helps increase gross margin Control, Customization, and Cost Management

13. Build a unique product that solves a real problem with proprietary AI models built with proprietary data.

14. THANK YOU!

15. UP NEXT: FY25 Strategies for Growth Denise Persson CMO Snowflake Nicole Baer CMO Carter Gail Moody-Byrd VP Marketing LinkedIn for Sales Carilu Dietrich CMO, Advisor Formlerly Atlassian