Key Moments
Why Most SaaS Companies Will Fail at AI (And How to Avoid It) with Intercom's CPO
Key Moments
SaaS companies must 'refound' themselves for AI, not just add features, to survive.
Key Insights
Transforming into an AI-first company requires a fundamental 'refounding' of the entire business, not just incremental feature additions.
The post-AI world shifts focus from 'seats' (human users) to 'outcomes' enabled by AI, making traditional SaaS models obsolete.
A successful AI transformation involves embracing radical change, eliminating 'sacred cows,' and a willingness to significantly disrupt existing operations and products.
Building true AI capability requires expertise across AI layers (RAG, custom models) and an 'app layer' for access, demanding a different approach to software development.
Companies must be prepared for difficult decisions, including potentially parting ways with employees who resist the necessary transformation.
Marketing AI products requires a different strategy, focusing on clear value, demonstrable performance at scale, and educating customers who are often new to AI.
The buyer landscape for AI solutions is more complex, involving C-level executives and AI specialists, necessitating tailored go-to-market strategies.
THE IMPERATIVE OF AI TRANSFORMATION
The advent of AI marks a fundamental shift, rendering traditional SaaS business models focused on 'seats' and manual 'CRUD' tasks obsolete. Companies must recognize that AI is an inevitable technological cycle, akin to the internet or mobile revolutions. To remain relevant, businesses need to ask themselves why they exist in a post-AI world. The future belongs to companies that can deliver superior outcomes built on robust AI, including advanced RAG systems and custom models. This transformation is not optional; it's a necessity for survival, requiring businesses to pivot rapidly to stay competitive.
INTERCOM'S JOURNEY: FROM DECLINE TO AI LEADERSHIP
Intercom, once a successful SaaS company, faced a critical juncture with five quarters of declining revenue growth. The company, in a potential IPO journey, was in a precarious position. The arrival of ChatGPT three years prior signaled a major technological inflection point. Recognizing this 'one-way door' for humanity, Intercom's leadership, including CEO Owen and CPO Paul Adams, made the bold decision to bet the entire company on AI. This led to the rapid development and launch of 'Finn,' an AI agent for customer service, which has since achieved over a million resolutions weekly with a 65% average resolution rate, transforming Intercom's trajectory.
THE 'REFOUNDING' OF A COMPANY: CULTURE AND SACRED COWS
Achieving AI transformation necessitates a complete 'refounding' of the company. This involves transforming the organization, product strategy, roadmap, development processes, metrics, go-to-market approach, and pricing. It's a deep cultural shift that requires challenging deeply ingrained habits and eliminating 'sacred cows.' Employees are naturally averse to change, and subtle resistance can manifest as delays or a preference for incremental additions over radical overhauls. Acknowledging and addressing this human tendency is crucial; without embracing profound change, a company is not truly adopting an AI-first mindset.
BUILDING AND MARKETING AI: A NEW PARADIGM
Developing AI products requires a different approach than traditional SaaS. It involves building and iterating on complex AI layers (RAG systems, custom models) and an accessible 'app layer.' This necessitates a shift in software development from a focus on user interfaces and easily observable features to robust, often invisible, AI infrastructure. Marketing AI products is equally challenging, as many products appear similar. Success hinges on demonstrating tangible outcomes, scientific rigor, and performance at scale, rather than just flashy demos. Customer education and advocacy become paramount as buyers navigate this new technological landscape.
THE NECESSITY OF 'SELF-HARMING' DECISIONS
True AI transformation often demands 'self-harming' decisions – actions that may negatively impact short-term revenue or comfort levels but are critical for long-term success. This can include letting go of a portion of the workforce that is not aligned with the new direction or revenue protective measures that hinder true AI adoption. The transition is inherently painful; if it doesn't feel difficult, the company likely isn't transforming deeply enough. Leaders must be willing to upset established norms and, at times, part ways with individuals unable to adapt.
REIMAGINING THE PRODUCT AND GO-TO-MARKET STRATEGY
Simply adding AI features to an existing SaaS product is insufficient; companies must reimagine their products entirely. Intercom's Finn, for example, is fundamentally different from its original SaaS offering. The go-to-market strategy also needs a complete overhaul. Customers are often not equipped to evaluate or purchase AI solutions. This requires significant investment in customer education, clear identification of the strongest differentiators (often infrastructure and performance, not just UI), and a multi-faceted approach to engage a new buyer landscape that includes C-level executives and AI specialists alongside traditional stakeholders.
THE EVOLUTION OF SOFTWARE DEVELOPMENT AND DESIGN
The process of building software has fundamentally changed. AI development is characterized by empirical evaluation, continuous experimentation, and a deep understanding of discrete steps within a workflow to ensure reliability. Unlike traditional SaaS where UI and UX were paramount, AI's complexity lies in its infrastructure and model layers. Design itself has become more accessible, with product managers and engineers capable of prototyping and iterating rapidly, empowering designers with coding skills to fix front-end issues and enhance the development velocity. This shift requires adapting long-standing development principles and embracing a more chaotic yet potent development environment.
NAVIGATING THE CHAOS AND AVOIDING COMMON MISTAKES
The transition to an AI-first company is often chaotic and unpredictable. Common mistakes include failing to truly reimagine the product, protecting existing revenue streams at the expense of necessary transformation, diluting the vision, or delaying critical changes. Companies must also be wary of listening too much to customers who resist AI. Honest, soul-searching conversations within leadership teams are essential to maintain focus and adapt. Ultimately, success hinges on the courage to embrace radical change, confront difficult truths, and commit to building a fundamentally different, AI-native organization.
Mentioned in This Episode
●Products
●Software & Apps
●Companies
●Organizations
Common Questions
Traditional SaaS models focused on 'seats' for human users performing CRUD tasks. In a post-AI world, the 'seats' concept is obsolete, and products are increasingly invisible, with value derived from AI-driven outcomes.
Topics
Mentioned in this video
An AI agent for customer service launched by Intercom in March 2023, which has achieved over a million resolutions per week and has over 6,000 customers with a 65% resolution rate.
Mentioned as the catalyst for Intercom's strategic shift towards AI, appearing three months after the co-founder returned as CEO.
Mentioned as an example of AI technology that has advanced significantly in recent years, contributing to the magical capabilities of current AI.
Mentioned as Apple's entry into the AI race, with a postponed launch to Spring 2026, serving as an example of the 'marketing overhang' and the gap between demos and production reality.
The speaker mentions working on the mobile team at Google when the iPhone came out, highlighting experience with significant technological disruptions.
Cited as an example of a major technological disruption, comparable to the impact of AI on the software industry.
Mentioned as a publication where companies that successfully transform to AI might be featured on the front page.
A company that has undergone a significant transformation from traditional SAS to an AI-native one, launching the AI agent 'Finn'.
Found this useful? Build your knowledge library
Get AI-powered summaries of any YouTube video, podcast, or article in seconds. Save them to your personal pods and access them anytime.
Try Summify free