Upending SaaS Common Sense: Bret Taylor on 'AI x Business Model Reconstruction'
In recent years, the rapid evolution of AI technology has shaken the foundations of the software business. Traditional models such as 'product licensing' and 'subscriptions' are being redefined by the emergence of AI agents. In this article, based on a lecture by Bret Taylor, former CTO of Google/Facebook and current leader of Sierra, we will explain how AI is reconstructing software business models, incorporating specific examples and his own statements.
1. Shifting Perspective from Engineer to Executive
While Bret Taylor is praised as a '10,000x engineer,' he analyzes himself as a 'single-issue voter,' stating that one cannot scale an organization by being obsessed only with the product. In an episode where he was reprimanded by COO Sheryl Sandberg when he became CTO of Facebook at age 29, he learned the lesson:
'You won't scale if you don't manage the team properly instead of trying to do everything yourself.' Through this experience, he learned
Breaking out of the self-contained 'comfort zone' and changing roles according to organizational challenges
Prioritizing 'what needs to be done now' over 'one's own strengths' in task allocation
Continuous growth through self-reflection
He preaches the importance of this mindset.
2. Sierra's Vision: 'New Customer Touchpoints' Opened Up by AI Agents
Taylor speaks as follows about Sierra, which he founded with co-founder Clay.
'Every company's digital interface will eventually become an AI agent.'
While websites and mobile apps were once the core of the customer experience, he holds the vision that 'the majority of digital interactions will shift to being via agents within the next five years.'
2-1. A 'One-of-a-Kind Agent' That Embodies the Brand
At Sierra, they collaborate with diverse companies such as ADT and SiriusXM to build 'primary customer agents' that embody each company's brand.
'If a website is the brand's "signboard," then an AI agent is the "actor" who plays it.' By automating complex use cases such as customer inquiry handling, troubleshooting, and contract renewals, they are reconstructing traditional touchpoints in a unified manner.
3. Redefining Business Models: Performance-Based Billing and Microservices
Taylor presented the following three classifications for the business models of AI-adopting companies.
3-1. Foundation Models
He predicts that the foundation model market, which requires massive computational resources, will become an oligopoly like cloud infrastructure, creating a structure that collects 'tax revenue' with low margins and high scale.
3-2. Tool Market (Pickaxe/Gold Rush)
The data tool market, such as Snowflake and Databricks, is seeing intensified competition due to horizontal expansion from foundation models. This is an area also exposed to threats from foundation companies.
3-3. Applied AI Agents Market
The most exciting development is this 'vertical-specialized agent.' He suggests that agents specialized by domain—such as law, marketing, and customer support—will emerge as a new form of SaaS.
Furthermore, the 'outcome-based pricing' adopted by Sierra sets it apart from traditional per-seat or usage-based billing.
By charging based on the number of problems solved and providing escalations for free,
the goal is to align the ROI of client companies with the profits of the vendor.
4. Implications for Startups and Incumbents: Vertical Specialization and Go-to-Market Strategy
Mr. Taylor expresses skepticism toward 'horizontally scalable platforms' and emphasizes 'vertical specialization' in niche areas.
'A platform that is just slightly better horizontally tends to get buried by existing internal solutions. Vertical specialization to “alleviate pain” is the only breakthrough.'
Also, in Sierra's own Go-to-Market strategy,
Grasping needs through deep research
Demonstrations and value propositions tailored to specific challenges
Building teams well-versed in vertical industries
are strictly implemented to differentiate themselves in the saturated AI vendor market.
5. Future Outlook: The Birth of Trillion-Dollar Companies and a Business Model Revolution
Mr. Taylor predicts that 'the first trillion-dollar company in the SaaS market will be born through outcome-based sales via agents.'
Going beyond productivity improvements to directly sell the value of 'outcomes'
The evolution of business models from licenses to subscriptions to outcome-based billing
New entrants possess the 'greatest advantage' to disrupt legacy business models
From now on, the key to determining the next generation of winners will be how to design a 'business model that connects customers with outcomes,' rather than just embedding AI technology into products.
What we can learn from Bret Taylor's insights is that the true value lies not in the 'technology itself,' but in 'how to convert technology into value.' By looking beyond the realms of engineering and data science and re-evaluating AI from a management perspective, the software business can open up new horizons—that challenge is beginning right now.
Related Articles
