The Death of SaaS

🌈 Abstract

The article discusses the evolution of software development, from the early days of in-house built software to the rise of SaaS (Software-as-a-Service) solutions, and the potential impact of AI on the future of software development.

🙋 Q&A

[01] The Rise of SaaS

1. What were the issues with in-house software development in the early days of the author's career?

• In-house software was often built by poorly motivated summer interns, resulting in terrible "legacy software" that everyone complained about.

2. How has the software landscape changed with the advent of SaaS?

• The basic startup operating system is now a "spiderweb of SaaS", with many SaaS tools being used without the knowledge of the CFO.
• SaaS providers have become complacent, with little innovation, as they are "fat on their gross margins".

3. What are the downsides of the current SaaS landscape?

• The bills for SaaS tools can add up quickly for businesses.
• There is a need for additional SaaS tools to manage the SaaS bills.

[02] The Future of Software Development

1. How does the author envision the future of software development?

• Businesses will build features specific to their industry and use case, deploying them as lightweight web apps or microservices, and using APIs for things they don't want to build.
• AI and large language models (LLMs) will be the "beating heart" of this software, with entire workflows potentially being automated without the need for a user interface.

2. What are the potential benefits of this AI-powered software development approach?

• It can be more cost-effective than paying for SaaS tools, as demonstrated by the example of an intern building an AI-powered invoice processing pipeline in just two hours.
• AI is expected to increase developer productivity, potentially making code cheaper and more accessible.

3. Which types of companies are likely to be more exposed to the impact of AI on software development?

• Pure SaaS plays, where the software itself is the product, are at risk, as AI-powered alternatives could potentially replace them.
• Technical platforms like Datadog are also at risk, as their services could be replicated by AI.
• Companies that offer a service where the software just facilitates access to the product, such as Paypal, Stripe, and Shopify, are better positioned to survive the AI disruption.