SaaS isn’t dead, it urgently needs a new purpose
On December 12th, Satya Nadella, CEO of Microsoft referred to SaaS applications as nothing more than “CRUD databases with a bunch of business logic.” A “CRUD” database describes databases that perform the basic functions of Create, Read, Update, and Delete operations on data like a CRM.
The business logic then emerges from the rules, processes and workflows that define how the data is used to achieve business goals - pretty straightforward.
What Microsoft's numbers say
Contrary to Satya Nadella's characterization of SaaS applications, Microsoft's own financial success tells a different story. Their SaaS offerings have evolved into powerful platforms that deliver substantial value, contributing approximately $100 billion (40.8%) to their total annual revenue of $245.1 billion.
The Productivity & Business Processes segment accounts for $81.2 billion through platforms like Office 365 and Dynamics 365. LinkedIn further demonstrates how SaaS can combine professional networking with AI-driven insights to create unique value. The Intelligent Cloud SaaS segment contributes $18.8 billion through specialized solutions like Microsoft Defender and Virtual Desktop.
These numbers demonstrate that modern SaaS applications deliver value far beyond basic data management.
From CRUD Silos to Smart, Integrated Agents
Nadella suggests that the trend around AI Agents will most likely lead to a redundancy of traditional SaaS applications.
In his view, agents won’t just be limited to a single database or application; they will be capable of handling CRUD operations across multiple data sources and simultaneously handle the business logic by adding context.
A daily example for this can be an e-commerce platform that is relying on multiple SaaS applications such as Salesforce for customer and marketing data and Shopify for inventory management.
These SaaS applications rely on separate CRUD databases to manage customer information, inventory and marketing campaigns. However, the business logic is defined within each platform, which creates silos that require manual integration.
An AI agent on the other hand is now capable of connecting to all of these data sources, accessing customer data and marketing insights from Salesforce as well as inventory data from Shopify.
By doing this, the agent can perform CRUD operations across these databases simultaneously and is now able to generate personalised marketing campaigns, updating inventory in real time, and managing customer orders autonomously. A real-life example could look like this: A customer abandons a cart on the website, the Agent reads the data from Shopify, updates the CRM with customer behaviour data, and triggers an email campaign in Salesforce tailored to that customers preferences.
What’s beneath the surface?
It’s true that many SaaS products today have become indistinguishable, offering the same basic functionalities with strikingly similar polished UIs. Beneath the surface, many of these tools fail to deliver meaningful innovation or efficiency, prioritizing locking customers into expensive subscriptions over long-term user satisfaction.
Along with the AI buzz, platforms are presenting shallow “AI-powered” features that add little value and can be done just as easily - and often more flexibly - by users directly in ChatGPT etc., rather than making genuine improvements to their core offerings.
However, there is a crucial distinction between superficial AI features and those built on robust foundations. The key differentiator lies in how deeply AI capabilities are integrated with a platform's underlying data structure - specifically, its system of record.
Business-critical SaaS applications have a powerful foundation for AI development: their systems of record. This concept evolved from early platforms like Salesforce and HubSpot, which initially created structured databases organizing sales data in traditional rows and columns.
As businesses adopted more digital tools, they faced a new challenge. Sales, accounting or customer support processes began generating diverse data across multiple platforms, creating disconnected information silos. This fragmentation made it difficult to gain comprehensive insights from your own data.
The advent of LLMs is transforming these systems of record. Modern SaaS platforms nowadays can now process three key data types:
Unstructured data (text, images, voice, video)
Multi-source information (emails, calls, product usage)
Real-time updates and context
This evolution makes these platforms ideal for AI features because they:
Provide complete visibility across entire business processes
Enable AI learning from diverse data types and sources
Support real-time adaptation and connect previously isolated information
The extensive data foundation allows AI to not just analyze historical patterns but actively improve business processes through intelligent automation and personalization.
After taking a closer look at the capabilities of modern SaaS tools and how they serve LLMs and AI initiatives, it can be considered to be more than just a CRUD database with some business logic.
The future we see for SaaS
Understanding these foundational elements raises an important question: How are successful SaaS companies leveraging these capabilities to create genuine value? Our answer lies in two key strategic approaches:
Core Components
Smart SaaS companies are weaving AI deeply into their products - not as flashy add-ons, but as core components. The future isn’t about standalone AI replacing SaaS; it’s about hybrid models where SaaS leverages AI to supercharge workflows, uncover insights, and simplify user experiences. At the same time, enterprise SaaS is evolving to prioritize deeper integration, robust security, strict compliance, and advanced data insights to maintain a competitive edge.
Slite - which integrates AI to transform knowledge management by enabling instant information retrieval and simplifying documentation is a leading example. Their approach is not to release just add-ons; they fundamentally transform the way teams interact with information, improving efficiency and usability at every level.
Hyperfocus
Generic, cookie-cutter products are at the greatest risk of extinction. From now on, SaaS will thrive by targeting niche markets with highly specialized solutions. Companies that innovate beyond CRUD operations and basic business logic solely in one industry will attract loyal customers who value specialized features that go beyond what AI agents alone can provide.
Bezahl.de is a great DACH example of a hyperfocused platform, providing a specialized payment management solution for the automotive industry. By automating invoicing, payment tracking, and financial integration, they address the specific needs of these customers (e.g., car dealerships, manufacturers, and service providers), offering a tailored solution beyond generic financial software.
SaaS isn’t dead - it’s just gotten lazy. The era of coasting on basic features, cookie-cutter interfaces, and inflated pricing is over. As AI integrates deeper into the core of SaaS products and hyper-focused solutions cater to specific industries, the market demands real innovation. The future of SaaS is about offering specialized, value-driven solutions that transcend basic CRUD operations, delivering real impact and long-term user satisfaction. Right now, SaaS is in the process of reinventing itself, evolving to meet the demands of a rapidly changing, AI-driven landscape.