The rise of OpenAI’s API and Anthropic’s Claude model brings up a big question: Is this the end of SaaS?
The concept is bold: Train AI models on company data to handle boring tasks like employee onboarding and customer-relationship management. The allure is to do it all in-house, bypassing the need to pay for a slew of off-the-shelf software tools. It’s a direct challenge to the established software-as-a-service (SaaS) model. A do-it-yourself future for businesses.
It’s an easy argument to make. More than 2 million engineers have downloaded OpenAI’s API, while Claude and other open-source models are quickly gaining traction in the enterprise. Consulting firms are offering services to help companies implement AI. And blog after blog contend that SaaS as we know it is all but dead.
That’s ludicrous.
We’re not seeing the death of SaaS. Instead, we’re just moving to its new phase: SaaS 3.0.
Evolution of SaaS
SaaS 1.0 represented the initial leap from physical software installations to cloud-based solutions, changing the way businesses accessed and scaled technology (think Salesforce). SaaS 2.0 built on the concept, embracing mobile integration, enhanced user interfaces, and collaboration tools to meet the evolving demands of enterprises (think Slack).
For decades, machine learning and iterations of AI have been embedded into these B2B applications. Everything from semantic search, automated content generation, recommendation engines, and big data have found their way into product suites. This made managing backend operations much easier for enterprises.
Now, with the emergence of large language models (LLMs), AI isn’t just an add-on but at the core of the application.
Enter SaaS 3.0: In this new phase, AI and machine learning are enhancing B2B applications to more effectively handle business-critical tasks. Through LLMs and deep learning, B2B vendors can automate entire processes, including payments, HR, CRM, content creation, and so much more.
To be fair, OpenAI’s API and enterprise-grade chatbot, along with other LLM vendors, have enabled some tech-savvy startups and hyperscalers to build their own tools to automate tasks.
But the reality is that most companies suddenly aren’t engineering powerhouses, capable of pumping out their own vertical software just by downloading an API. Engineers lack the time and expertise to build proprietary tech. And executives don’t want to invest in developing tools that already exist — ones that benefit from the experiences of thousands of customers and come with the assurance of being third-party provided.
The SaaS business model is alive and well.
SaaS 3.0 in Action: Insights from our portfolio
Our portfolio companies are living proof of the demand for ready-made software solutions, now supercharged by AI:
AI speech coach platform Yoodli teamed up with Toastmasters, one of the largest international speech improvement communities, to offer a customized version of its speech-training platform to Toastmasters’ hundreds of thousands of users.
Mesh, which uses AI, is partnering with large enterprise customers like Amazon and Plaid to help with business identification processes; meanwhile, Vouched helps customers like Hims and Tennessee State Bank verify identities through its AI tech.
Clarity is stepping in to combat the rise of AI-generated deepfakes, providing essential content moderation tools. The startup’s tech recently detected a deep fake audio message from President Joe Biden, even as other digital-detection software tools and officials failed to sound the alarm.
While Coca Cola might dabble in a custom-built AI art competition, its CRM backbone remains Salesforce. The scenario is similar across industries; long-standing businesses are not rushing to create in-house tax management software or identity verification solutions. Yet, these functions remain essential cogs in the machinery of modern business.
Investor confidence in the space remains strong, despite the broader pull back in venture capital funding. Essential.AI raised $56 million to automate workflows. Slope scooped up $30 million for B2B payments. And Kognitos landed $20 million to streamline back-office tasks.
To suggest that the rise of in-house AI development will eclipse SaaS is to overlook the complexity of technology development and conventional business wisdom.
As we enter the SaaS 3.0 era, expect to see an undeniable surge of its market capture.