Will AI wipe out the SaaS economy?

Are We Witnessing the End of Traditional SaaS?

When Klarna announced its decision to discontinue both Salesforce and Workday implementations, it wasn't just another cost-cutting measure in a tech company. Instead, it marked a game-changing moment that could reshape the way organizations think about software solutions. This courageous move, coupled with the recent 20% drop in Salesforce's stock following its Q1 2025 earnings announcement, suggests that we are about to witness a meaningful shift in enterprise software.

The shakeup of this change has been building over the past few months. While market analysts initially attributed Salesforce's challenges to economic headwinds, a closer examination reveals something more profound: the emergence of AI as a disruptive force in enterprise software. The traditional SaaS model that has ruled enterprise software for the past two decades is facing its first serious challenge since it supplanted on-premise solutions in the early 2000s.

To understand the scale of this change, we need to look at Klarna's transformation. Their AI implementation isn't just a minor upgrade - it's a complete reimagining of how enterprise software works. Within the first month, their AI assistant was handling two-thirds of all customer service interactions, achieving - according to the company - the same level of satisfaction as human agents and cutting processing time from 11 minutes to under 2 minutes. What is perhaps most impressive is that the system works 24/7 in more than 35 languages, demonstrating capabilities that traditional SaaS solutions simply cannot offer.

The economic drivers for this change are convincing. Traditional SaaS providers have long worked with premium pricing models that are billed by user seats or endpoints. This model has worked well for providers, but it is increasingly at odds with the value proposition of modern AI solutions. Companies are finding that they can develop customized, AI-powered alternatives that if well designed and implemented might not only cost less, but often offer superior functionality. This realization is leading to a major shift in enterprise IT spending, with budgets increasingly flowing from SaaS subscriptions to AI initiatives.

Major Changes go Hand in Hand with new Challenges

However, this change goes far beyond pure cost savings. We are seeing a profound change in the way software fulfills business needs. Traditional SaaS platforms are based on structured databases with rigid workflows. In contrast, next-gen AI platforms can [always based on the assumption that the required data is available in the necessary quality] process unstructured, multimodal data - from text and voice to video and images. They can gain insights from across the organization without the need for manual data entry and can adapt to changing business needs in real time.

This change is also leading to changes in the pricing and valuation of software. The traditional per-seat model is giving way to outcome-based pricing structures that better align costs with the value delivered. This new paradigm could transform the way companies think about their software investments and how vendors structure their offerings.

However, not everyone is convinced that this transition will be smooth or inevitable. Critics raise valid concerns about the challenges of maintaining custom solutions, the complexity of state consistency and integrations, and whether in-house development is the best use of capital. They point out that internal tools have rarely developed into successful products in the past. These are valid concerns that cannot be easily dismissed.

For business leaders, this transition brings both opportunities and challenges. The potential for significant cost savings and improved functionality must be weighed against the added complexity of creating and maintaining customized solutions. IT leaders are seeing their role expand from license negotiation to solution architecture, requiring new expertise in AI implementation and integration.

SaaS providers are not standing still either

Many are rapidly integrating AI capabilities into their platforms and trying to stay ahead of this change. The question is whether they can innovate fast enough while making the transition from their traditional pricing models to something that better reflects the new reality of AI-powered solutions.

Salesforce may have had its “ChatGPT moment” in enterprise software development with AI agents. Salesforce has unveiled 'Agentforce' at Dreamforce 2024 - this not only signals a new product launch, but possibly a major shift in the way companies interact with and derive value from enterprise software. 'Agentforce' aims to enable users to create autonomous AI agents capable of performing complex tasks across the Salesforce ecosystem with minimal human intervention.

Salesforce CEO Marc Benioff's enthusiasm for AI agents is not mere management hyperbole, but a carefully calculated response to a persistent challenge in enterprise software: the gap between the promise of AI and its practical implementation. While chatbots and AI assistants are now commonplace, they have largely failed to deliver the transformative automation originally promised [see Fig.2: the Gartner? Hype Cycle? for Artificial Intelligence, in whose 2024 representation Generative AI has already passed the peak and is moving downwards into the trough of disillusionment].

First-Mover Advantage and Ecosystem Strength

The timing of this launch is significant. Just as Microsoft and OpenAI's partnership in early 2023 triggered a “code red” reaction from competitors, Salesforce's early lead in AI agents could trigger a similar reaction across the enterprise software landscape. Microsoft and ServiceNow have already announced their own agent initiatives, suggesting that the industry is indeed moving in this direction.

What sets Agentforce apart is its approach to implementation. Instead of offering a “DIY AI” solution, Salesforce has created a closed ecosystem in which companies can deploy agents while maintaining control over their data. This addresses one of the main problems with the introduction of AI: Data security and governance. By keeping company data within its walled garden, Salesforce offers a compelling alternative to solutions that, as Benioff noted - as a side blow to Klarna CEO Sebastian Siemiatkowski - run the risk of “spreading data all over the floor”.

What is more interesting is the change in the pricing model. Instead of pricing future agent models in the traditional SaaS license per seat approach, Salesforce is moving to a consumption-based pricing model for the first time. Billing according to actual performance instead of the flat-rate calculation of a provision is more in line with the actual value delivered to customers.

It can be assumed that other SaaS providers will adopt this approach - at least for parts of their offerings - to avoid losing customers.

Salesforce's early advantage in the AI agent space could prove as significant as Microsoft's early lead in generative AI. The company's comprehensive ecosystem, combined with its focus on trust and ethical AI implementation, positions it uniquely to shape the next phase of enterprise software evolution.

Side Note

Salesforce's vision isn't about completely replacing human workers with AI agents. As evidenced by Benioff's response to Klarna's decision to replace traditional enterprise software with AI solutions, Salesforce sees agents as augmenting rather than replacing human capabilities. The company emphasizes that complex business processes still require human engagement, with AI agents serving as powerful enablers rather than complete replacements.

This perspective is particularly relevant given the ongoing debate about the role of AI in the workplace. A balanced approach could prove vital for companies looking to modernize their operations without sacrificing the human element, which remains essential to their business success.

Disruption will change SaaS-Providers at their Core

The change goes far beyond pricing models. Today's SaaS companies derive their competitive advantage from expertise, technical talent and a deep understanding of specific business problems. AI could democratize software development to such an extent that the traditional trench warfare of SaaS companies - data gravity, integrations and user inertia - may need to be reimagined. There are already early signs of this disruption: companies like Klarna have made headlines by replacing traditional SaaS systems with AI-powered alternatives.

However, this disruption is also creating unprecedented opportunities. The market-leading SaaS companies are larger and more efficient than ever before. The median value of the top 10 companies is over $40 billion - a dramatic increase from ten years ago, when no publicly traded SaaS company was valued at more than $50 billion. The companies that successfully integrate AI into their core offerings while maintaining enterprise-grade reliability and security could see a dramatic expansion of their total addressable market.

The Evolution of SaaS in an AI-First World

Looking ahead, it is unlikely that traditional SaaS will disappear completely. More likely is a new ecosystem in which AI-powered solutions increasingly compete with and complement established SaaS platforms.

The integration of AI into SaaS is not just an incremental improvement, but represents a fundamental shift in the way enterprise software is created, delivered and used. While some predict the death of traditional SaaS models, the reality is more nuanced. AI does not mean the end of SaaS, in fact it will make the industry even more valuable and transformative.

The figures speak for themselves: global sales of AI software are expected to rise from USD 9.5 billion in 2018 to USD 118.6 billion in 2025. This growth is not happening in isolation - it is changing the basis of SaaS delivery models. Traditional seat-based licensing could evolve into consumption-based models where AI agents, rather than human users, are the primary unit of value. This shift can already be seen with companies such as Salesforce introducing new pricing models for their AI agent services.

The current shift towards AI-driven solutions is something of a history repeating itself, as SaaS displaced on-premise software two decades ago. The parallels are striking: a new technological paradigm that enables better functionality at a lower cost, initially met with skepticism but gradually becoming inevitable. The question is not whether this transition will happen, but how quickly and dramatically it will take place.

The future likely belongs to those who can combine the best of both worlds: the reliability and scalability of traditional SaaS with the transformative power of AI agents. This hybrid approach could create hyper-personalized software experiences while maintaining the robust infrastructure that businesses need. Over time, the distinction between SaaS and AI is likely to blur as artificial intelligence becomes as fundamental to software as cloud delivery is today.

What is clear is that we are not witnessing the death of SaaS, but its evolution into something even more powerful and ubiquitous. The companies that successfully navigate this transition - whether established players like Adobe, Google, Microsoft, Salesforce, Siemens or innovative newcomers - will help shape an era in which enterprise software becomes smarter, more automated and more valuable than ever before.

Sources

• HFS Research, 06/2024 | Generative AI eats SaaS
• Forbes, 06/2024 | Salesforce Stock Dropped More Than 20% In One Day, What’s Next?
• a16z, 07/2024 | “Death of a Salesforce” - Why AI Will Transform the Next Generation of Sales Tech
• Meritech 09/2024 | SaaS Isn’t Dead (Yet) and AI Could Make it Bigger
• Inc 09/2024 | Klarna Plans to 'Shut Down SaaS Providers' and Replace Them With Internally Built AI.