The SaaSpocalypse. Plummeting SaaS valuations. AI replacing enterprise software.  

The AI disruption is real, but how far will it go?

Enterprise software has been the backbone of business for decades. Enterprise resource planning (ERP) systems manage finance and operations. Customer relationship management (CRM) systems track customer relationships. Business intelligence (BI) tools turn data into dashboards and reports.  

Together, they provide businesses with structured access to systems of record, engagement, and insight that are essential to getting work done.  

But that model is starting to crack.

As AI rapidly integrates into every corner of business and embeds itself into daily workflows, a larger question has emerged.

Is AI going to replace software altogether?

Consider what most CFOs must do to figure out why margins dropped last quarter. Log into multiple systems, click through dashboards, export data, and work with analysts to patch together an understanding.  

It takes time, and insights are often outdated by the time the reports are complete.  

But there’s another scenario where a CFO uses conversational AI to simply ask, “Why did margins drop in the first quarter, and what should we do to fix it?”  

Within seconds, AI agents retrieve relevant data from a cloud data platform and come back with a clear, accurate answer that includes causes, analysis, and recommended actions.  

This isn’t a vision of the future. It’s already happening in organizations that have modernized their data foundation and layered on conversational and agentic AI capabilities. That’s the key.

None of this can happen if you don’t tackle the data modernization part first.  

This glimpse into the next phase is the real threat that business- and operations-software vendors are facing.  

AI isn’t eliminating software entirely.  

It is disrupting the model that’s been a moneymaker to SaaS organizations for decades.  

AI is stripping away the core value of enterprise software (interfaces), forcing vendors to reshape and redefine where the value lies.  

Meanwhile, data and intelligence are becoming the primary sources of value.

In this article, we’ll explore what’s happening, why it matters to enterprise leaders, and what to do next.  

What Is Software, Really? An Interface to Your Data

Most software is just a layer between you and data (unless you’re playing games like Minesweeper and Solitaire or using general productivity tools like Microsoft Word and Excel).  

• Your accounting software is an interface between you and your financial data.

• Your inventory software is an interface between you and your operational data.

• ERP and CRM systems are interfaces between business users and organizational data.

These systems don’t run the business. The data does. But the systems provide a controlled way for people to input, access, and interpret the data that drives decisions.

Your ERP doesn’t run your operations. It provides a controlled way to access financial and operational data and standardize processes across the business.  

Your CRM doesn’t manage your customer relationships. It provides a framework to capture, organize, and interpret customer data.  

This model has served users well because working with raw data was historically extremely difficult. It required a deep technical understanding of underlying systems and the skills to know how to accurately query the data.  

Dashboards, reports, forms, and workflows removed that complexity, making data accessible to business users and a much broader cross-section of the organization.  

But that convenience came with tradeoffs.  

Business and operational software is typically:  

• Monolithic

• Rigid in how processes are structured

• Built around fixed user interfaces that don’t often align to how work is actually done

• Defined by the vendor’s assumptions, not unique business needs

• Expensive to customize and maintain

Because of inflexible interfaces, organizations have had to create workarounds and adapt to the software instead of the software adapting to their needs.  

Though not perfect, this was an acceptable solution because there were no alternatives.  

Now there are.  

What Has Changed? The Shift from Applications to Data

AI is shifting the focus from applications to data and from users to outcome, a transition that has been building for years.  

For example, today, data is no longer locked inside individual applications. More and more, it is centralized in cloud-based data platforms and made accessible to users across the organization.  

Data platforms like Snowflake and Databricks are already starting to eat ERP and CRM systems because data platforms are far more flexible and customizable for your business than any prepackaged software.  

The platforms have hastened this shift by providing the compute power and scalability needed to store, access, and quickly process massive amounts of data independent of the applications that generate it.  

As a result, data becomes a strategic, accessible, shared asset.  

That directly impacts the role of traditional software.  

Instead of being the primary location where data lives and is analyzed, enterprise applications become one of many inputs into a larger ecosystem.  

They are part of the backend rather than the center of interaction, especially as reporting, analytics, and cross-functional insights increasingly happen outside the ERP or CRM.  

Simultaneously, conversational AI agents are affecting how users interact with that data.

Users no longer rely on vendor-defined dashboards or reports. They just ask questions and receive answers tailored to their specific needs.  

They also save considerable time by making a single request rather than navigating multiple steps across multiple systems.  

Beyond the user experience, this shift directly affects the economics of software.  

Enterprise applications traditionally priced their licensing based on the number of users who need access. The more people using the system, the more licenses were required.  

But take away the value of providing an interface for humans to interact with the data, and the number of people who must interact with the system could decrease.  

If AI agents retrieve data, generate insights, and execute tasks on behalf of the users, this, too, could reduce the number of humans needing direct access.  

Based on that, when humans are no longer the primary users, what happens to a traditional software pricing model built on human access?

The shift is already underway. Vendors are already adjusting their pricing structure, moving away from seat licenses toward usage- or outcome-based models.

This amounts to a complete structural change in how software delivers value. And it also explains why SaaS vendors are scrambling.  

What AI Changes: From Interfaces to Intelligent Workflows

AI impacts traditional software by giving us the ability to do the following four things:

1. Eliminate Traditional Interfaces  

With AI, you no longer need to navigate systems and menus, dashboards, or reporting workflows.

Conversational AI gives us the alternative of replacing those interfaces with natural language so you can talk directly to your data and get answers instantly. AI becomes the interface.

2. Create Your Own Flexible, AI-Generated Interface

If your organization really needs a structured user experience, you can use AI to quickly generate an interface tailored to your specific needs.  

Teams no longer have to rely on rigid, vendor-defined screens. Instead, they can create user experiences that adapt to how they work, without heavy customization or long development cycles.

3. Replace Static BI Reporting

Predefined reporting from traditional software is limiting and often doesn’t fit your requirements. AI enables you to ask a question, in natural human conversation, and get insights in whatever format you need in real time. And this happens instantly.  

4. Automate Workflows Using AI Agents

Agentic AI automation fundamentally changes how work gets done. AI agents can execute repetitious tasks and work across systems. For example, AI agents:

• Process invoices

• Update records

• Route approvals

• Trigger workflows

Whereas AI assistants would just explain to users how to do this, AI agents take action and complete them autonomously.

Gartner predicts that by the end of 2026, 40% of enterprise software will include AI agents that handle specific tasks. That number is up dramatically from the previous year, when less than 5% of enterprise apps included task-specific AI agents (Gartner, Aug. 26, 2025 press release).  

The Future of Enterprise Software: Where Are We Headed?

If traditional interfaces are no longer as valuable and if we’re moving to data platforms, agentic AI, and conversational AI to redefine how work gets done, what does that mean to the industry?

Traditional software applications enable people to interact with data through software.  

The emerging model is different: people interact with data through AI. AI agents retrieve data, and then they coordinate with and act on backend systems of record (your ERP and CRM) behind the scenes. The scale of change will be significant.

In this new architecture:

• AI is the primary interaction layer.

• Data platforms are the foundation.

• Applications operate behind the scenes.  

Users:

• Ask questions in natural language.

• Generate insights instantly instead of waiting days.

• Trigger actions across systems.

ERPs and CRMs don’t disappear. They continue to handle transactions, enforce controls, and maintain compliance.  

But they’re no longer the face of your data. AI is.

Users no longer will no longer think in terms of systems, either. They’ll think in terms of outcomes.  

Instead of logging into multiple applications, they’ll describe what they want to achieve, and AI agents will orchestrate the rest.  

What’s Holding AI Back? The Data Problem

This all sounds compelling and ready to execute. So if AI is powerful enough to replace large portions of enterprise software interfaces, why hasn’t it happened already?

The reality is that AI is advancing faster than the data it depends on.  

AI is not the limitation. Data is.  

Most organizations still operate with fragmented, inconsistent data that’s spread across multiple, disparate systems:  

• ERP systems

• CRM platforms

• Marketing tools

• Finance and HR applications

• Spreadsheets

Organizations were built in functional siloes. Departments were siloed, the applications that supported them were siloed, and their data was siloed. At the time, this made sense.

However, as a result, each enterprise application today has its own structure, definitions of metrics, and assumptions. Each team operates with its own version of truth. Even within a single team, it’s often hard to locate where that source of truth exists.  

When AI is layered on top of that foundation, it still produces answers quickly. But if the underlying data is incomplete, inconsistent, or inaccurate, the results will be as flawed as the data.  

AI doesn’t eliminate data problems. It amplifies them, increasing risk across your organization.  

SaaS vendors face a related challenge. If their platforms are built on fragmented data models, their AI capabilities will struggle to deliver meaningful value.  

So it’s not enough for SaaS vendors to layer conversational AI or AI agents on top of weak data foundations. The features might look impressive, but they won’t be able to deliver real value.  

For AI to deliver accurate, meaningful results at any organization, it requires:

• Removal of data silos

• A modern data foundation and architecture

• High-quality, reliable data

• Consistent definitions across systems

• Strong governance frameworks and ownership

• AI integration across the enterprise

It boils down to a strategic challenge, not just a technical one.  

Cloud data platforms like Snowflake and Databricks provide the infrastructure to unify and scale data, but they don’t solve the AI readiness problem on their own.

There is a lot of hype around AI, but the need for an AI-ready data foundation is a real requirement.  

If you want to use AI to accelerate work, deliver accurate answers, and uncover insights at scale, there’s only one path forward: invest in a modern data foundation that makes your data usable, trustworthy, connected, and accessible.  

Organizations that ignore this advice will struggle to move beyond experimentation and risk falling behind.

What This Means for Software and the SaaS Model

The role of software is being redefined as AI changes how users interact with data and as data becomes the primary constraint to enabling AI to find results.  

The traditional model of business software, with dashboards, workflows, and a structured user experience, is weakening.  

When conversational AI becomes the main way for users to engage with systems, and when AI agents begin executing workflows, the interface will no longer be the center of value.  

Value will shift to:

• The quality and accessibility of data

• The intelligence layered on top

• The ability to execute processes efficiently

And the tech stack will evolve accordingly:  

• Conversational AI as an interface

• AI agents to handle workflows

• Data platforms to enable scale and speed

• SaaS applications operating as backend services.  

Software vendors are responding by embedding AI capabilities (such as copilots) into their platforms. Oracle, SAP, and Microsoft are already making huge investments in AI-driven governance features to stay competitive.

Forrester predicted that 50% of enterprise ERP vendors will introduce autonomous, AI-driven governance modules in 2026 (Forrester blog: “Predictions 2026: AI Agents, Changing Business Models, and Workplace Culture Impact Enterprise Software,” Nov. 2025).  

Vendors are also rethinking how they package and price their offerings, shifting away from seat-based pricing toward usage- or outcome-based price structures.  

IDC expects that seat-based pricing will be gone by 2028, with 70% of vendors replacing those pricing models based on other factors like outcomes or consumption (IDC FutureScape: Worldwide Agentic AI 2026 Predictions, Oct. 2025).  

Enterprise software as we know it is changing forever. It will become less visible and more dependent on the data and intelligence behind it.  

What to Do Next: Practical Takeaways for Enterprise Leaders

If you’re wondering what to do with all this information, here are the actions you should prioritize.  

1. Shift your focus to data rather than applications. AI is only as effective as the data it uses. Siloed, inaccurate, inconsistent, or poorly governed data will produce poor results.

2. Make data readiness a business priority, not just a technical one. Invest heavily in improving data quality, aligning definitions, and establishing governance across the organization. A data strategy with a roadmap helps align teams, identify gaps, and prioritize what to fix first.

3. Rethink how work gets done. Don’t just optimize existing workflows. Challenge whether they should exist at all and create more practical workflows organization-wide. Identify high-volume, repetitive work that you can simplify or automate using AI agents.  

4. Prepare for a shift in how software will deliver value. User-based pricing models and rigid, predefined workflows are giving way to more flexible, outcome-driven approaches. Evaluate what will be most effective and cost-efficient as this shift accelerates.

Conclusion

The question isn’t whether AI will change enterprise software. That change is already happening.  

The real question is whether your organization is ready for a world where:

• Conversational AI becomes the primary way users interact with data.

• AI agents execute work.

• Software fades into the background, almost invisible.

In that world, the tools you use today won’t give you a competitive edge. If you want AI to deliver real value, the priority now is to get your data foundation in place.

Then you can build intelligence that actually drives your business forward.