This month we take on the oft-repeated assertion that “SaaS is dead.” We chatted with our CEO, Chief Customer Strategy Officer, Chief Product Officer and lead Industry Analyst to go beneath the headlines and sweeping statements for their views on the topic.
There has been no shortage of chatter lately about the so-called “death of SaaS.” The argument is familiar: AI is changing how software is built, how it’s consumed and what businesses expect from it. If intelligent agents can execute tasks and surface answers on demand, why keep paying for layers of software in between?
It’s a compelling narrative. But it’s also an incomplete one.
What came through clearly in our conversation is that this isn’t really about the end of SaaS. It’s about a shift in what B2B software needs to deliver and a clearer distinction between tools that are easy to replace and platforms that are fundamental to how firms operate.
Unpacking the “SaaS is dead” narrative
The narrative itself isn’t coming from nowhere. AI is starting to change how users interact with software, while at the same time SaaS as a category is facing more scrutiny, from investors and buyers alike. Saas market growth has normalized from the hypergrowth area (which shows up in public company growth rates and valuations), expectations have tightened and cost versus value is being looked at more closely.
That combination has led to the view from many quarters that software is becoming less and less necessary.
But as Leigh explains, that conclusion is too general: “The headlines are collapsing nuance. What’s actually under pressure is generic, thin SaaS – tools with less robust workflows and limited integration. That’s not the same as saying SaaS as a model is going away.” So the more useful lens is not whether SaaS is ‘alive’ or ‘dead,’ but which types of software are being challenged, and why.
That combination has led to the view from many quarters that software is becoming less and less necessary.
But as Leigh explains, that conclusion is too general: “The headlines are collapsing nuance. What’s actually under pressure is generic, thin SaaS – tools with less robust workflows and limited integration. That’s not the same as saying SaaS as a model is going away.” So the more useful lens is not whether SaaS is ‘alive’ or ‘dead,’ but which types of software are being challenged, and why.
What’s actually changing
At the center of this shift is a change in how software is accessed. Users are moving away from navigating systems towards asking for outcomes; querying, prompting and triggering actions through an AI layer. That’s a meaningful change in experience. What it doesn’t change is the need for structure underneath. Nadine brings it back to first principles: “AI only works if the underlying data and workflows are sound. If they’re not, you don’t remove complexity, you just accelerate it.”
That’s where the conversation becomes more grounded. The requirement for structured data, defined workflows and auditability doesn’t disappear. If anything, it becomes more important as automation increases. So the shift isn’t about software being replaced. It’s about where value sits and what software is expected to do.
That’s where the conversation becomes more grounded. The requirement for structured data, defined workflows and auditability doesn’t disappear. If anything, it becomes more important as automation increases. So the shift isn’t about software being replaced. It’s about where value sits and what software is expected to do.
Where the pressure actually sits
That distinction shows up clearly when you look at where pressure is being felt. The tools most exposed are those that sit lightly on workflows; horizontal, loosely integrated and not essential to how work actually gets done.
As Ian puts it: “If a product doesn’t really own the workflow or the data, it’s much easier to route around it. That’s where AI has the biggest impact.”
By contrast, platforms that act as systems of record play a different role. They hold the data firms rely on to operate.
Leigh frames that distinction directly: “The platforms that tend to hold up are the ones that hold contractual, financial and regulatory truth. That’s not something you can replace with an agent.”
In market data management, that difference is particularly pronounced. The combination of fragmented sources, licensing complexity and vendor policies means firms need systems that can interpret and structure that commercial landscape. A big part of that is taxonomy; having a consistent way to classify datasets, services and usage. In practice, the same data can be described, consumed and licensed in different ways across the organization. Without agreed nomenclature and expertise how to orchestrate and map data points in support of business outcomes it becomes difficult to track, interpret or govern, and that’s where both reporting and automation start to break down. As Suzanne explains: “This isn’t just data tracking. It’s about interpreting rights, obligations and usage in a way that stands up commercially and from a compliance perspective.” AI doesn’t remove that requirement. It increases the need for it to be handled well.
As Ian puts it: “If a product doesn’t really own the workflow or the data, it’s much easier to route around it. That’s where AI has the biggest impact.”
By contrast, platforms that act as systems of record play a different role. They hold the data firms rely on to operate.
Leigh frames that distinction directly: “The platforms that tend to hold up are the ones that hold contractual, financial and regulatory truth. That’s not something you can replace with an agent.”
In market data management, that difference is particularly pronounced. The combination of fragmented sources, licensing complexity and vendor policies means firms need systems that can interpret and structure that commercial landscape. A big part of that is taxonomy; having a consistent way to classify datasets, services and usage. In practice, the same data can be described, consumed and licensed in different ways across the organization. Without agreed nomenclature and expertise how to orchestrate and map data points in support of business outcomes it becomes difficult to track, interpret or govern, and that’s where both reporting and automation start to break down. As Suzanne explains: “This isn’t just data tracking. It’s about interpreting rights, obligations and usage in a way that stands up commercially and from a compliance perspective.” AI doesn’t remove that requirement. It increases the need for it to be handled well.
AI and SaaS: a shift in roles
Where AI does have a major impact is in how users interact with software. The interface is becoming less central. Less time navigating systems, more time querying outcomes. But that doesn’t remove the platform underneath. It changes its role.
Suzanne captures that shift: “AI changes the interface, not the need for the platform. The platform becomes the structured layer that everything else depends on.”
Increasingly, that interaction is not just a user querying a system, but software acting on behalf of the user. Agentic capabilities are starting to blend software with service-layer expertise; combining automation with domain knowledge to execute tasks, interpret outputs and drive decisions.
That doesn’t remove the need for platforms. It raises the bar for them. The underlying systems still need to provide the structure, data integrity and governance that those agentic layers rely on to operate effectively.
That reframes SaaS as infrastructure rather than destination. The value sits in the data, workflows and logic, even if interaction happens elsewhere. And that has implications for what makes a platform defensible. At the same time, expectations are moving. Platforms need to be faster, more open and able to evolve alongside how their users want to work.
Suzanne captures that shift: “AI changes the interface, not the need for the platform. The platform becomes the structured layer that everything else depends on.”
Increasingly, that interaction is not just a user querying a system, but software acting on behalf of the user. Agentic capabilities are starting to blend software with service-layer expertise; combining automation with domain knowledge to execute tasks, interpret outputs and drive decisions.
That doesn’t remove the need for platforms. It raises the bar for them. The underlying systems still need to provide the structure, data integrity and governance that those agentic layers rely on to operate effectively.
That reframes SaaS as infrastructure rather than destination. The value sits in the data, workflows and logic, even if interaction happens elsewhere. And that has implications for what makes a platform defensible. At the same time, expectations are moving. Platforms need to be faster, more open and able to evolve alongside how their users want to work.
What this means for market data teams
For market data teams, this shift is already translating into more practical questions. The focus is moving away from simply rationalizing cost, and towards understanding what is genuinely critical.
Ian frames it directly: “The question isn’t ‘do we need this tool?’ It’s ‘does this system hold something authoritative, or is it just an interface?’”
That distinction becomes especially important as firms look to simplify their stack. Alongside that, data quality moves into sharper focus.
As Nadine points out: “If your data is fragmented or inconsistent, AI will expose that very quickly. It doesn’t fix weak foundations, it amplifies them.”
There is also a clear shift in expectations from the business. Leigh highlights what’s coming through most consistently: “Customers want speed and visibility. Faster answers and a clear, real-time view of what’s happening across their data estate.” That creates a different kind of opportunity. Less time spent on administration and more on insight and decision making.\
Ian frames it directly: “The question isn’t ‘do we need this tool?’ It’s ‘does this system hold something authoritative, or is it just an interface?’”
That distinction becomes especially important as firms look to simplify their stack. Alongside that, data quality moves into sharper focus.
As Nadine points out: “If your data is fragmented or inconsistent, AI will expose that very quickly. It doesn’t fix weak foundations, it amplifies them.”
There is also a clear shift in expectations from the business. Leigh highlights what’s coming through most consistently: “Customers want speed and visibility. Faster answers and a clear, real-time view of what’s happening across their data estate.” That creates a different kind of opportunity. Less time spent on administration and more on insight and decision making.\
Market reality: pressure is real, but selective
The pressure on SaaS is real. Buyers are more selective. Expectations are higher. The gap between stronger and weaker propositions is becoming more visible. But it’s not uniform. As Suzanne summarizes: “This isn’t about SaaS declining across the board. It’s a reset in what’s considered valuable.”
Some tools will struggle. Others will become more central.
Some tools will struggle. Others will become more central.
What the narrative gets wrong
The limitation of the “SaaS is dead” narrative is that it treats all software as if it plays the same role. In reality, that role is diverging. Some software becomes easier to bypass. Some becomes more embedded. Some becomes more important.
Leigh captures that direction: “The future isn’t about broader software. It’s about platforms that are trusted, embedded and able to evolve quickly. That’s where value concentrates.”
That’s the shift playing out. Not whether SaaS survives. But which platforms remain fundamental to how firms operate.
Leigh captures that direction: “The future isn’t about broader software. It’s about platforms that are trusted, embedded and able to evolve quickly. That’s where value concentrates.”
That’s the shift playing out. Not whether SaaS survives. But which platforms remain fundamental to how firms operate.
Thanks for taking time to read Market Data Matters! Don't forget to subscribe if you haven't yet.
