Recent stock market losses among software providers have prompted some analysts to predict a coming “SaaS-pocalypse” as software companies are threatened by artificial intelligence that can write code and build software quickly and cheaply.
The doomsayers may be premature, however. While AI undoubtedly has the ability to supplant some of those firms, it also presents a powerful tool that can strengthen them.
By incorporating agentic AI into their products, software developers and SaaS providers can improve their own products and offer a compelling alternative to financial institutions who are considering ending their contracts to create in-house software-building AI agents.Carrie Osman, chief executive and founder of Cruxy, which advises private equity funds and investors on their technology and data needs, including software pricing, packaging and product value, likens the present situation for software builders to that of Kodak, the camera film company that turned down the opportunity to invest in digital photography.
“We face that Kodak moment where software firms have the opportunity to say, ‘we believe AI is coming, we believe it’s going to be revolutionary and we are going to start ruthlessly moving our software to embed as much AI and as much agentic as we can,” Osman told Data Management Insight.
Standing Still Is not an Option
Software providers are the “masters of their own destiny” and have the ability to adapt, she said. Having built the business model, they can change it, she argued. They have the software and they can change that by embedding AI. They’ve also built the integrations and they have an element of control over the data that will be used.So there’s no excuse for standing still.
“Blaming public markets isn’t really effective – you can blame them for a day and probably be a bit sad about your valuation but then you’ve got to stand back and say right what are we doing?”
The suitability of technology and data services in private markets has become a key issue for financial institutions and other participants in the space, as they allocate more and more capital. A recent report from Aviva shows that a quarter of all financial institutions surveyed had dedicated up to 20 per cent of their portfolio to private markets last year. That was up from a tenth in 2023.
Osman’s view is similar to those of SAP chief technology officer Philipp Herzig and Oracle chief executive Mike Sicilia. Both said recently that AI’s test will be the making of their firms as they embrace the technology to boost their own.
Osman said she is already seeing AI agents performing the tasks provided in some SaaS offerings and believes the technology can definitely challenge the viability of some software makers.
Seat-Based Pricing Model Poses a Hurdle
Preventing them from meeting the challenge, she warned, is their pricing model. Most SaaS products are charged on a per-seat basis.
SaaS providers adopted the model, in which they charge for licences issued to individuals in an organisation, long ago. No matter the job of the person occupying the seat that holds the licence, the cost is usually the same. The model has endured because it is simple.
But its rigidity prevents the customisation and specialisation that agentic AI can offer. And the model overlooks a simple fact: it doesn’t reflect the value derived from utilisation of the software. Whether an end user leverages the entire functionality of the software, or just dips in every now and then, the price of doing so remains the same.“Having one-size-fits all and assuming that every user is equal is inherently flawed,” she added. “That doesn’t really conceptually make sense. It’s not tied to value.”
Institutions are realising this, she said.
Software Buyers Are Ready to Change
Research carried out for Cruxy has found that 20 per cent to 40 per cent of software buyers plan to cut their seat counts as they see the value of agentic AI-led automation.
“Seat-based models limit expansion of SaaS or software businesses to the seat,” she said. “They make the unit of value the individual using the software. It was considered right because it was simple for the customer to understand. But was it ever necessarily completely the right way to price? Was it ever truly aligned to value?”
Osman illustrated the shortcomings of the pricing structure with the example of a loan origination system. The commercial loan officer will bring a different degree of value to a retail loan officer or a mortgage originator, an yet they might all be paying the same amount for the same software.
The model is also flawed from a business growth perspective for the software developers, she argued.
If the software brings a 10 per cent cost saving to a team utilising 200 seats then, logically, the client could reduce its seat spend by 20. By extension, then, the more effective the software becomes, the fewer seats will be needed and the less income the SaaS provider will collect.
She concedes that the model is still useful for some use cases.
“For some organisations and businesses that use seat-based pricing the fact it’s so ingrained in the way they operate perhaps means it still might be the right thing to do – an HR technology system, for example, which is truly tied to the number of employees could find an argument that it might be right to stick,” Osman said.
“But the danger is that in the world of agentic AI, the product is really serviced as a software and tying yourself to the human being as the only driver of production obviously feels quite limited.”
Change is Already Happening
Osman sees AI agents already making some software obsolete. The risks of agentic usurpation varies across institutions and within their departments. Tasks that have been all but commoditised, such as cyber risk management, are more likely to be managed by AI agents than those that give organisations a competitive edge, such as front-office and trading workflows.
But she sees software companies realising the challenge the technology is throwing at them and they are beginning to incorporate agents into their own offerings.
This change is more apparent among US providers because, Osman says, they have recognised that change is the best way for SaaS firms to ride out the SaaS-pocalypse.
“Many of the software businesses ultimately forget a little bit that they’re the masters of their own destiny,” she said.
The fact that not all providers will change means there is likely to be substantial M&A activity within the software space in the near future, she predicts.
“There are these slices of companies that have been devalued in the SaaS-pocalypse that have very unique specialist data sets, for example; treating them all as equal is flawed. I expect we will see M&A as software businesses start to need a lot more of this data in their core offering,” she said.
Subscribe to our newsletter
