We’ve been thinking about how much AI could pose a genuine threat to parts of the SaaS market.
Not because every company will suddenly start building all its own technology but because the economics and practicalities of building software are beginning to change.
For years, the accepted route has been relatively straightforward. A company identifies that it needs some new technology, runs an RFP, evaluates a long list of features and ultimately signs a lengthy contract with a SaaS provider.
It may then use 30–40% of the product, while continuing to pay for 100% of it.
This is not necessarily the fault of the provider. SaaS businesses are designed to build products that can serve multiple customers. They can offer some configuration and perhaps a small amount of customisation, but they cannot build a genuinely bespoke product for every client.
This is where I think many RFP processes have become discombobulated.
Companies produce enormous lists of features and requirements, but often spend far less time clearly articulating the problem they are actually trying to solve.
The process becomes:
“Can your platform do these 150 things?”
Rather than:
“This is the operational problem. This is how it affects our business. What is the simplest and most effective way to solve it?”
Historically there was little point designing a procurement process around the perfect solution when the available products could only be configured within relatively fixed boundaries. But AI, combined with modern development tools and infrastructure, could begin to change that dynamic.
We may be moving towards a model where companies do not always need to buy a large, established product and adapt their processes around it.Instead, smaller providers could build focused products around a company’s actual problem, workflows, data and existing technology.
This is ultimately moving from “software as a service” towards “service as software”. The provider is not simply selling access to an existing platform. It is providing the expertise and technology required to build and operate a solution.
The commercial model might include:
• The initial product build
• Integration with the company’s existing systems
• Ongoing maintenance and development
• Hosting and infrastructure
• Governance and information security
• Human expertise and oversight
Over time, the client could choose to continue working with the provider as the product evolves. Alternatively, it could take ownership of the code and move the product into its own infrastructure.That feels materially different from paying a recurring licence indefinitely for a broad platform that only partially addresses the original need.
None of this means SaaS is going away.
There are plenty of standardised business problems where buying a mature, proven platform will remain the most sensible and cost-effective option.Building something simply because it is now possible to build it is not a strategy either. Without the right expertise, architecture, governance, security and ongoing ownership, companies may simply replace underused SaaS products with a collection of poorly maintained internal tools. But for more specialist or strategically important problems, the balance between buying and building is changing.
The biggest constraint may no longer be whether the technology can be developed. It may be whether procurement, legal, information security and RFP processes are ready to evaluate a solution that does not already exist. Perhaps technology procurement needs to start with a much simpler question:
What problem are we actually trying to solve?
At Alligator Solutions, we have been helping brands navigate this shift by first getting clear on the problem they are trying to solve, the data and processes that underpin it, and where existing technology is creating unnecessary cost or complexity. From there, we help lay the groundwork for a more flexible “service as software” approach, whether that means building a focused solution around a specific business need, improving how existing platforms are used, or creating a clearer path towards greater ownership and control. Done well, this is not simply about reducing technology spend. It is about creating products and capabilities that are better aligned to the business, easier to evolve and more capable of supporting future growth.