Service through software, and not Software as service

It is no secret that the enterprise software market is far from perfect. Every product manager believes their product fits all customer needs, while every enterprise swears by its unique processes, so much so that it’s almost amusing for them to think anyone else could claim to know it.

This gap between belief and expectation is today worth $1.7 trillion USD and is growing at 7%. For every dollar of software sold, the service opportunity is about four times greater, covering implementation, operation, customization, and a inhouse IT support team. That's assuming the implementation goes as planned. In reality, 90% of projects miss either timelines or budgets, which drives up the total cost of ownership (TCO). That’s certainly not good news.

So, what’s stopping enterprises from building solutions themselves, or even better, software product companies from offering “service through software” instead of software as a service?

Why not offer a base platform, commit to customization, and provide ongoing support to better align with enterprise needs? Most large data centers today run on commodity hardware and open-source infrastructure stacks. Is it time to rethink application software with similar standards—commoditized platforms customized by domain experts to meet specific needs?

With a “service through software” mindset, the focus shifts to outcomes rather than the tools themselves. Towards smaller engineering teams larger delivery teams with domain experts and customer engagement teams.

More than ever, AI adoption should make this smoother than in the past. AI can improve coding efficiency by around 30-50%, significantly reducing manual coding time. It can reduce understanding existing code base by 50%. It can also increase efficiency in writing unit tests by approximately 50-60%. On documentation, AI tools can assist in automating about 80-90% of structured documentation tasks. Unlike any time in the past, now may be the right moment to reconsider using a low-cost base product and customizing the stack.

The high cost of subscriptions and cloud services is prompting enterprises to reconsider the value they are getting from the SaaS stack. This could signal a natural transition.

It would be unwise to think that it is the end of software products. A software product still provides stability, scalability, security, ongoing support, and an industry standard that may not be feasible for all custom developments. A custom build might also not work in areas like security where threat sensitivity and cost might be a high wall to cross.

However, it’s not unrealistic to consider that the time may have come to reduce premiums for complex software platform stacks and move towards companies that treat their base offerings as commodities and "service delivered" as the product.

The only factor potentially slowing this transition might be the inertia to change. But as enterprises increasingly look to optimize costs and align technology more closely with business outcomes, there is momentum to move towards real service-oriented future. The evolution of AI should push the industry toward a new era where software becomes the foundation, and service delivery defines success.

Dr. Pandian Angaiyan Partha Mukherjee Raj Lakkamraju Marco van Hurne Robotics Artifical Intelligance Sathish Elumalai Nitin Kumar Jayakrishnan Sasidharan Dan Nguyen-Huu James Cham Sara Choi

#AI #GenAI #SaaS #ServiceThroughSoftware