There's SaaS products for every enterprise form and function: Treasury SaaS for Finance teams, Analytics SaaS for data teams (an entire Modern Data Stack), Headless E-commerce SaaS for marketing teams, and more. These department-specific tools and products all have two things in common:
Composability – Headless/API-first products that integrate easily into existing stacks. Best-of-breed means that even the most platformized SaaS need to play well with others.
Data sovereignty – Data is safer in the cloud than it is on premise. But customers want data to be safe on their cloud. Not only for composability with other tools, but to reduce lock-in and often egress fees. This isn't incompatible with closed-source software, many new companies are pivoting to Bring-your-own-database (BYODB).
The important and unsolved question is who manages these software stacks embedded in different departments? It used to be IT. But now these are real software stacks that needs to be versioned, tested, deployed, and managed (even if you're using a managed service). Composability requires engineers to do the plumbing. Data sovereignty comes with a cost of ownership.
Workers are becoming more literate with technology.
The answer so far has been to upskill the analytical-but-not-software-engineer workers (see The Modern Data Stack) to learn how to use version control, write tests and documentation, and clean data. If you squint, a data engineer looks a bit like a DevOps engineer. Everyone is pushing around different flavors of configuration and YAML.
Maybe platform teams will become more important in engineering organizations. Managing Kubernetes or other generic platform endpoints for self-service within marketing, sales, finance, and data. But organizationally managing software across the entire company is going to get even harder.