I recently wrote about the impending SaaS crisis, driven by companies’ newfound ability to use AI to build software they used to have to buy. I predicted this phenomenon would make it even harder for SaaS vendors to drive growth, and that elite SaaS margins would fall from the mid-70s to the mid-60s as companies leaned more into their data and AI.
As SaaS vendors watch companies like Klara publicly share their evolution with vendors like SalesForce, it feels like a “the sky is falling” moment. I don’t think the whole sky is going to fall, or that SaaS vendors are powerless to defend themselves. But in the age of AI, SaaS vendors will have to fight harder than ever to defend their value.
But that’s only half the story, and it depends on the other half coming true: that companies actually start building software to run their operations in-house again.
That’s where IT departments have an opportunity to increase their value.
For IT departments, the AI story presents a bit of a back-to-the-future moment. They could retake a role they used to play in the years before the SaaS revolution — that is, assuming they avoid one major pitfall.
For IT departments, the AI story presents a bit of a back-to-the-future moment. They could retake a role they used to play in the years before the SaaS revolution — that is, assuming they avoid one major pitfall.
How AI Could Transform IT Departments
In the old days (i.e., the mid-to-late 20th century), IT departments built software to run internal business processes. The goal was the same then as it is now: to automate complex and/or tedious tasks. Companies identified processes that were slowing them down, and in-house “computer people” (as IT pros were commonly and perhaps pejoratively referred to in the early days of programming) set about building applications to solve them.
The emergence of commercial software and integrated enterprise solutions (like SAP and Oracle in the 1990s — not surprisingly, Larry Ellison became the richest person in the world recently, much of his fortune built in those days) dramatically changed this role. Companies all struggled with a similar set of problems; so, all could buy the same software to solve it. IT departments no longer needed to reinvent the wheel, and their role shifted from building novel software to implementing packaged software.
Then came SaaS and the cloud.
Salesforce replaced Siebel.
ServiceNow displaced Remedy.
AWS disrupted datacenters.
With this transition, IT’s primary directive became governance, and IT departments again needed to recreate their sense of identity. IT started as system builders, evolved into software implementers, and wound up becoming SaaS brokers and managing and securing access points to systems and data.
But now there’s AI.
With AI, companies have the power to build certain in-house software applications that they used to have to provision externally. They can’t build everything they get from a SaaS vendor — not yet, anyway — but they can start chipping away at what has likely become a sprawling software footprint for the world’s largest enterprises.
And that duty could fall squarely on IT departments.
IT departments have an in-depth understanding of where internal capabilities end and external help is required. They have the technical expertise necessary not just to build additional tooling but integrate it with existing processes. If IT capitalizes on the opportunities that co-pilots and vibe coding bring, they could return to their roots of developing software for internal business purposes.
But there’s a catch: This assumes the cost to build and manage all that AI tooling does not spiral out of control.
How To Build Cost-Efficient In-House Software
Building software in-house only makes sense if it will 1) get up and running faster than external SaaS platforms, 2) align more tightly with the business, and 3) cost less. While in-house software promises to be more efficient, it won’t be free. Every engineering decision is still a buying decision — every new cloud/AI instance still costs money — and you’ll still have to manage the costs and headaches of bespoke in-house software.
You’ll still have to monitor software for sudden cost spikes, allocate software costs back to the business, and ensure healthy unit economics. You’ll still have to understand how its cost aligns with corporate profitability and budgets, and figure out how to maximize its margins. You’ll still have to get engineers engaged in the practice of cloud financial management (FinOps) to make sure cost efficiency starts on day one.
CloudZero’s fundamental value is shifting the cloud and AI conversation from, “What did that cost?” to, “Was it worth it?” We put every penny you spend on cloud and AI services in a business context, enabling impactful collaboration between technical and financial FinOps stakeholders. We manage cloud and AI costs for the world’s most sophisticated cloud spenders, and recently became the first cloud cost management platform to integrate directly with Anthropic in addition to OpenAI.
What’s old is new again; the AI revolution could put IT organizations squarely at the center of companies’ innovation goals. Proactively managing the costs of those innovations will set organizations up for long-term viability and profitability.
The Cloud Cost Playbook
The step-by-step guide to cost maturity
