About this series - The best business insights are happening in the room right now. This series takes the real conversations we're having with clients and hands them to Bertie, our AI host, to break down — no fluff, no filler, just the topics that matter delivered in a format built for busy people.
The build vs. buy question is a classic topic in enterprise tech. It comes down to whether you create your own solutions or use those from vendors. Right now, two major changes are happening: SaaS vendors are gaining more control over customers, and AI is making it much easier and cheaper to build custom software. Each trend matters on its own, but together, they are forcing everyone to rethink long-held assumptions.
Why Are Enterprises Questioning Their SaaS Investments?
For years, SaaS offered significant value by providing access to robust software without the burden of in-house development and maintenance. However, this balance has shifted as industry dynamics have changed.
Lately, vendors have pushed customers away from self-managed setups and onto cloud platforms they fully control. Now, every seat, API call, and click comes with a cost. At the same time, teams have lost control over uptime, data location, and maintenance schedules. What began as a promise of simplicity now feels more like a SaaS tax for many teams.
The most critical platforms, such as ERPs and CRMs that serve as the organization's single source of truth, are the most difficult to replace. Vendors are aware of these high switching costs, which is often reflected in their pricing.
How Has AI Changed the Case for Building Custom Software?
Now, here's where things get interesting. For years, the main reason to buy instead of build was simple: building took too long and cost too much. But AI has quietly turned that logic on its head.
Thanks to AI-powered development tools, it's now faster and more affordable to spin up working software than ever before. Projects that would have been out of reach for a custom build just a couple of years ago are suddenly on the table. And when that new option shows up right as SaaS frustrations are hitting a high, it's no wonder the whole decision process is changing.
But let's not get ahead of ourselves. There's a big difference between code that works and code that's truly ready for the enterprise. AI can generate something that passes a quick test, but that doesn't mean it's secure, scalable, or easy for another team to pick up down the road.
What Is Ghost Code and Why Should Enterprises Care?
With AI making building easier, it’s no longer just engineering teams creating internal tools. Now, marketing, operations, and finance are all involved. Where a company might have managed ten apps before, now they’re juggling two hundred, many built by people without a software background and running on infrastructure that no one really owns.
When someone builds a tool and then leaves, and no one else knows how it works, that tool becomes what we call ghost code. It keeps running, but there’s no owner and no plan for its future. Over time, these tools quietly add up, creating bigger technical, security, and operational risks for the whole organization.
Moving quickly without the right structure can cause problems that are expensive to fix later.
So, Build or Buy — What's the Right Answer?
The best way to look at this isn't as a simple either-or. Instead, ask yourself: where do we want to take on risk, and what do we have in place to manage it?
If we buy from a vendor, we're signing up for their roadmap and their pricing schedule. If we build, we're taking on the responsibility for the code, the architecture, and everything that comes after. Neither path is automatically right or wrong. The real question is: which risks are we best equipped to handle?
The teams getting this right aren't just picking one side. They're moving quickly on both fronts, but with clear purpose. They know exactly what they're building, why it matters, and who's responsible for it over time. For them, managing the software lifecycle is a core discipline, not something to tack on at the end.
This is the essential discussion organizations must address. The focus should be on effective governance of software assets to ensure that short-term solutions do not become long-term challenges.
We explore these topics in greater detail, including the SaaSpocalypse, vendor lock-in, and practical approaches to intentional software strategy.
Frequently Asked Questions
What is vibe coding, and how do you manage the risks of AI-generated code in an enterprise environment?
What is the build vs. buy decision in enterprise software?
What does SaaS vendor lock-in mean?
What is ghost coding?
Share this Blog with colleagues
