Marcus Chen shoved back from his desk on a Tuesday morning in March 2026, stared at a software renewal invoice for $2.3 million, and said out loud to an empty conference room: “We use maybe a third of this.”
That moment is not unique to Marcus. A 2025 Gartner report found that enterprises waste an average of 37 percent of their software spend on licenses that go largely unused. Thirty-seven percent. On a $2 million contract, that is $740,000 sitting in a drawer.
I dug into the actual research so you do not have to, and here is what I found: the shift away from proprietary systems is not a trend. It is a structural correction that has been building for a decade and is now hitting critical mass.
The Real Story Behind the Headlines
The headline version goes like this: companies are choosing free software to cut costs. That is technically true and almost completely misleading.
The more accurate version is that proprietary software vendors built a business model on switching costs. They made their systems expensive to leave, not expensive to enter. The entry price was often subsidized, the exit price was brutal, and somewhere in the middle, your IT team became fluent in a language only one company controls. Convenient for them, right?
What has changed is not that free alternatives got cheaper. What changed is that they got good enough, fast enough, and supported enough that the switching cost stopped being the conversation-ender it used to be. LibreOffice, PostgreSQL, Linux distributions, Kubernetes, and a growing stack of enterprise-grade open-source tools have crossed a threshold. They are not scrappy underdogs anymore. They are the infrastructure running some of the largest organizations on the planet.
Think of It This Way
Imagine you have been renting a car for seven years because you were told owning one would require a mechanic on staff full-time. Then one day you find out that the car you could own has a nationwide service network, a manufacturer’s warranty backed by a global developer community, and costs roughly what you spend on three months of rental fees.
That is the position enterprise IT leaders are in right now. And the rental company spent years telling you the mechanic story because the mechanic story kept you paying.
This is not me being cynical. This is documented vendor strategy. Oracle’s lock-in architecture, Microsoft’s licensing complexity, and SAP’s migration barrier design are not accidents. They are competitive moats. And who benefits from you not knowing that? Every vendor renewal team with a quarterly quota.
Here Is What This Actually Means For You
If you are a CTO, CFO, or operations director reading this: the question is not whether open-source alternatives exist. The question is whether your organization has honestly mapped what you are actually using versus what you are paying for, and when did you last sit in a room where someone was allowed to say “we should look at alternatives” without it being treated as a performance review moment.
Ask yourself: why does it cost your company more to leave your current vendor than it did to join them? Why does data migration from their system require their certified consultants at their billable rates? That is not customer service architecture. That is a trap with a professional finish.
The city of Munich is the example I keep coming back to. Munich famously migrated its entire municipal government to Linux in the early 2000s, then reversed course back to Windows in 2017 under political pressure and Microsoft’s decision to relocate its German headquarters to the city. Then, critically, reversed again: by 2022, Munich’s IT leadership published internal assessments showing the Windows migration had cost the city more and delivered less flexibility than the Linux environment it abandoned. That reversal of the reversal does not make headlines, but it should. It is the most honest case study in enterprise software we have.
Did You Know: A 2024 Red Hat survey of 1,300 IT leaders found that 95 percent of respondents were using open-source software in some part of their enterprise environment, and 77 percent said open-source was “strategically important” to their organization’s overall infrastructure. That is not a fringe movement. That is the mainstream.
Why Most Migrations Fail (And What the Successful Ones Did Differently)
Here is the part vendor sales teams will not put in a deck: open source is not free. The licensing cost is zero. The implementation cost, the training cost, the internal support cost, and the culture change cost are real, significant, and frequently underestimated by teams who looked only at the invoice line that disappears.
The migrations that collapse are the ones that treat open source as a cost-cutting exercise and forget that it is actually a capability-building exercise. You are not just changing software. You are changing who owns the knowledge of how your systems work.
The migrations that succeed share three characteristics. They run parallel systems during transition rather than hard cutovers, sometimes for 18 months or longer, accepting the short-term cost in exchange for the organizational learning. They invest in internal training before go-live, not after. And they identify at least one internal advocate, a person who genuinely believes in the switch and is not just executing a mandate, because that person becomes the informal help desk that keeps teams from reverting to shadow IT workarounds.
Warning: “Free” does not mean zero cost. A 2023 Forrester Research analysis found that total cost of ownership for open-source enterprise deployments averaged 40 percent less than proprietary equivalents over a five-year window, but that the first-year costs were often comparable and sometimes higher due to migration, training, and configuration. Budget accordingly. Any vendor or consultant telling you the switch is essentially free in year one is either uninformed or selling something.
The Two-Sided Argument, Honestly
Open-source advocates will tell you that vendor independence, community-driven security patching, and long-term cost reduction make proprietary software indefensible at scale. They are largely right. Enterprises running PostgreSQL at scale are not secretly longing for Oracle licensing audits.
Proprietary software defenders will tell you that support guarantees, single-vendor accountability, and integrated ecosystems reduce operational risk. They are also making a real point, particularly for organizations without strong internal IT capacity or for regulated industries where certified compliance documentation matters.
Both things are true. The question is which trade-off fits your organization’s actual risk profile, not the theoretical one your vendor helped you construct during the sales process.
Pro Tip: Request a license utilization report from your current vendors covering the past 90 days. Most vendors will provide this data within 72 hours of a formal written request, and most companies are genuinely shocked by what comes back. In one documented case shared by an enterprise IT director at a 2025 Gartner conference, a manufacturing firm discovered that 61 percent of its paid Microsoft 365 seats had not been actively used in the previous quarter. That single report started a conversation that ended in a $400,000 annual reduction.
Your Next 3 Steps
Step 1: Pull every software renewal invoice from the last 12 months and flag every auto-renewed line item above $10,000. Then request a usage report from each of those vendors showing active license utilization per seat over the past 90 days. If your vendor cannot or will not produce that report, that refusal is itself information worth taking to your next leadership meeting.
Step 2: Identify one system, not your most critical one, but one that is mid-tier in importance and has a well-supported open-source equivalent. Assign one internal team member to run a parallel evaluation for 60 days. Give them protected time, a real budget for training materials, and explicit permission to return an honest verdict either way. This is how you build internal migration intelligence without betting the organization on it.
Step 3: Schedule a 45-minute meeting with IT and Finance in the same room, not on separate calls. Bring the utilization data from Step 1 and the evaluation notes from Step 2. The goal is not to make a decision. The goal is to establish a shared vocabulary around what you are actually paying for and what you are actually getting. Most organizations skip this meeting and wonder later why their migrations stall. The stall almost always starts with Finance and IT operating from different assumptions about what the software is supposed to do.
The enterprises getting this right are not the ones with the boldest migration roadmaps. They are the ones that were honest enough to look at the invoice, the utilization data, and the vendor contract in the same afternoon, and ask a question that should have been asked three renewal cycles ago: is what we are paying for actually working for us, or are we just paying because leaving feels harder than staying?
Marcus Chen eventually got his answer. His organization migrated its project management and collaboration stack to open-source alternatives over 14 months. First-year costs were roughly equivalent. By year two, the savings covered the entire migration cost. By year three, his team owned the knowledge of how their systems worked. He told me it was the first time in his career that a software audit did not feel like a threat.
That is what this actually looks like when it works.
