Platform decisions create switching costs that only become fully visible when organizations try to move. AT&T discovered this after Broadcom acquired VMware: the company faced a proposed price increase of 1,050%. Migration costs were estimated at $40-50 million. The multi-year timeline consuming operational capacity to execute the move created the binding constraint.
"A significant investment in engineering work to replace the VMware software, which will be costly and time-consuming. Just choosing an alternative and getting it ready for lab testing could take a year. Completing the migration could take years."
Organizational capacity becomes the measure of cost. The one CISPE member that successfully migrated away from VMware entirely invested several months mobilizing all 400 employees in an intensive development effort to implement open-source alternatives. A complete business transformation that most organizations simply cannot afford in operational bandwidth.
Platform migration consumes years of organizational bandwidth—that's the cost that matters. Your teams execute the move instead of building new capabilities.
Engineering teams that could be building new capabilities spend years executing migration instead. Operations teams that could be optimizing production systems spend bandwidth managing parallel infrastructures during transition. Product roadmaps that assumed certain capabilities get deferred because the organization's capacity is consumed by migration work.
Organizations report cloud migration costs running 3x higher than initial estimates. The multiplier includes accumulated bandwidth consumed by unexpected complexity. A healthcare provider budgeted $1.2M for migration with $100K allocated for training and external expertise. They ultimately spent over $900K on consultants because their team struggled with the cloud learning curve. Budget overruns represent operational capacity redirected from other work.
Infrastructure decisions create this tax through mechanisms that only become visible at production scale. When Broadcom reduced VMware's product offerings from 168 products to four main bundles, organizations that needed only specific components found themselves forced into full bundles. The economics shifted: pay for the full bundle or migrate everything. Starting April 2025, VMware enforced a minimum 72-core license subscription, up from 16 cores. Organizations running smaller deployments suddenly needed to license capacity far exceeding their needs or consume years of capacity migrating away.
The UK Cabinet Office warned in 2024 that overreliance on AWS could cost public bodies as much as £894 million. Switching expenses represent only part of the burden. Add the operational capacity consumed over years to migrate, the features you can't access because you're locked into a specific platform, and the capabilities you can't build because your organization's bandwidth is spent on migration.
Organizations face a choice: pay increased pricing or consume operational capacity over years to migrate. Switching costs compound with the operational bandwidth required to execute the move. When that capacity could otherwise build new capabilities, migration costs accumulate through what you're not building while your organization executes the move.
Platform decisions that optimize for current costs create constraints that only reveal themselves when pricing changes or capabilities shift. Infrastructure choices can consume years of organizational capacity to reverse. Whether your organization can afford to spend years of operational bandwidth on migration rather than building capabilities that matter to your business—that determines what's possible.