A newly published study by Backblaze and Dimensional Research has uncovered that a significant majority of enterprises dealing with substantial data volumes are facing unforeseen cloud storage costs. This situation poses challenges for infrastructure planning, especially as the adoption of artificial intelligence accelerates.

The study reveals that a staggering 95 percent of participants reported unexpected cloud storage charges, primarily attributed to egress fees and unclear billing systems. The survey encompassed 403 IT executives responsible for managing over 250 terabytes of data in the cloud. The findings highlight that the expense associated with accessing and transferring data is now a key impediment to embracing multi-cloud strategies, as indicated by 58 percent of those surveyed.

The report suggests that enterprises managing more than five petabytes of data are particularly susceptible to budget overruns. As AI and analytics workloads demand faster, larger datasets, organizations are being compelled to scale back or adjust priorities to manage costs.

Addressing the issue of escalating storage expenses, Gleb Budman, CEO of Backblaze, stated, “The data shows what many IT leaders already know from experience: innovation is being throttled not by technology limits, but by cloud economics via egress fees and unpredictable charges.”

The study also points to a trend toward independent storage providers. Sixty-two percent of respondents expressed a preference for best-of-breed vendors over single-provider ecosystems, citing the necessity to avoid restrictive pricing models and vendor lock-in. However, the study also discovered that 97 percent of organizations view data movement costs and technical complexity as significant obstacles to switching providers.

“As AI and other data-intensive workloads expand, the ability to move data freely and affordably is no longer optional — it’s a strategic requirement,” noted Diane Hagglund, principal at Dimensional Research and the study's author.

The “2025 Cloud Storage Survey” was carried out between May 24 and June 5, 2025.