Panzura Named as a Sample Vendor in the Gartner® Hype Cycle™ for Strategic Cost Management 2026

Our View on the Rise of File Data Management and AI Capabilities in Enterprise Storage and the Benefits, Obstacles, and Future of the Broader Hybrid Cloud Landscape

Gartner® has published the Hype Cycle for Strategic Cost Management 2026 [1], and Panzura is named as a Sample Vendor in the Hybrid Cloud Storage category.

“The Hype Cycle for Strategic Cost Management examines how organizations manage enterprise technology cost across the life cycle from funding and design decisions through execution and optimization. As enterprise technology becomes increasingly consumption-driven, cost is no longer set at planning, but shaped continuously by how technology is designed, used and scaled. A smaller set — including financial management for AI, business capability modeling and composable ERP — shapes cost at the point of funding and design, where economic outcomes are largely determined. In contrast, a larger group — including FinOps, observability and optimization technologies — focuses on improving cost efficiency after deployment.”

The Hybrid Cloud Storage category is authored by Julia Palmer and appears in the Sliding into the Trough section of the curve.

Gartner defines the category:

“Hybrid cloud storage bridges on-premises, edge, and cloud storage services, allowing organizations to manage data across different locations. This offers flexibility, scalability, and cost-effectiveness across multiple environments by facilitating seamless data movement, improved management, and collaboration. Supporting file, block, and/or object storage, it provides tools for life cycle management, synchronization, and disaster recovering, ensuring data accessibility, protection, and resilience.”

Why Panzura thinks file data belongs in a cost conversation

For decades, unstructured file data has been budgeted as infrastructure. It has been owned by storage teams, refreshed on a capital cycle, and negotiated in terabytes. Panzura considers that framing to be incomplete. In cost terms, in most distributed organizations, the majority of the money attached to file data never appears in the storage budget.

Some of it resides in branch infrastructure that has to be refreshed at every location, on a schedule set by a vendor’s lifecycle policy rather than by the business. Moreover, some unstructured data is assigned to the separate backup estate, licensed by capacity, that exists to protect data the organization already owns and pays for. Some resides in disaster recovery arranged site by site, where each office carries its own contract because there is no way to protect all of them together.

The largest share of costs, in Panzura’s experience, does not appear on an invoice. It is the hours engineers spend reconciling the same working files across offices that cannot see each other’s work. That includes chasing a version, unraveling a conflict, and explaining to a project team why a file is out of date. That labor is rarely counted as storage cost. It is counted as engineering time, absorbed into project schedules, and largely endured as a part of the job of IT.

Then there is the cost of not knowing. Most organizations of any age are carrying file shares nobody has opened in years, project directories belonging to teams that no longer exist, and duplicate working sets created during migrations that were never cleaned up afterward. Each has costs in terms of capacity, backup windows, and the attack surface that must be defended. Panzura sees data visibility as a cost control in its own right because a share nobody can account for is also a share where decisions cannot be effectively made.

Consumption pricing has changed the shape of the problem as well. When storage was bought as equipment, cost was fixed at purchase and stayed fixed. When it is consumed as a service, cost moves with behavior. That is, with how often data is read, where it is read from, and how much of it crosses a boundary to get there. Panzura considers this a large part of why file infrastructure decisions have moved closer to the CFO. An architectural choice made in the past transforms into a variable line item in the time that follows.

Panzura sees this as the practical case for consolidating distributed file data onto a single platform, which is exactly what Panzura CloudFS is built to do. The savings that matter are rarely about storage alone.

Table 1. Panzura view of cost savings with CloudFS

How cost savings compound as sites are added

In a site‑by‑site architecture, every location that needs local‑speed access to the shared dataset has to provision for all of it. Add a location and the requirement grows by that amount again. CloudFS holds the authoritative dataset once in object storage, the least expensive tier in the estate, and gives each site a node with a cache on fast local storage. Adding a location adds a cache. It does not add the dataset again.

The cache is not a percentage of capacity. It is sized to the working set, and the working set does not grow simply because the namespace does. That is why the advantage compounds with each location rather than holding flat, and why it is pronounced as sites multiply.

Two lenses on the same business case

Panzura encourages anyone with access to the research to read the profile’s user recommendations in full to fully understand the authors’ intent. What Panzura would add is drawn from our own customer deployments and on-the-ground experience. As we see it, organizations tend to evaluate file infrastructure through one of two lenses, and both are legitimate. They answer different questions, and the lens chosen in the first instance tends to shape the contours of both the business and IT decisions well in advance of acquisition and roll out.

Table 2. Business case view from Panzura perspective

Panzura considers the workflow lens more difficult to understand for many customers, but also more likely to produce a business case that survives long term. Panzura also accepts that plenty of sound purchases are made primarily through the first lens, and it should be a part of every decision without reservation, particularly where the requirement is genuinely capacity and the workflow already works. Taken together, both lenses provide a more complete picture in our view.

What a workflow pilot has to prove

The difficulty with the workflow lens is that it asks for evidence that a pricing matrix cannot fully deliver. Panzura’s suggestion, for teams building that case, is to pick the workload that causes the most friction or drag for the organization, rather than the one that shows best in a demo. A pilot run on a convenient dataset, one that is limited in size and scope but fits into a demo environment well, proves very little about a large or complex one.

Three measurements tend to settle the argument. The first is how long it takes a user at the furthest site to open the largest file they genuinely work on, measured on their connection rather than a test link. The second is how long a recovery takes when run by the team, at the time of day when they would be doing it. The third is how many administrator hours the estate consumes in a month, counted honestly, including the inevitable interruptions and complications that typically arise.

Panzura considers the last point as the aspect that is least quantifiable and most often skipped, yet perhaps the most decisive. Hardware costs are visible and negotiable. Administrative load is invisible and compounds with every site added. Capacity runs the other way, and it is the aspect of the storage estate that gets proportionally cheaper with CloudFS as locations are added.

Where AI enters the discussion

Panzura’s own view is that enterprise AI has changed the stakes as to file data, and that this belongs squarely inside the cost conversation rather than adjacent to it. It is an inseparable aspect of the discussion, and needless to say, will become increasingly so.

AI systems run on enterprise-class data, and in most organizations the bulk of that data is unstructured file data spread across sites, backups, and archives. Panzura’s position is that an AI system cannot produce a trustworthy answer from a dataset that cannot be easily identified, located, and for which the validity cannot be verified.

As we see it, there are two challenges that get in the way. The first is reach. Where part of the file estate is offline, archived into a format that cannot be indexed, or sitting on a branch server where central systems do not know about it, it is effectively invisible to an AI system regardless of how capable that system may be. The answer comes back confident and incomplete, which is worse than no answer at all.

The second is permission. An AI assistant inherits whatever access model the underlying file estate already carries. Where permissions have accumulated through years of reorganizations and departures, an assistant that searches faithfully across everything a user can technically reach will surface material that user was never meant to see. Panzura considers this to be a governance risk that is far too often left unexamined before a pilot. Importantly, it is also a cost risk, because remediation after the fact is expensive and the permission leak cannot be undone.

Keeping file data live, governed and available from a single source of truth addresses both, which is why Panzura Nexus exists to connect that data directly to AI systems including Microsoft 365 Copilot. Panzura sees this as the reason data consolidation is necessary across at least two horizons. It removes duplicated operational cost, and it determines whether an AI investment in the future has data that is reliable to read.

What Panzura takes from the recognition

Panzura is pleased to be named in the Gartner report and we are grateful to the customers whose deployments have and continue to inform our understanding of the category.

As we see it, this inclusion as a Sample Vendor is not a recommendation to buy. From our view, the report validates that hybrid cloud storage is being examined as part of how organizations manage technology cost, and we consider that the right conversation to be having about file data.

 Most organizations cannot say what their distributed file estate truly costs, because the number is scattered across branch refresh cycles, backup licensing, per-site DR contracts, and administrator hours nobody totals. A Panzura expert will help you add it up against your own estate.

[1] Gartner®, Hype Cycle for Strategic Cost Management, 2026, Cesar Lozada, Robert Neagle, Lauren Wheatley, 25 June 2026

GARTNER and Hype Cycle are trademarks of Gartner, Inc. and/or its affiliates.

Gartner does not endorse any company, vendor, product or service depicted in its publications, and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner publications consist of the opinions of Gartner’s business and technology insights organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this publication, including any warranties of merchantability or fitness for a particular purpose.

Frequently Asked Questions

• How do you calculate the total cost of distributed file storage across multiple sites?
  

  Start with three inputs: the number of sites, the working set each site actively uses, and the current per-site spend on primary storage, backup licensing, disaster recovery contracts, and hardware refresh. Most organizations count only the first, which is why file infrastructure looks cheaper than it is. Add administrator hours as a fourth input, counted across a full month including interruptions. The total is almost always larger than the storage budget suggests, and that gap is what a consolidation business case is built on.

• Why does the cost advantage of a hybrid cloud file platform grow as you add sites?
  

  In a site-by-site architecture, every location that needs local-speed access to shared data has to provision for the entire dataset, so cost tracks the number of sites. Panzura CloudFS holds the authoritative dataset once in object storage and gives each site a node with a local cache. Adding a location adds a cache, not the dataset again. Because the central requirement is spread across every site added, the advantage is pronounced as sites multiply.  

• How much local cache does each site need with Panzura CloudFS?

  With Panzura CloudFS, cache is sized to the working set, meaning the data people at that site are actively using, rather than to a percentage of total capacity. The distinction matters because the working set does not grow simply because the namespace does. A site actively touching a fixed volume of live project data needs the same cache whether the shared namespace holds one hundred terabytes or four hundred. Rules of thumb expressed as a percentage of capacity understate the outcome at scale and overstate it for smaller estates.

• What file infrastructure costs do not appear in the storage budget?

  Four categories are outside the storage budget. Branch hardware refreshed at every location on a vendor lifecycle rather than a business one. A separate backup estate licensed by capacity to protect data the organization already owns. Disaster recovery arranged site by site, with each office carrying its own contract. And engineering hours spent reconciling working files across offices that cannot see each other, which gets recorded as project time rather than storage cost. That last category is usually the largest and the least measured.

• Should file infrastructure be evaluated on cost per terabyte or on workflow cost?

  Both, but they answer different questions. Cost per terabyte answers what a system costs to buy, uses price sheets normalized to capacity, and produces a faster decision. Workflow cost answers what it costs to operate and what it changes and requires a pilot on live production work. The capacity lens is the right primary measure when the requirement is capacity alone and the existing workflow already functions. Where distributed teams are losing time to version conflicts and remote file access, the workflow lens produces the business case that survives.

• What should a file infrastructure pilot measure?

  Three things, run on the workload causing the most friction rather than the one that demos well. First, how long a user at the furthest site takes to open the largest file they actually work on, measured on their own connection. Second, how long a recovery takes when the team runs it, at the hour they would really be doing it. Third, administrator hours consumed across a month, counted honestly with interruptions included. A pilot on a small convenient dataset proves little about a large one.

• Why consolidate file data before deploying Microsoft 365 Copilot?

  Two problems make an unconsolidated estate a poor foundation. The first is reach. File data that is offline, archived in an unindexable format, or sitting on a branch server central systems do not track is invisible to Microsoft 365 Copilot, which returns answers that are confident and incomplete. The second is permission. An assistant inherits whatever access model the file estate carries, so permissions accumulated through years of reorganizations will surface material users were never meant to see. Panzura Nexus addresses both issues by connecting governed Panzura CloudFS file data directly to Copilot without migration.