Leadership1 publisher3 min readPublished
StorMagic's CEO tells retailers to separate the VMware migration from the hardware refresh
Sue Odle argues that a retailer with a thousand stores has to settle one repeatable deployment model before it touches a single site. She runs a company that sells the replacement. Her sequencing argument is testable.
The Board Room · Leadership desk
What happened
- Sue Odle, CEO of the on-site virtualization vendor StorMagic, wrote in Forbes Tech Council that VMware licensing changes have pushed many retailers to transform their virtualization strategy.
- She treats remote deployment as a baseline: image and configure systems before they reach the store, and limit store staff to basic physical tasks while technical work happens centrally.
- Where deployment depends on local troubleshooting, she writes, travel costs increase, maintenance windows get harder to coordinate and central IT spends more time on individual exceptions.
Compiled by The Board RoomSomething wrong?How this is made
Why it matters
- decision Splitting the software migration from the hardware refresh takes the platform decision off the capital calendar, so the choice becomes which budget carries the migration and in which quarter.
- constraint With no skilled engineer available at most sites, the ceiling on onsite skill sets how fast a thousand-store estate can move.
- exposure The satisfaction figure rests on a study the piece does not name and the advice comes from a supplier of the replacement, so a board using it as a plan is leaning on interested evidence.
- precedent If maintenance windows are set by workload risk, migration planning gets organised by application tier across the fleet, and site-by-site scheduling becomes the exception.
Odle discloses her interest in the first line: she is CEO of StorMagic, which helps organizations simplify on-site virtualization for distributed IT environments [1]. A vendor of edge virtualization has an obvious reason to call a licensing renewal an estate-wide problem, but her description of the problem can be tested. For organizations with hundreds or thousands of locations, she wrote, the difficulty is managing the platform migration without disrupting store operations, increasing support requirements or leaving configurations that become hard to manage over time [3]. Her first recommended step asks for facts the retailer already holds: which workloads must run locally without cloud dependency, what hardware is in place and whether it can be reused, and how latency and bandwidth vary by location [15].
A virtualization migration does not require a hardware refresh, Odle wrote, and retailers should assess whether existing servers can still carry the local workloads, particularly where the hardware has useful life left [9]. This is the sequencing decision with cost consequences. Replacing software and hardware at the same time increases cost and adds more moving parts to each deployment. Separating the two decisions gives IT leaders greater control over capital planning and lets hardware be refreshed according to condition, workload requirements and lifecycle plans [10].
Odle cites a study finding that only 39% of retailers are satisfied with their edge computing support [5], which leaves 61% who are not [6]. She does not identify the study, and the article does not mention Broadcom or describe what the VMware licensing changes were [16]. Her stated reason the pressure is rising is that AI workloads, "from near real-time video analytics for loss prevention to predictive maintenance, move closer to the store level" [7].
Applications supporting sales transactions, inventory, security and other daily functions often run locally with no onsite IT support [4], and most retailers cannot put a skilled engineer in every store during a migration [11]. The constraint on the schedule is staffing. So the work moves to the center: image and configure systems before they reach the store, plan for minimal intervention by store staff, and limit the onsite role to basic physical tasks while the technical work is handled remotely [12]. When deployment depends heavily on local troubleshooting, Odle wrote, travel costs increase, maintenance windows become harder to coordinate and central IT spends more time managing individual exceptions [13].
A point-of-sale application carries a different risk than a system that can tolerate several hours of downtime [14], and connectivity is not guaranteed and varies by location [17]. What Odle argues has to hold across the estate for years is a short list: consistent deployment methods, unified management and orchestration, and support programs during and after the migration. "These are nonnegotiables," she wrote [8].
What to watch
- Identification of the study behind the 39% edge satisfaction figure, so retailers can check it against their own estates.
- Any retailer publishing per-site cost or timeline data from a completed multi-site virtualization migration.
- Further changes to VMware licensing terms at renewal. A second change would send retailers back through the estate assessment.