News & Press: Redress Compliance

Redress Compliance Completes 500th Enterprise Engagement

Redress Compliance has reached a significant milestone with the completion of its 500th enterprise engagement, marking five years of independent software licensing advisory since the firm's founding in 2019. The milestone reflects sustained growth across the advisory practice, expansion of vendor coverage to 11 practices, and deepening specialisation in compliance and cost optimisation for enterprise customers.

The firm has expanded its practice coverage to include Oracle, Microsoft, SAP, Salesforce, IBM, Broadcom/VMware, AWS, Google Cloud, ServiceNow, Workday, and GenAI vendors. Over this period, the advisory team has grown to incorporate specialists from the vendor community, corporate procurement, and software compliance backgrounds, bringing cumulative experience spanning more than two decades.

Reflecting on the milestone, co-founder Morten Andersen commented: "When we started Redress in 2019, the market for independent software licensing advisory was fragmented. Customers were caught between vendor sales narratives and consultant recommendations influenced by vendor relationships. Five hundred engagements later, we have demonstrated that buyer-side independent advisory is not only viable but essential. Our customers have identified over $800 million in cost savings and compliance gaps that vendor-led assessments consistently missed."

The 500th engagement milestone was marked by the appointment of new advisor roles spanning executive leadership, vendor advisory, and account management. The firm has also launched formal certification programmes for junior advisors, establishing consistent methodologies across the growing team.

Broadcom VMware Subscription Transition Creates Audit Urgency for 14,000+ Customers

Broadcom's completion of its acquisition of VMware in late 2023 has triggered a forced migration of the entire VMware customer base from perpetual licensing to subscription-based models, affecting an estimated 14,000+ enterprise customers. The transition, which becomes mandatory during 2024 renewals, carries significant financial and operational implications: support costs are increasing 3-5x on average, and customers face a compressed timeline to evaluate alternatives or renegotiate terms.

The subscription transition is creating immediate audit and cost review urgency. Many customers who purchased VMware perpetual licenses 3-7 years ago paid between $2,000 and $8,000 per processor for perpetual rights. The equivalent annual subscription cost now ranges from $1,200 to $3,500 per processor per year—meaning a 10-processor deployment that cost $20,000-$80,000 upfront now costs $12,000-$35,000 annually, with no purchase equity remaining. Over a five-year period, the subscription model can exceed the original perpetual purchase price by 30-150%, depending on initial licensing tier.

Beyond cost, the transition is prompting customers to reassess their virtualisation architecture. Nutanix hyperconverged infrastructure and Azure VMware Solution are emerging as viable alternatives for organisations seeking to avoid vendor lock-in or reduce annual support obligations. Customers with multi-year contracts locked into Broadcom pricing through 2026-2027 are exploring tactical license deactivation or server consolidation strategies to reduce their VMware footprint before renewal.

Organisations affected by the VMware transition face a narrow window to conduct proper cost analysis and alternative evaluation. Historical support entitlements are being discontinued, and Broadcom is accelerating the sunsetting of perpetual support agreements, typically allowing 12-18 months of overlap before mandatory transition to subscription support at the new pricing tiers.

IBM ILMT Compliance Failures Spike Post-PVU-to-VPC Transition

IBM's transition from Processor Value Units (PVU) to Virtual Processor Core (VPC) licensing for enterprise software has created a documented spike in ILMT (IBM License Metric Tool) configuration failures and sub-capacity licensing misstatements. The transition, which became effective in 2023 for new software releases, introduces new licensing complexity that many customers are still resolving in their 2024 compliance assessments.

The core compliance issue centres on ILMT configuration accuracy. Under PVU licensing, ILMT could measure workload-based compliance at the server level. Under VPC licensing, ILMT must accurately track virtual processor allocation at the hypervisor level—a significantly more complex requirement. Organisations migrating from PVU to VPC are discovering that their existing ILMT deployments lack sufficient hypervisor integration, leading to undercounting of virtual processor assignments and invalid sub-capacity licensing claims.

Proper ILMT configuration is essential for any sub-capacity licensing claim to be valid under IBM's audit methodology. When IBM audit teams discover ILMT gaps during compliance reviews, they typically reject the customer's sub-capacity position, reclassifying usage as full-capacity and triggering significant true-up invoices. In several recent cases, IBM has widened audit scope retroactively when initial ILMT data is inadequate, extending assessment periods from three years to five years or more.

The compliance failures are occurring even in organisations with dedicated software licensing teams, indicating that the transition complexity is widespread. Customers are now implementing enhanced ILMT monitoring, hypervisor integration verification, and quarterly reconciliation processes to prevent audit exposure. The PVU-to-VPC transition has become a critical compliance refresh point for IBM customers, with implications extending beyond the initial transition period into ongoing true-up and renewal negotiations.

Oracle Java Licensing Claims Intensify Under New Per-Employee Model

Oracle's shift from named user/processor licensing models for Java SE to an enterprise-wide per-employee licensing model (Java SE Subscription) has triggered a significant increase in licensing claims and audit activity. The change, which took effect in January 2023, affects any organisation deploying Java SE in production environments and removes the ability to segregate Java usage to specific business units or departments.

Under the legacy named user model, organisations could license Java SE for a specific set of users or applications. Under the new per-employee model, Java SE licensing is assessed on the entire employee base regardless of actual Java exposure. This creates a stark difference in licensing cost: a 10,000-employee organisation with 800 actual Java developers and users now requires licensing for all 10,000 employees at approximately $40-60 per employee per year, versus approximately $3,200-$4,800 per named user under the previous model.

The transition has prompted a sharp increase in Oracle compliance claims, with many organisations receiving audit notices specifically addressing Java SE licensing scope and employee population calculations. Oracle is taking an aggressive position on Java deployment breadth, arguing that organisations deploying Java in any production system (including databases, middleware, and third-party applications that embed Java) must license the entire employee population, not just direct developers.

Organisations defending their Java licensing positions are increasingly engaging independent advisory support to validate their employee population calculations and establish defensible licensing scopes. The Java licensing transitions highlight a broader trend: Oracle is leveraging enterprise-wide licensing models (per-employee, per-user) to increase contract values and audit exposure across customers who previously had more granular licensing flexibility.

Microsoft EA Tier Elimination Effective November 2025: What Customers Must Do Now

Microsoft has announced the elimination of volume-based pricing tiers from its Enterprise Agreement (EA) framework, effective November 2025. The change removes the historical three-tier structure (Level C, Level D, Level E) that provided cost advantages for larger organisations and will replace it with a single standard pricing model applied across all customer sizes. For many enterprise customers, this represents an effective price increase of 8-15% for Level D/mid-market customers, with smaller impact for Level E (large) customers who already received minimal tier benefits.

The tier elimination reflects Microsoft's shift toward consumption-based and per-user licensing models that inherently scale pricing with usage, reducing the negotiation leverage that historically accrued to large customers. Organisations currently benefiting from Level D discount positioning (typically 15-25% below Level C pricing) face the most significant exposure: their November 2025 renewal will apply Level C-equivalent pricing to their entire user and seat base, creating an effective 8-15% cost step-up.

The timeline creates an urgent planning window for impacted customers. Any EA renewal completed before November 2025 will lock in existing tier-based pricing through the contract term (typically 3-5 years). Organisations currently on Level D or Level E agreements should prioritise EA renewal negotiations for 2024-2025, well ahead of the November 2025 effective date. This approach allows customers to negotiate new terms while tier-based pricing is still available, creating a "price hold" before the tier framework eliminates.

Beyond cost implications, the tier elimination affects EA negotiation strategy broadly. Tier-based pricing has historically served as the primary negotiation lever for large customers, justifying investment in EA renewals and driving competitive dynamics between Microsoft and competing platforms. With tier-based pricing removed, customers will need to negotiate value through other mechanisms: service credits, advanced feature access, or targeted investments in specific workloads (such as AI/Copilot licensing or managed services).

GenAI Vendor Lock-In: OpenAI Enterprise Agreements Under Scrutiny

OpenAI's enterprise agreement structure for ChatGPT and custom GPT deployments is facing increasing scrutiny from enterprise customers and procurement teams concerned about vendor lock-in, pricing unpredictability, and consumption-based cost escalation. Unlike traditional enterprise software licensing with fixed per-seat or per-year costs, OpenAI enterprise agreements use token-based consumption pricing that scales with API usage, creating budget uncertainty for organisations deploying ChatGPT across business units.

The core concern centres on cost unpredictability. OpenAI's enterprise agreements require organisations to commit to minimum annual spend (typically $30,000-$100,000+) with pricing that scales dynamically based on token consumption. An organisation that initially projected 100 million tokens per quarter could see actual consumption spike to 300+ million tokens as adoption accelerates, driving quarterly costs from $7,500 to $22,500+. Over the course of a contract term, unforeseen consumption growth can create budget overruns of 50-200%.

Additionally, OpenAI's single-vendor positioning creates architectural lock-in risk. Organisations that embed OpenAI's GPT models directly into applications and workflows face significant migration costs if they later want to switch to competing models (Azure OpenAI, Anthropic Claude, open-source alternatives). Unlike traditional cloud providers that support multi-vendor competition, the GenAI market is still establishing competitive parity, leaving customers with limited fallback options if OpenAI pricing becomes uncompetitive.

Azure OpenAI offers an alternative that provides more pricing stability (committed capacity pricing similar to Azure reservations) and reduced vendor lock-in (token consumption is portable across Azure services). However, Azure OpenAI functionality lags OpenAI direct access in some domains, creating a trade-off between cost control and feature availability. Organisations are increasingly negotiating Azure OpenAI for production workloads with strict consumption budgets, while reserving direct OpenAI access for pilot programmes and evaluation use cases.

Industry Outlook: Six Strategic Priorities for H2 2024

1. Accelerate VMware Alternative Evaluation: Organisations on VMware perpetual licenses should complete architecture reviews of Nutanix, Azure VMware Solution, and alternative hypervisors before the 2025 renewal cycle locks in multi-year subscription commitments. The cost differential between legacy perpetual and new subscription pricing is substantial enough to justify migration evaluation.

2. Validate IBM ILMT Configuration: Any organisation with IBM database or middleware deployments should conduct ILMT compliance verification, particularly if they recently transitioned from PVU to VPC licensing. Configuration gaps identified during internal reviews are far less costly to remediate than gaps discovered during IBM audit.

3. Conduct Oracle Java Licensing Scope Review: Organisations deploying Java in any production capacity should validate their employee population calculations and licensing scope assumptions against Oracle's current audit methodology. Java licensing disputes are among the highest-value audit findings Oracle is pursuing in 2024-2025.

4. Prioritise Microsoft EA Renewal Before November 2025: Any organisation on Level D or Level E EA pricing should move EA renewal discussions forward to capture tier-based pricing before the November 2025 elimination date. Delaying renewal past November 2025 will apply new standard pricing to the entire contract base.

5. Establish GenAI Licensing Governance: Organisations piloting or deploying OpenAI or competing GenAI models should establish consumption monitoring, budget controls, and cost allocation processes before adoption accelerates. Early governance prevents budget surprises as adoption spreads across business units.

6. Engage Independent Advisory for Multi-Vendor Assessments: The complexity and pace of vendor transitions (VMware, IBM, Oracle, Microsoft, GenAI) are creating a need for comprehensive, independent cost and compliance assessments. Vendor-led assessments are inherently biased toward consolidation narratives that may not reflect customer economics.