Software licensing analysis
Pillar · Cluster IX · Editorial reading list

Cloud licensing, read in full.

A buyer-side reading list on AWS, Microsoft Azure and Google Cloud. The Enterprise Discount Program, the Microsoft Azure Consumption Commitment, the Google Cloud committed-use envelope. Reserved Instances and Savings Plans. BYOL bridges into the hyperscalers. The Bedrock, Azure OpenAI and Vertex AI commercial surfaces. The pages below cluster the firm’s cloud commentary into a single editorial reading set.

ClusterIX · Cloud
Sections10
Spoke pages40
PublishedDecember 2024
Independent. Buyer-side. Not a partner, reseller, or affiliate of Amazon Web Services, Microsoft, Google, or any other software vendor.

Inside the pillar

  1. Why cloud is a licensing problem
  2. The three commitment instruments
  3. AWS EDP and the Private Pricing Addendum
  4. Microsoft Azure MACC and the consumption envelope
  5. Google Cloud EDP and committed-use discounts
  6. Reserved Instances and Savings Plans
  7. BYOL bridges into the hyperscalers
  8. Generative AI commitments
  9. FinOps versus licensing
  10. Reading list
Section i

Why cloud is a licensing problem.

A cloud bill reads, on the surface, as consumption. A closer read finds three load-bearing commercial layers: the published rate-card, the cloud-vendor commitment instrument (EDP, MACC, EDP Cloud) and the embedded software entitlements (BYOL, marketplace, native services). Each layer carries its own renewal cycle, its own ramp, its own audit posture.

The procurement implication is that a cloud commitment is not a FinOps optimisation problem. It is a multi-year commercial commitment that interacts with the buyer’s wider software portfolio. An AWS engagement, an Azure engagement or a Google Cloud engagement sizes the commitment against the deployed workload, structures the discount curve against the right tenor and protects the buyer against the publisher’s standard ramp design.

This pillar groups the cloud commentary into ten editorial sections. Each section names the load-bearing mechanic, links the deeper spoke articles and points to the practice page and the relevant white papers for the buyer who wants the engagement methodology.

Section ii

The three commitment instruments.

AWS, Microsoft Azure and Google Cloud each operate a private discount instrument that sits above the public rate-card. AWS calls it the Enterprise Discount Program (and its variants, the Private Pricing Addendum and the Strategic Collaboration Agreement). Microsoft calls it the Microsoft Azure Consumption Commitment. Google calls it the Google Cloud Enterprise Discount Program.

The three instruments share a common architecture. The buyer commits to a multi-year consumption envelope (typically three years, sometimes five). The publisher returns a discount that grows with commitment size and tenor. The envelope counts against named eligible services. The unconsumed envelope is, by default, forfeit at expiry. The buyer-side methodology designs the envelope size against the deployed and forecast workload (not against the publisher’s aspirational forecast), the eligible-service mix and the ramp curve.

The full methodology sits in three companion white papers: AWS EDP Commitment Design, Microsoft Azure MACC and Google Cloud EDP Design.

Section iii

AWS EDP and the Private Pricing Addendum.

The AWS Enterprise Discount Program is the private discount instrument that converts a multi-year consumption commitment into a stepped discount against the AWS public rate-card. The standard tenor is three years. The discount curve grows with commitment level, eligible-service breadth and the architectural commitment to AWS-native services.

The buyer-side question is the commitment’s size relative to the deployed workload. A commitment sized to the publisher’s forecast typically over-commits by 15 to 35 percent against the realised consumption. A commitment sized to the firm’s own trailing-twelve-month consumption (with a defensible growth curve added on top) sits inside the realised envelope and leaves the buyer protected against over-commitment.

The Private Pricing Addendum is the contractual document that records the EDP discount. The PPA carries protective clauses (RI consumption inclusion, marketplace inclusion, services exclusions, true-down protections, mid-term re-sizing rights) that the buyer should negotiate explicitly rather than accept on default. The full reading sits in the EDP Commitment Design paper.

Section iv

Microsoft Azure MACC and the consumption envelope.

The Microsoft Azure Consumption Commitment is the Microsoft equivalent. The MACC is a three-year (sometimes five-year) consumption commitment that counts against Azure consumption (and against Marketplace transactions billed through Azure) in exchange for an enterprise discount and access to publisher incentive programmes.

The buyer-side question runs on the same axis as EDP: is the commitment sized to the realised consumption or to the account team’s aspirational forecast. The Admodum methodology sizes the MACC against trailing consumption, against the named workload migrations with realistic timelines, and against a contestable growth assumption. Marketplace inclusion is the load-bearing clause: a MACC that consumes Marketplace transactions extends the envelope to a wider eligible-spend base than a MACC restricted to Azure native services.

The MACC is also a negotiating instrument inside the wider Microsoft Enterprise Agreement. A MACC commitment can be packaged with the EA renewal as a commercial vehicle that releases discount across the wider Microsoft portfolio. The full reading sits in the Microsoft Azure MACC paper.

A cloud commitment is not a FinOps optimisation. It is a multi-year commercial position with renewal and ramp consequences for the wider portfolio.
Section v

Google Cloud EDP and committed-use discounts.

Google Cloud operates two related commercial instruments. The Google Cloud Enterprise Discount Program is the multi-year consumption commitment that delivers an enterprise discount against the GCP rate-card. Committed-Use Discounts are the workload-level commitments (one-year and three-year) that bind specific resource families (Compute, Memorystore, Cloud SQL) to a fixed discount in exchange for a usage commitment.

The buyer-side methodology stacks the two. The Google Cloud EDP delivers the enterprise-wide discount and access to Google’s solution funding. CUDs deliver the workload-level discount on the steady-state base. The combination sits below the rate-card and protects the buyer against the publisher’s tendency to design the EDP around a single discount layer.

Vertex AI consumption sits inside the EDP envelope where the contractual document includes Vertex AI as an eligible service. The Admodum methodology negotiates the Vertex AI inclusion explicitly rather than accepting the default. The full reading sits in the Google Cloud EDP Design paper.

Section vi

Reserved Instances and Savings Plans.

Reserved Instances and Savings Plans are the workload-level commercial instruments that sit inside the wider EDP / MACC / EDP envelope. AWS operates Standard RIs, Convertible RIs, Compute Savings Plans and EC2 Instance Savings Plans. Azure operates Reserved Instances and the newer Azure Savings Plan for Compute. Google operates Committed-Use Discounts on Compute and on a growing list of managed services.

The buyer-side question is the right mix of tenor (one-year versus three-year), the right mix of instrument type (Standard / Convertible / Savings Plan / CUD), and the right coverage ratio (the percentage of the steady-state base covered by commitment instruments). The Admodum methodology runs the trailing-twelve-month consumption against the instrument families and produces a recommendation that protects the buyer against tenor lock-in while delivering the discount envelope.

The instruments interact with the EDP / MACC / EDP envelope in non-obvious ways. RI consumption counts against the EDP envelope (so an aggressive RI position can over-commit the EDP envelope and waste the discount). The Admodum methodology sizes the two layers together to avoid double commitment.

Section vii

BYOL bridges into the hyperscalers.

A material share of the cloud-licensing position runs through the BYOL (Bring Your Own Licence) bridges that connect on-premise software entitlements to hyperscaler deployments. Oracle Database on AWS, Azure or GCP runs against the Oracle Cloud Authorisation document. SQL Server on AWS or GCP runs against the Microsoft Product Terms with the Azure Hybrid Benefit exception. Windows Server on the three hyperscalers runs against License Mobility through Software Assurance.

The buyer-side methodology reads the BYOL bridges as commercial instruments and not as technical artefacts. A BYOL bridge that is read correctly extends the buyer’s on-premise licence portfolio into the cloud at no incremental publisher cost. A BYOL bridge that is read incorrectly creates a compliance gap that the publisher will surface inside the next audit.

The Admodum methodology produces the BYOL inventory at the start of any cloud commitment engagement. The inventory reads as a per-publisher, per-product, per-deployment list of contractually eligible BYOL workloads. The inventory becomes the input to the EDP / MACC / EDP sizing exercise.

Section viii

Generative AI commitments.

AWS Bedrock, Microsoft Azure OpenAI Service and Google Vertex AI are the three principal hyperscaler generative-AI commercial surfaces. Each operates on a Provisioned Throughput Unit construct (with naming variations) that converts on-demand token-based pricing into a reserved-capacity commitment at a per-hour rate.

The buyer-side methodology sizes the PTU commitment against trailing sustained consumption (not against the publisher’s aspirational rollout assumption). The break-even point between on-demand and PTU sits at a defined utilisation ratio per model; the Admodum methodology runs the arithmetic on the buyer’s own consumption telemetry and sizes the PTU envelope at the right utilisation ratio.

The PTU commitment interacts with the EDP / MACC envelope. Bedrock spend on AWS counts against the EDP envelope where the contractual document includes Bedrock as an eligible service. Azure OpenAI spend counts against the MACC envelope on the same basis. The Admodum methodology negotiates the inclusion explicitly. The full reading sits in three papers: AWS Bedrock Commitment, AI Vendors PTU Design and the AI Vendors practice page.

Section ix

FinOps versus licensing.

FinOps is the operational practice that optimises cloud consumption against the realised workload (right-sizing, scheduling, storage class management, idle-resource cleanup). Licensing is the commercial practice that designs the multi-year commitment instruments against the realised consumption. The two practices interact but are not interchangeable.

The buyer-side methodology runs FinOps as the input layer to the licensing exercise. The FinOps team produces the trailing-twelve-month consumption telemetry, the realised utilisation curve and the optimisation potential. The licensing team uses the telemetry as the sizing input for the EDP / MACC / EDP envelope, the RI / Savings Plan / CUD coverage ratio and the PTU commitment.

The Admodum methodology runs both layers in sequence. The FinOps optimisation pass establishes the realised consumption baseline. The licensing design pass sizes the commitment envelope against the baseline plus a defensible growth curve. The combined output sits below the rate-card and inside the realised utilisation envelope.

Section x

Reading list.

The pillar groups cloud commentary into ten sections above. The spoke band below opens the forty named articles inside the cluster, each one a deep-read on a specific cloud-commercial mechanic. The white papers below sit alongside the pillar as the methodology deliverables; the practice pages sit alongside as the engagement entry points.

A short follow-up checklist for the reader who is closing this page: visit the AWS, Microsoft or Google Cloud practice for the engagement entry point; visit the AWS knowledge hub for the aggregated reading; request the cloud-commitment papers (AWS EDP, Azure MACC, Google Cloud EDP, AWS Bedrock, AI Vendors PTU); or open a private conversation with a senior Admodum cloud advisor through /contact/.

Cluster IX · Forty cloud articles

Deep reads inside the pillar.

Forty named articles inside the cloud cluster. Each one is a deep-read on a specific cloud-commercial mechanic, written from the buyer’s seat.

i.
The AWS EDP construct in plain language
Tenor, discount curve, eligible-service breadth and ramp design.
Read →
ii.
Reading the AWS Private Pricing Addendum
The clauses that matter and the standard wording the buyer should contest.
Read →
iii.
AWS Marketplace inside the EDP
When Marketplace transactions count against the commitment envelope.
Read →
iv.
RI versus Savings Plans on AWS
The two instrument families and the right coverage ratio.
Read →
v.
EDP ramp design
Year-one, year-two and year-three commitment shape.
Read →
vi.
AWS Bedrock: on-demand versus PTU
The break-even utilisation curve per model family.
Read →
vii.
EDP renewal posture
The six-month timeline ahead of the renewal window.
Read →
viii.
Running Oracle on AWS under BYOL
The Oracle Cloud Authorisation read against AWS EC2.
Read →
ix.
The Azure MACC construct in plain language
Tenor, discount, eligible spend and Marketplace inclusion.
Read →
x.
Azure Hybrid Benefit, considered
Windows Server and SQL Server portability into Azure.
Read →
xi.
Azure Reservations versus Savings Plan for Compute
The two instrument families and the right coverage mix.
Read →
xii.
Azure OpenAI Provisioned Throughput Units
Sizing the PTU envelope against trailing token consumption.
Read →
xiii.
MACC renewal posture
The six-month timeline and the EA-renewal interaction.
Read →
xiv.
MACC inside the Microsoft EA
The packaging of the cloud commitment with the wider EA renewal.
Read →
xv.
Azure Arc and the hybrid licensing position
Reading the management surface against the licensing surface.
Read →
xvi.
Microsoft Fabric pricing
SKU consumption, capacity reservations and the analytics envelope.
Read →
xvii.
The Google Cloud EDP construct
Tenor, eligible services and the discount stack with CUDs.
Read →
xviii.
Google Committed-Use Discounts
Resource-based versus spend-based CUDs and the right blend.
Read →
xix.
Vertex AI pricing and Gemini commitments
Reading the PTU equivalent inside the Google generative-AI stack.
Read →
xx.
GCP EDP renewal posture
The six-month timeline and the renegotiation moves.
Read →
xxi.
Multi-cloud portability, considered
The abstraction interface and the routing protocol across hyperscalers.
Read →
xxii.
EDP versus MACC versus GCP EDP
Side-by-side reading of the three commitment instruments.
Read →
xxiii.
Cloud egress economics
The hidden cost line and the architectural mitigation.
Read →
xxiv.
Storage tiering across the three clouds
Cold, archive and the lifecycle policy that pays for itself.
Read →
xxv.
FinOps versus licensing
Two practices, one consumption telemetry, distinct outputs.
Read →
xxvi.
Right-sizing as a licensing input
How the FinOps pass changes the commitment-sizing output.
Read →
xxvii.
SQL Server on the three clouds
License Mobility, Azure Hybrid Benefit, the BYOL boundary.
Read →
xxviii.
Windows Server on the three clouds
The Software Assurance lever and the AWS dedicated-host carve-out.
Read →
xxix.
Red Hat subscriptions on the three clouds
Cloud Access programmes versus marketplace subscriptions.
Read →
xxx.
SAP on AWS, Azure and GCP
RISE deployments, BYOL, infrastructure interaction.
Read →
xxxi.
Databricks commitments across the clouds
DBU commitments and the marketplace passthrough.
Read →
xxxii.
Snowflake credit commitments
Capacity contracts, the marketplace path and the EDP read.
Read →
xxxiii.
The cloud renewal cycle
A twelve-month buyer-side timeline that survives the publisher cadence.
Read →
xxxiv.
Cloud-vendor quarter-end and the deal calendar
The point at which the account team becomes commercially flexible.
Read →
xxxv.
Reading the cloud account team
Named accounts, regional structures, escalation paths.
Read →
xxxvi.
The cloud BATNA
What credible alternatives look like across AWS, Azure and GCP.
Read →
xxxvii.
Recovering from an over-committed EDP
The mid-term re-sizing rights and the workload-shift options.
Read →
xxxviii.
Reading cloud billing anomalies
The audit posture and the credit-recovery moves.
Read →
xxxix.
Cloud commitments inside an M&A transaction
Transferability, carve-out posture and the publisher consent gate.
Read →
xl.
Cloud sustainability reporting
Reading the publisher carbon dashboards against the procurement story.
Read →
Engage

Speak with a cloud senior advisor.

A senior Admodum cloud advisor will run the methodology through with your CIO, CFO, procurement team or FinOps lead on a private call. The engagement runs as fixed fee, contingency or annual retainer. EDP, MACC and EDP renewals route into the Renewal Programme.

Independence
Admodum is not a partner, reseller, or affiliate of Amazon Web Services, Microsoft, Google Cloud or any other software vendor. No reseller margin, no referral commission, no audit-subcontract relationship.