Sales Snapshot
~2,600
VMs in Scope
107 VMware hypervisors across 2 DCs — 1,516 Prod (NYC2 East) + 1,071 Non-Prod (SJC2 West)
408
IBM Power LPARs
26 IBM Power hosts — iSeries/AS400 workloads assessed in parallel via Connectria (Advisory IBM partner)
$4.25M
AWS As-Is (Annual)
No discounts/optimization. Extrapolated from CloudScape 100-license sample to full ~2,600 VM estate
$7.61M
Azure As-Is (Annual)
Higher Azure as-is due to Windows-heavy estate; EA + CSP discounts close the gap significantly
$2.3M
AWS Optimized (Yr 3)
46% reduction from as-is through RIs, rightsizing, EDP discount, and storage tiering
$4.0M
Azure Optimized (Yr 3)
47% reduction from as-is through CSP (3-yr partial upfront), Hybrid Benefit, Azure discount
The Unique Angle — Why Global Consumer Products Company Matters as a Reference
Global Consumer Products Company is one of the largest consumer products companies in the world — owner of consumer brand, Sharpie, outdoor brand, Ball, Calphalon, Yankee Candle, and 80+ other brands. Their IT estate reflects what decades of acquisition-driven growth looks like at scale: two production-grade VMware datacenters, 26 IBM Power hosts running iSeries workloads that go back to the Jarden/JAH Global era, and a mixed application stack anchored by SAP, Oracle EBS, and PKMS Warehousing alongside hundreds of custom .NET applications. The engagement was a full Cloud Strategy and TCO Assessment — dual-cloud financial modeling (AWS and Azure), server inventory analysis iterated through five versions, a 5-year cost optimization roadmap, and a strategic plan for the IBM Power platform that x86 cloud tooling can't address natively.
What made this engagement technically complex was the combination of scale, heterogeneity, and existing vendor relationships. Global Consumer Products Company was already a Advisory hosting customer (NA DC1/DC2 co-location plus a separate London charge) — the question was not "should we go to cloud?" but "which cloud, on what timeline, and what does the 5-year financial picture look like compared to our current on-prem contract burn-down?" We modeled that scenario explicitly: year-by-year on-prem contract expiry against rising cloud run rates, optimized for Reserved Instance coverage, rightsizing, and enterprise discount programs.
Use this when selling to: Any large enterprise with a multi-DC VMware estate and legacy IBM Power workloads in the mix — common in manufacturing, CPG, retail, and distribution. Also strong for clients who already have Advisory hosting and need to understand the financial crossover point to cloud. The multi-iteration inventory methodology (5 passes) and the extrapolated CloudScape analysis approach are directly applicable to any engagement where full estate data isn't available on day one.
IT Estate — What Was In Scope
~2,600
x86 VMs Total
DC2 Prod: 1,516 VMs (NYC2 / East region) · DC1 Non-Prod: 1,071 VMs (SJC2 / West region)
107
VMware Hypervisors
VMware vCenter-managed infrastructure; multiple vCenters across DC1 and DC2 — excluded from cloud pricing model (not needed in cloud)
26
IBM Power Hosts
~408 LPARs — IBM iSeries/AS400 workloads; modeled separately via Connectria (Advisory IBM Power partner)
3.56 PB
Block Storage (Total)
1.89 PB used · Dell EMC Isilon for file; Cohesity for backup — both assessed for cloud-equivalent tiering
2.63 PB
File/Object Storage
1.12 PB used · Storage tiering and modernization identified as Year 1 cost optimization lever
100
CloudScape Licenses
Randomized sample across DC1 & DC2; TCO extrapolated to full estate using scale factor methodology
Application Stack — Enterprise Workload Complexity
ERP / Finance
SAP
Enterprise SAP suite — BW, ECC, and satellite systems. Dedicated HANA and AnyDB instances identified. Cloud path assessed for SAP on Azure (preferred Microsoft/SAP partnership) and AWS (SAP certified instances). Licensing implications modeled. 3-year partial upfront CSP favored for SAP Windows+Linux blend.
ERP / Finance
Oracle EBS
Oracle E-Business Suite — financials, supply chain, and HR modules. BYOL and License Mobility options assessed. Oracle on EC2/Azure VM paths modeled. Database servers separated from standard compute inventory for dedicated licensing analysis and lift-and-shift instance sizing (all DBs mapped to IaaS vs. RDS per scope).
Warehouse / Logistics
PKMS Warehousing
Manhattan Associates PKMS warehouse management system — legacy Windows-based deployment supporting distribution operations. Assessed for lift-and-shift to EC2 / Azure VM. Critical dependency mapping for distribution network operations flagged as a pre-migration requirement.
Legacy Platform
IBM iSeries / AS400
26 IBM Power hosts, ~408 LPARs — not addressable via x86 cloud native migration. Assessed via Connectria (Advisory's IBM i managed services partner). Strategic options: IBM i Managed Cloud (Connectria), modernization to .NET/Java, or co-location extension. Strategic plan document included in SOW deliverables.
Custom Development
Custom .NET Applications
Large portfolio of custom Windows .NET applications spanning acquired brands — Jarden-era and Global Consumer Products Company-era systems running side-by-side. Windows Server 2008–2022 mix. Hybrid Benefit applied to Azure costing for this estate. EOS risk (2008/2012 R2) flagged as migration accelerator.
Support / Tooling
Infrastructure & DevOps
Cisco networking (Smartnet), VMware ELA, SUSE Linux Enterprise, Citrix, Cohesity backup, Dell EMC Isilon, LaserVault document archival, and multiple monitoring/management tools — all mapped to current contract expiry dates (2021–2025) in the 5-year financial model to identify the contract burn-down crossover point to cloud.
AWS vs. Azure — Dual-Cloud Financial Model
Amazon Web Services
DC1 Non-Prod (Oregon) As-Is$1.60M/yr
DC2 Prod (Virginia) As-Is$2.45M/yr
Support (5%)$203K/yr
Grand Total As-Is$4.25M/yr
Optimized Year 3 Total$2.30M/yr
Optimized Year 4 Total$2.26M/yr
Optimized Year 5 Total$2.23M/yr
5-Yr Reduction vs. As-Is~46%
Regions: AWS Oregon (DC1 West) + AWS Virginia (DC2 East). Optimization levers: EDP Discount (10%), RI/3-yr CSP (up to 75% compute), Rightsizing (20%), Modernization (20%), Storage tiering (10%/yr).
Microsoft Azure
DC1 Non-Prod (West US) As-Is$2.70M/yr
DC2 Prod (East US) As-Is$4.35M/yr
Support (8% Unified)$564K/yr
Grand Total As-Is$7.61M/yr
Optimized Year 3 Total$4.01M/yr
Optimized Year 4 Total$3.98M/yr
Optimized Year 5 Total$3.96M/yr
5-Yr Reduction vs. As-Is~47%
Regions: Azure West US (DC1) + Azure East US (DC2). As-Is is higher due to Windows-heavy estate; Azure EA discounts (up to 45%), Hybrid Benefit, and 3-yr CSP Partial Upfront materially close the gap by Year 3. Support modeled at 8% Unified (Azure + O365).
All figures are x86 VM estate only (DC1+DC2). IBM Power/iSeries is modeled separately via Connectria. Costs extrapolated from 100-license CloudScape sample using 2-batch 7-day collection across DC1 and DC2. As-Is benchmark uses 150% 95th-percentile sizing to ensure headroom; rightsizing reduces this in Year 1–2. Year 1–2 optimization partial due to 80% RI coverage guideline (to avoid overcommit on a first-year migration). IBM TCO will merge into final analysis.
5-Year Optimization Strategy — Levers Applied
Year 1–2
Reserved Instances & CSP Commitment
Year 1: 80% of compute covered under RI/CSP (conservative to avoid overcommit on first migration cohort). 3-Year Partial Upfront CSP blends Windows + Linux. Compute savings of up to 75% vs. on-demand. AWS: EDP discount (10%); Azure: EA discount (12.5%).
Year 1–2
Rightsizing — 150% to Actual
As-Is benchmark intentionally uses 150% / 95th percentile to ensure sizing headroom. Majority of instances can be downsized in Year 1–2 once actual utilization is confirmed post-migration. 20% compute reduction modeled for rightsizing pass. CloudScape performance data drives instance selection.
Year 2–3
Instance Generation Modernization
AWS x86 → ARM (Graviton) / Azure x86 → newer generations: 20% additional compute reduction. Non-production workloads (DC1) are prime targets — lower migration risk, faster modernization cycle. SAP and Oracle workloads excluded (certifiable instance constraints).
Year 1+
Storage Tiering & Modernization
2.63 PB File/Object + 3.56 PB Block — significant opportunity to move cold/archival data to S3 Glacier / Azure Archive. 10% storage cost reduction per year modeled from storage generation modernization and tiering. Dell EMC Isilon and Cohesity direct equivalents identified for cloud-native replacements.
Year 1–3
Azure Hybrid Benefit (Windows)
Large Windows Server estate (custom .NET, SQL Server, Citrix, SUSE) across the acquired brand portfolio. Azure Hybrid Benefit applied to Windows VM licensing — removes Windows OS license from VM cost for license-mobile SA-covered servers. Significant per-VM savings on the Windows majority of the estate.
Ongoing
Enterprise Volume Discounts
At Global Consumer Products Company's estate scale (~$4–7M/yr), both cloud providers offer Private Pricing. AWS EDP (Enterprise Discount Program) up to 15%; Azure Enterprise Agreement with bundled M365/O365 provides context for Azure discount negotiation up to 45%. Modeled conservatively (AWS 10%, Azure 12.5%) in the base case.
Methodology — How We Built the TCO
- CloudScape sampling (100 licenses): Advisory provided 100 CloudScape discovery licenses at no cost. Global Consumer Products Company selected the server cohort. 2 batches of 7-day collection periods were run across DC2 (Prod) and 1 batch across DC1 (Non-Prod). CPU, memory, network, and storage I/O metrics collected at 95th percentile.
- Scale factor extrapolation: CloudScape sample data extrapolated to cover the full ~2,600 VM estate assuming the sampled servers are representative of the broader population. vCenter management VMs excluded (not needed in cloud). Scale factor applied per DC separately (DC1/DC2 have different workload profiles).
- Server inventory iterations (v1.0 → v1.4): Five passes of the server inventory as discovery data improved — OS validation, Oracle/SAP workload separation, rightsizing application, client validation, and final sign-off. Each pass fed updated inputs into the AWS and Azure TCO models.
- On-prem contract burn-down model: Global Consumer Products Company provided their full IT infrastructure contract register (CapEx/OpEx) with expiry dates (2021–2025). All contracts mapped to DC1/DC2 and IBM split. Year-over-year on-prem cost trajectory modeled as contracts expire — revealing the financial crossover point where cloud becomes clearly cheaper even without optimization.
- As-Is benchmark (150% / 95th percentile): Cloud instances sized to 150% of peak (95th percentile) observed utilization — conservative by design. This ensures the As-Is cost comparison is defensible and the rightsizing savings in Year 1–2 are demonstrably real rather than assumed.
- IBM Power handled via Connectria: iSeries/AS400 LPARs cannot be directly migrated to x86 cloud. IBM Power TCO provided by Connectria (Advisory's IBM i partner) and merged into the final consolidated analysis. Strategic options (IBM i Managed Cloud, modernization to .NET/Java) delivered as a separate workstream in the Cloud Strategy Report.
- Dual-cloud final report: Three TCO model iterations (v1.0, v1.1, v1.2 for AWS+PC; v1.0 for Azure+PC) reflecting updated inventory inputs. Final deliverables: Cloud Strategy Report (strategy assessment + IBM iSeries plan), TCO Analysis Report (AWS and Azure monthly run rate, 1/3/5-year ROI).
Forward Opportunity — What Follows the Strategy
Wave 1 Migration
DC Exit — Non-Prod First
DC1 (SJC2 Non-Prod, 1,071 VMs) is the natural Wave 1 candidate — lower migration risk, faster feedback cycle, and significant run-rate reduction once the Advisory co-location contract for DC1 is terminated. Migration SOW with Advisory Factory or cloud migration services is the direct follow-on from the TCO strategy.
IBM Power Modernization
Connectria Managed iSeries
The 26-host IBM Power estate (408 LPARs) has no x86 cloud path. Connectria IBM i Managed Cloud is the Advisory-family solution — same commercial relationship, no re-procurement. This is a parallel track that doesn't block x86 migration. Connectria IBM i cloud likely reduces on-prem IBM hardware, software, and Meridian IT maintenance costs.
Managed Cloud Services
Post-Migration Operations
At ~$2.3M–$4M/yr cloud run rate, a Advisory Managed Cloud layer (monitoring, patching, security operations, FinOps cost governance) has clear ROI vs. building internal cloud ops capability at Global Consumer Products Company's scale. AWS Managed Services or Advisory Azure Managed are the products; the TCO analysis already established the cloud run-rate foundation for commercial conversations.
Talk Track — Using Global Consumer Products Company as a Reference
Large Enterprise Multi-DC Exit, CPG / Consumer Goods, and AWS vs. Azure Fence-Sitters
"Our estate is too large and complex for a quick cloud assessment — we have thousands of VMs, legacy IBM systems, and SAP. We don't know where to start."
The Global Consumer Products Company engagement is exactly this scenario. Global consumer products company — 2,600+ VMs across two datacenters, 26 IBM Power hosts running iSeries, SAP, Oracle EBS, PKMS warehousing, and hundreds of custom .NET applications. Their estate was a post-merger patchwork: JAH Global systems (pre-Jarden acquisition) running alongside Global Consumer Products Company corporate systems, different VMware clusters, different Cisco networking configurations. We ran the engagement in two parallel tracks: a CloudScape-based x86 inventory and TCO model for AWS and Azure, and a separate IBM Power assessment via Connectria. The x86 track used 100 CloudScape licenses — a representative sample from their 2,600 VMs — collected over two 7-day periods and then extrapolated to the full estate with a documented scale factor. We iterated the server inventory five times as discovery data improved. The deliverable was a defensible 5-year financial model — "here is what you spend on-prem year by year as contracts expire, here is what AWS or Azure would cost at three optimization levels, here is the crossover point." That's the model that gets a cloud migration on the executive agenda.
"We're not sure whether AWS or Azure is right for our Windows-heavy environment. Azure looks more expensive on the surface."
This is exactly the question the Global Consumer Products Company TCO addressed, and the answer is nuanced. On a pure as-is, lift-and-shift basis, Azure came in at $7.6M/year vs. AWS at $4.3M/year for the same ~2,600 VM estate. That gap alarmed the team. But it's almost entirely a function of two factors: first, Azure's list pricing for Windows VMs without Hybrid Benefit is materially higher than AWS; second, Azure's support tier (Unified Support at 8%) vs. AWS (5%) adds a structural cost difference. When you apply Azure Hybrid Benefit to the Windows estate, a 3-year CSP commitment, EA discounts, and rightsizing, the optimized Year 3 cost for Azure is $4.0M vs. AWS at $2.3M — the gap narrows but AWS remains lower for this estate. If the client has a Microsoft ELA and significant O365/Dynamics 365 spend, the EA bundling and consolidated support discount can shift the math materially. The answer isn't "which cloud is cheaper in a spreadsheet" — it's "which cloud has the most contractual leverage given your current Microsoft relationship."
"We already have a Advisory relationship for our datacenter hosting. Does that complicate a cloud migration conversation?"
Global Consumer Products Company was already a Advisory customer — NA DC1/DC2 hosting and a separate London hosting contract. The cloud strategy engagement was additive to, not in conflict with, that relationship. We built the on-prem cost model to include the Advisory hosting fees as part of the current-state cost structure — which actually accelerates the cloud conversation, because clients can see the exact contract renewal dates and expiry years where their on-prem cost line drops. When the Advisory co-location for DC1 renews, that's the moment to migrate DC1 to cloud instead. The migration SOW follows the TCO naturally: "you've already decided cloud is the destination, you know the timeline, now let's plan Wave 1." The existing Advisory relationship means there's trust already in the room — we're not competing with ourselves, we're upgrading the engagement from hosting to advisory and managed cloud.
Context & Reference Notes
Client: Global Consumer Products Company Inc. — global consumer products company, owner of 80+ consumer brands (consumer brand, Sharpie, outdoor brand, Ball, Calphalon, Yankee Candle, and others). Corporate headquarters Hoboken, NJ. Legacy systems carry JAH Global / Jarden Corp references (acquisition completed 2016). Anonymized as "Global Consumer Products Company" or "Enterprise CPG" in external references. Engagement: Cloud Strategy & TCO Assessment, SOW executed December 2021 (SOW_-_Cloud_Strategy_and_TCO__-_Global Consumer Products Company_121321). Engagement team: Paul Herrick (Enterprise Client Executive), Abhinav Mishra (Account Principal), Kash Humayun (Engagement Manager), Raj Ravada (Advisory Consultant), Lokesh Taneja (FinOps Consultant). Deliverables: Cloud Strategy Report (strategy assessment + IBM iSeries strategic plan), TCO Analysis Report (AWS and Azure monthly run rate, 1/3/5-year ROI). Multiple working file iterations: Server Inventory Analysis v1.0–v1.4, AWS+PC Report v1.0–v1.2, Azure+PC Report v1.0, TCO Case drafts, Azure Migrate import template. IBM Power partner: Connectria (Advisory IBM i managed services partner). Use "Global Consumer Products Company" or "Enterprise CPG" in external references unless otherwise confirmed with account team.