Market Research Logo

What's Price Got to Do With It? Why Enterprises Struggle to Calculate and Compare Costs to Run a Cloud Workload

What's Price Got to Do With It? Why Enterprises Struggle to Calculate and Compare Costs to Run a Cloud Workload

Despite the well-publicized cloud price wars, many enterprises have little understanding of what they pay for cloud services as compared with other infrastructure options, and (more importantly) what they could be paying. While three-quarters of cloud users believe that the cloud is helping them reduce costs, few can articulate how much they are saving or expect to save.2

Users who are looking to save often gravitate toward the mass market cloud providers that have dominated the “cloud pricing wars”—Amazon Web Services (AWS), Google, and Microsoft—rather than the high-touch, enterprise-grade service providers (such as IBM SoftLayer, Rackspace, and Internap).3 The mass market providers’ frequent and conspicuous rate decreases make it easy for would-be buyers to assume they are getting the best deal. Often that assumption is valid.

Yet, in recent months, enterprise-grade competitors have started taking aim against AWS and its ilk with price- and performance-based analyses showing specific use cases where their own services cost less and/or perform better than comparable competitive services. Stratecast has had an opportunity to review some of their calculations, and we find many of the analyses to be credible. Does that mean that IBM, Internap, CenturyLink and other enterprise-grade cloud service providers are the new “low-cost” providers? Not at all. Instead, the findings simply highlight the need for enterprises to do their homework in selecting the right service provider and the right service type for each workload—a recommendation that the mass-market providers tend to agree with.

But putting the onus on enterprises may be unfair. The lack of standardized cloud “units,” the dizzying array of service components and options, and the absence of a baseline for understanding what cloud costs should be, all make it difficult for enterprises to make an informed assessment regarding price.

  • Introduction1
  • The Difficulty in Determining the "Best" Cloud Deal
    • Few enterprises understand their costs for running workloads in the traditional data center.
    • IaaS is not a consumption-priced model.
    • Hardware matters for performance.
    • A la carte or bundled? The best deal depends on how services are used.
  • What Goes into IaaS Service Pricing
    • Table Smallest available Linux Unit, On-Demand, Hourly Pricing
    • Workload-Optimized Units
    • Term and Volume Commitments
    • Private, Dedicated, and Bare Metal Clouds
  • Apples to Apples: Comparing Cloud Costs across IaaS Providers
    • Burstable General Compute Workload (e.g., Test & Dev)
    • Compute-Intense Workload (e.g., eCommerce)
      • Table Compute-intense, 8 Core, 1 TB, US East, Three-Year Term
      • Table Compute-intense, 8 Core, 1 TB, 4000 IOPS, US East, Three-Year Term
    • Data Transfer Intense Workload (e.g., mobile app)
      • Table Outbound Data Transfer Fees among Cloud Service Providers
  • The Risk: Being Penny-Wise and Pound-Foolish
  • What Enterprises Should Do
  • How Cloud Service Providers can Help
  • The Last Word

Download our eBook: How to Succeed Using Market Research

Learn how to effectively navigate the market research process to help guide your organization on the journey to success.

Download eBook

Share this report