Since cloud computing entered the mainstream, the market has seen dozens of price reductions, often from a single cloud service provider. Competition in pricing and improved economies of scale are expected to lower overall cloud service pricing, but the dizzying pace of price reductions suggests there's more at work here. Understanding the factors behind pricing reductions helps cloud buyers get the best ROI out of their clouds -- now...
and in the future.
Providers may drop prices for competitive reasons, but only to a point that doesn't compromise a profit. In an immature market such as cloud computing, providers most often drop prices to increase total addressable market (TAM). Reducing cloud prices reduces costs, and this makes it easier to negotiate deals.
In a market with falling prices, consumers must consider pricing protections on their service contract. Ensure your service provider will adjust the contract if reductions occur so you're not locked into a higher price. This adjustment is common practice among larger cloud service providers, so you may need to look a bit deeper into pricing to find new ways to benefit.
Cloud buyers should also prepare project justifications so they can update them to reflect changing prices. A surprisingly large number of enterprises report they don't revisit cloud project considerations when service provider prices change; many enterprises don't even monitor provider pricing to know when prices change. If a cloud project approval is withheld because it doesn't meet corporate targets for return on investment (ROI), project materials should be added to a watch list and reviewed each time vendor prices change.
Moving from Capex to Opex
Since providers want to increase TAM by reducing prices, buyers may overlook the potential benefits in cloud projects beyond just simply cost savings. Current cloud projects that succeed are likely low-use applications running on dedicated servers, IT applications for small and medium-sized businesses with limited technical support, or both. Moving up the value chain to mainstream IT resources means moving beyond the traditional cost-of-server value proposition -- and moving from a capital expense (Capex) to an operating expense (Opex).
Opex costs for applications include hardware management; however, businesses report operating costs for software are actually three to five times higher than hardware expenses. Companies with higher-than-usual support costs find it's easier to justify cloud computing, providing that their cloud service actually affects operational costs.
Infrastructure as a Service (IaaS) replaces servers with virtual servers but still requires users to maintain the software images for loading into the cloud. Therefore, IaaS displaces only the hardware portion of Opex.
On the contrary, adopting Platform as a Service (PaaS) or Software as a Service (SaaS) may improve an enterprise's cloud justification. Because these platforms move system software (PaaS) or all software (SaaS) into the cloud, operational costs for software are reduced significantly. That reduction could mean the difference between a cloud project making its business case or being unapproved because it doesn't appear to offer enough benefits.
Data storage choices for cloud apps affect your bottom line
When cloud price reductions qualify new applications as cloud candidates, it's important to review, the data storage and access costs of the cloud carefully. Sometimes highly promoted price reductions for compute instances aren't matched by corresponding cuts in data storage and access costs, which limits the effects of cloud price changes on total service pricing.
Users who move regularly accessed applications to the cloud because of lower cloud prices may find data costs are prohibitive when these projects move from pilot runs to production-level runs. Sometimes cloud pricing changes enable users to deploy more reliable or predictable forms of cloud services, including dedicated instances and geographic diversity.
It may be valuable to re-examine cloud project proposals that require higher performance or high availability to see if projects once unfit for the cloud because of cost may now work.
Cloud-pricing models continue to evolve
Most users wonder how long these cloud price cuts can continue. Some companies have considered factoring in 15% annual reductions in cloud pricing when doing their five-year total cost of ownership calculations. This is likely an overly optimistic assumption, particularly as increased cloud consumers begin to move more mainstream applications to the cloud.
In any event, economies of scale don't grow at a linear rate with the size of the cloud; they grow at an ever-decelerating rate. Therefore, the current price declines cannot be sustained indefinitely. The current trend toward very low use of cloud applications allows operators to pack more onto a single cloud server than expected, which means more can be sustained as projects mature.
Additional price declines may be in the cards. Cloud compute costs have fallen faster than cloud data costs; data pricing may be the next target as cloud operators try to strike a balance between prices attractive enough to move buyers and earning enough profit to sustain their own businesses.
Cloud service providers will likely introduce special data services designed for backup and archiving, which will have restrictions on access frequency but offer more attractive pricing. This kind of price change -- one that affects cloud services selectively -- requires more analysis but may offer special benefits for users whose applications qualify.
Pricing models continue to evolve and it's important to watch the trends carefully for the best near- and long-term deals. It's also important to recognize that industry consolidation could easily reverse price reductions in cloud services. Enterprise IT needs to carefully monitor applications with ROI tied to their rate of return for changes in both reward and risk.
About the author:
Tom Nolle is president of CIMI Corp., a strategic consulting firm specializing in telecommunications and data communications since 1982.