Peaks and Valleys: Software Licensing for Elastic Cloud

Grazed from Wired.  Author: Richard Northing.

One of the most compelling aspects of cloud computing is its elasticity. With cloud provisioning models such as SaaS, PaaS and IaaS, organizations can scale up capacity to support business growth and/or temporary spikes in demand much more easily than they can with traditional IT architectures.  They can also scale down capacity when business slows and/or spikes are over.

Just as important in an uncertain and unstable global economy, the cloud mitigates the risk of being locked into fixed IT costs in the vent of a business downturn. So even if the business continues to grow, the rental model offered by the cloud still feels more prudent than the capital model...

Discussions about cloud elasticity, however, tend to focus on infrastructure—in particular compute, storage and network. Infrastructure virtualization and automation are enabling vendors, cloud service providers and enterprise IT organizations alike to achieve a great deal of flexibility when it comes to supporting peaks and valleys of demand.

In fact, a general best practice of “buy the base, rent the spike” is beginning to emerge whereby IT organizations are driving down the cost of supporting well-known workloads internally with private on-premise cloud architectures while shifting more dynamic, less predictable workloads to private and public off-premise cloud resources.

Conversations about the efficacy of this model, unfortunately, quickly deteriorate when the issue of software licensing comes up.  That’s because ISVs have not all responded with equal speed or with similar policies to the phenomenon of the elastic cloud.  Some want their customers to pay for elasticity with a fixed enterprise software license.  Others want to charge for peaks.  And many are still adopting a wait-and-see attitude—even though the move to the cloud is well underway.

None of these responses seem useful for a market that is demanding elasticity as an intrinsic component of any technology buy, rather than as a value-added option. Any failure to address this demand for elasticity will simply drive customers to open-source solutions, to SaaS resources that may be functionally inferior but economically more compelling or to other vendors that understand the new rules of the cloud game.

ISVs therefore need to ask themselves some basic questions about how they want to play in the elastic cloud, including:

  • How should use and software entitlements be properly defined in computing environments where processor cores and end-user devices are no longer fixed assets—but are instead themselves becoming services that can be spun up, spun down, moved and morphed on an as-desired basis?
  • How will revenue and sales compensation be affected if the customer’s cost becomes dynamic?  Are pure-play upfront software licenses and SaaS-style subscriptions the only viable models for ISVs?  Or does the advent of the elastic hybrid cloud call for new elastic hybrid software licensing models?
  • How will cloud-friendly innovation in software licensing have a positive impact on brand value, customer experience and competitive positioning in the long term? Will gains in these areas offset potential short-term losses in revenue and increases in business complexity?

Infrastructure vendors are moving aggressively to offer customers new value propositions around elasticity, economies of scale and the offloading of low-value technology ownership tasks. In doing so, they are breathing new life into markets that were heading towards stagnation. It will be interesting to see how—and if—ISVs successfully follow suit.

In fact, being the first to meet the needs of the market with the right software licensing model that monetizes the cloud might just be the differentiator needed to push your solution over the top and give you the competitive advantage.