In June 2012, Everest Group released a flagship report titled “Impending Contract Renewals – Back to the Future.” The report covers in extensive details trends around outsourcing contracts that are reaching end of term during the period between April 2012 and September 2013. Analyses performed include industry, geography and service provider category trends.
Everest Group’s analyst Soumit Banerjee shares some of the key headlines from the report.
The market conversation around enterprise adoption of private, public and hybrid cloud models has been surprisingly light on facts. Data, surveys and analysis tend to focus either on predicting overall market sizes for cloud services (which often strain credulity of even the most ardent cloud supporter), or on high level surveys around planned cloud adoption and perceived issues. While interesting from a broad market perspective, they provide little insight for IT executives facing hard choices around cloud migration. Decision makers are faced with little hard data on the use cases that are actually being implemented in the enterprises and the value that they’re generating. This gap creates challenges not just for enterprise CIOs but also for cloud service providers and consultants.
In conjunction with Cloud Connect and UBM TechWeb, Everest Group is excited to announce a new tracking survey focused on better understanding where the “rubber is hitting the road” with enterprise cloud adoption. Targeted at enterprises and vendors alike, our Enterprise Cloud Adoption Survey will focus on identifying global enterprise cloud adoption trends and patterns and where enterprises are seeing value today from the cloud. Our survey will help readers gain visibility and insight into questions such as:
What are the use cases that are driving adoption of SaaS, PaaS, IaaS and private cloud?
How are cloud adoption patterns and uses cases are differing by vertical? By geography?
What cloud infrastructure models are most frequently being deployed (private, public, hybrid)?
What cloud management platforms are gaining traction in the enterprise? Where are open source options (OpenStack, CloudStack) being adopted?
How does the value being delivered by cloud deployments compare to expectations?
We think some of the more interesting insights will come from seeing how these responses change and trend over time. Our goal is not to provide just a one-time shapshot of adoption, but to conduct an ongoing survey several times a year to surface key trends and patterns. The results from our first joint survey will be announced in conjunction with Cloud Connect Chicago, to be held September 10-13 at the Hyatt Regency O’Hare.
Given how much of the typical large enterprise IT budget is consumed by ERP, we’re not surprised to find a growing curiosity among many CIOs to understand how cloud delivery models could reduce costs. On the surface, you wouldn’t think that production ERP applications would be at the top of the list for cloud migration. ERP apps are mission critical, complex and highly customized, often with significant data security and compliance requirements.
That’s why we think one of the more interesting, underreported stories in cloud are the examples of large enterprises that have migrated existing ERP environments to private, hybrid and community cloud models. We’re actually finding quite a number of quite interesting, global scale ERP cloud deployments particularly among SAP customers. Why SAP? While Oracle is obviously the other large enterprise ERP heavyweight, as we’ve discussed here before, Oracle’s licensing policies are creating roadblocks for customers to migrate to even virtualized models, let alone private or public clouds.
The market for SAP cloud services is surprisingly robust with at least 10 major service providers that deliver SAP ERP capabilities via managed or host private or hybrid cloud models, including IBM, T-Systems, Fujitsu, Accenture CSC, CapGemini and others. T-Systems alone already supports 500 customers and 1.9 million SAP users via cloud-based models. Not surprisingly, most of these service providers started by originally providing SAP hosting services and have since extended their offerings. What’s the customer value proposition for SAP in the cloud?
Cost variablization – given the significant capex investments associated with SAP deployments and upgrades, cost variability is central to cloud-based SAP offerings. Nearly all providers offer consumption-based pricing models for SAP cloud services.
TCO reduction – many service providers are claiming the ability the reduce TCO for customer SAP environments by 30+% through the typical cloud levers. Several providers have customer references that have achieved these efficiencies and more in live production.
Flexibility – service providers are touting the ability of cloud-enabled deployments to more rapidly and easily provide new capabilities to users.
Standardization – in conjunction with cloud migration, many enterprises desire to consolidate data centers, rationalize SAP instances and standardize global processes to drive efficiency and flexibility.
Unlike other enterprise cloud use cases focused more on business agility and flexibility, in most cases cost appears to be the major driver of SAP cloud migration. Some of the more interesting examples include:
British American Tobacco (BAT) – just last month BAT announced a seven-year, US$160 million deal with T-Systems to consolidate its current SAP deployments into a single, cloud-based instance by 2016. The deal will enable BAT to variabilize its SAP costs through a usage-based pricing model.
Domino Sugar – leveraging Virtustream’s virtual private cloud platforms, Domino Sugar has been able to reduce SAP costs by over 30%, while actually improving availability and performance for several thousand users. As with BAT, SAP costs are variabilized and based on actual resource consumption.
Shell – to drive standardization, increase flexibility and shift to consumption-based pricing, Shell migrated its SAP environment to private cloud models (delivered by T-Systems) in support of 102,000 global employees across 100 countries.
Other notable enterprise examples include Audi, Freeport McMoran, Siemens, and Suntory.
Why haven’t we heard more about these and other examples? With the exception of IBM, most leading SAP cloud service providers and many of the early enterprise adopters of SAP in the cloud aren’t U.S.-based and are outside of the cloud hype and “echo chamber.” Also, details on many of these deployments tend to be tightly held both by both service providers and customers.
While many segments of enterprise cloud appear to be stuck in pilots and proofs of concept, ERP is surprisingly providing some early examples of large scale enterprise cloud migration.
Less than three years back, there was widespread excitement (and alarm and despondency) in many quarters about the impact of cloud computing on traditional IT outsourcing providers.
Cloud computing was predicted, though not by us, to greatly disadvantage the incumbent players, but as of today, such a prediction is difficult to stand by (just take a look at TCS’s and Accenture’s results since then). Sure, public cloud providers continue to grow rapidly, and the traditional license model is increasingly giving way to the pay-as-you-go paradigm. Yet most leading providers of outsourced IT services seem to be adapting well through a combined strategy of alliances, acquisitions, and in-house cloud solutions. Cloud computing appears to be increasingly well integrated as part of the delivery model for most traditional ITO providers. Consider the following statistics from our recently released report, Enterprise Cloud Adoption: Role of Cloud in Global Services:
In the second half of 2011, approximately eight percent of all ITO/BPO deals serviced by traditional outsourcers (excluding SaaS product companies, and public cloud and hosting providers) included cloud delivery models or platforms within their scope. This is up from four percent in the first half of 2011.
The average total contract value (TCV) of 2011 global services deals with cloud delivery in scope was US$168 million, compared to US$95 million for deals without cloud in scope.
Cloud deals seem to be more transformational in nature, almost at the cutting edge of ITO capabilities if you will. 53 percent of all ITO deals with cloud delivery in scope involved significant infrastructure transformation of test, development, and production environments. Clearly, traditional ITO providers view cloud computing as an important solution component for large, transformational deals.
Cloud computing seems to be helping service providers get access to markets that were previously unprofitable or too complicated to serve. Approximately 38 percent of all global services contracts with cloud in scope were awarded by enterprises with less than US$500 million in revenues. And government and non-profit sectors together account for 20 percent of all global services deals with cloud delivery in scope.
Clearly, there’s a big pot of gold somewhere amidst all these clouds, but what’s interesting to note is that few service providers have all of what it’s going to take to win all of it:
Design and Consulting – Service providers, such as Accenture, with a consulting legacy and orientation are going to have an advantage when it comes to advising clients on how to build their cloud solution from scratch.
Host and Implement – Players like IBM and HP with a deep legacy of asset-based infrastructure transformation will have an advantage in providing these services
Management and Professional Services – Offshore players such as TCS, with their global delivery models, have an advantage in offering the “cloud management” role
The problem is that these activities are seldom commissioned in isolation. This is not something where a best-of-breed approach always works, despite buyers being wary of lock-in risks. The opportunities are tightly coupled, and service providers need intelligence on the characteristics of relevant opportunities as they are torn between focusing on what they have, and plugging the gaps through alliances and acquisitions.
The fact of the matter is that there will be winners and losers, and the market today is too dynamic to predict who will play which part. It will be interesting to see if there are ground-breaking disruptions (e.g., a major public cloud provider making a headline acquisition of a giant system integrator, thereby making its move in the private cloud market, potentially disintermediating a lot of other system integrators, and at one stroke making a deep thrust in the enterprise market) as the stakes get higher. Or an asset-light provider marking a strategic u-turn by investing in physical infrastructure to build its own cloud solution, complete with consulting, system integration, and management services delivered through a global platform?
As discussed here before, a number of different enterprise cloud adoption paths are emerging. These patterns range from “Observers,” who are taking a reactive, wait and see approach to migration to “Transformers,” who are using private and public clouds to drive wide scale IT transformation and modernization programs. Not surprisingly these transformers are the Holy Grail being pursued by many cloud service providers and enterprise IT vendors. The opportunity to drive significant pieces of an enterprise IT environment to cloud environments (private and public) in multi-year transformation efforts creates visions of big services, hardware, and in some cases software dollars.
While many service providers are crafting go-to market strategies around these types of client opportunities, they’re running into an interesting challenge. They’re not finding a lot of Transformers out there yet. Enterprise cloud adoption, particularly for IaaS, is still largely focused on specific use cases or initial pilots. While many CIOs have long-term visions for cloud-centric future state environments, few CIOs are actually doing it today.
So why aren’t we seeing more Transformers in the market? Our experience suggests that in many cases there are a set of tactical (and often mundane) issues preventing CIOs from getting to the cloud more aggressively. While by no means comprehensive, several of the issues we frequently see are:
Licensing handcuffs – legacy enterprise software vendors clearly understand the business model disruption that cloud represents. Not surprisingly, most enterprise software houses are in no hurry to get their customers to the new world. For example, nearly all of Oracle database licensing policies are still based on physical server CPUs. One notable exception is with Amazon AWS, for which Oracle does support a “BYOL” (bring-your-own license) model based on virtual cores; at this time, Amazon is the only cloud service provider certified by Oracle. In addition Oracle software licensing also provides no or limited technical support for major non-Oracle virtualization platforms such as VMware, KVM, Xen and Hyper-V. Needless to say, if you’re a CIO running an Oracle shop (as many Fortune 500 companies are), there are significant constraints to migrating to even private cloud environments. While not every legacy enterprise software vendor has staked out a position as extreme as Oracle, many are still using licensing as leverage to drive clients to preferred models (or keep them there).
Shortage of skills – cloud expertise and experience is hard to find. Without cloud architecture and solution skills, enterprises are finding it difficult to drive wide-scale transformation efforts. While retraining would seem to be the obvious answer, CIOs that have tried going down that path are finding it to be a dead end. As discussed at our Organizational Readiness track at Cloud Connect Santa Clara last February, IT leaders are finding that the cloud paradigm shift is a bridge too far, and that most of their current employees are unable to make the shift. The lack of internal talent, combined with the wariness to trust vendors and service providers, is leading to a real constraint to further adoption, particularly in IaaS and private cloud models.
Analysis paralysis – private cloud provides an interesting example of the proliferation of options facing enterprise IT. Private clouds can be provided in a variety of flavors, with important choices to be made around delivery model (VPC vs dedicated), location (on-premise or hosted), asset ownership (customer or service provider), platform (proprietary vs open source) and, of course, vendor. Given the skills shortage mentioned above, even sophisticated enterprise IT shops are challenged with the variety of vendor and service options in the market, particularly given the pace of change. Of course the recent flare-up of IaaS platform wars doesn’t help make these choices clearer for risk-averse CIOs. The result of too many choices? It’s not uncommon for us to see clients experiencing “vapor lock,” not really knowing what to analyze, let along what methodology to use. Clients are finding the frameworks, methodologies and tools they’ve historically used to make similar decisions in the past aren’t applicable or relevant in the cloud paradigm. As simple as it seems, many of our clients simply don’t know where to get started.
Why isn’t security and compliance on the list? Because in many cases, we’re finding that security and compliance is a red herring that IT is hiding behind. This is not to say that there are not workloads and use cases where security and compliance issues prevent certain public cloud models; however, these situations in reality are the minority. A variety of examples exist of enterprises leveraging the cloud today while still maintaining compliance with PCI, HIPAA and other mandates (most of which are open to auditor interpretation anyway). Best practices, tools and architectures for addressing common security issues are also becoming more prevalent, as are more mature CSP offerings and security practices for common use cases. Net, net: where there’s a will there’s a way, and in most cases if CIOs are truly interested in getting to the cloud, there are secure, compliant ways of getting there.
Overall, we believe that the wave of transformation is coming in the enterprise. Early movers exist and are achieving the promised payoff. Unfortunately the timing and shape of the wave for the mainstream organization is not as clear as those in the enterprise IT world would like, and the pace is being shaped primarily by a set of factors that are largely non-technical and beyond the IT leader’s control.
As we work with our clients to understand the implications of Next Generation IT technologies, it’s clear that large enterprise adoption of public cloud IaaS is progressing more slowly than other types of cloud services (e.g., SaaS, private cloud). When we ask ourselves “why,” we continue to come back to three critical issues:
Vision and reality gap – we continue to be impressed with the sophistication that many of our client IT executives have around how private, public and hybrid clouds can be used to fundamentally transform their IT infrastructures. They then talk to vendors and face the disappointing gap between the state of cloud technologies today and their expectations and requirements (legitimate or not).
Risk aversion – it’s one thing for a CIO to passively support their VP of Sales as they roll out Salesforce.com. It’s quite another to own the decision to migrate critical IT workloads out of the data center to public cloud services. While early adopters are clearly out there experimenting with IaaS, don’t expect your typical Fortune 500 CIO to be eager to get on the diving board and jump in until they have to, or they feel it’s safe.
Market “noise” – just when CIOs think the drumbeat of vendor provider announcements around public, private and hybrid cloud offerings and standards can’t get any louder, someone dials it up a notch. The noise (and uncertainty) is now being amplified even further by the emerging battle around enterprise cloud platforms / operating systems like vCloud and Open Stack (more on this later).
Certainly we’re finding that these issues are reflected in enterprise IaaS adoption patterns that are not quite what many in the enterprise CSP vendor community had hoped for at this point. Namely we’re seeing:
Enterprises growing cloud usage from the “inside out” – nearly all the activity we see in the enterprise market around cloud and infrastructure today is focused around private cloud pilots or full deployments (hosted or on-prem). Rather than experiment with cloud with public service providers, they’re opting to try the model internally first. Some call it “server-hugging,” others a reactive move to keep IT spend in house, and still others a rational response to the current state of technology and services.
Heavy reliance on proprietary enterprise IT vendors – despite their vision, promise and industry support, new open source platforms (and Eucalyptus) have seen limited adoption in enterprise private clouds. While OpenStack has had success with service providers, many CIOs don’t consider it ready for prime-time yet in their data centers. CloudStack has had more success, but enterprise deployments still likely number only in the double digits. Perhaps not surprisingly we see enterprise cloud deployments (private cloud) dominated by VMware and IBM.
Selective, incremental migration of targeted use cases – where we do see enterprise IT migrating to public cloud or hybrid infrastructure models is for very targeted or smaller scale, lower risk use cases. Examples include test / dev environments, backup and archival, websites and batch data analytics. IT is dipping their “toe in the water” with public cloud, and not feeling a compelling need to drive widescale transformation – yet.
So where are we headed?
In general, enterprises are obviously not comfortable with the current risk / return profile associated with public IaaS and hybrid cloud models. We believe one of the few levers that would pull both components of this ratio would be a cloud management platform that would enable true workload portability / interoperability and policy enforcement across private, public and hybrid models. Not surprisingly, competing enterprise cloud vendor platforms, standards and ecosystems are emerging around VMware, Open Stack and Amazon (and to a limited extent Microsoft) to address this market gap. Several major announcements over the past several weeks that have served both to partially clarify and muddy this evolving landscape at the same time include:
The Amazon / Eucalyptus announcement around extended API compatibility for hybrid clouds
The Citrix announcement that they will be breaking away from Open Stack and open sourcing CloudStack to the Apache Software Foundation
HP’s announcement of the Converged Cloud portfolio of public, private and hybrid cloud offerings based on a “hardened” version of OpenStack and KVM.
Most major enterprise IT vendors are still hedging their bets and publicly keeping feet in multiple camps. With the marketing engines in overdrive it’s difficult to understand what commitments vendors are really at the end of the day making to the different platforms. In fact it’s quite instructive to take a look at who’s putting their money where their mouths are when it comes to open source efforts like Open Stack, not just in terms of sponsorship fees but also developer contributions.
Historically IT platform markets end up with a dominant leader and one to two credible challengers that end up with 2/3 to 3/4 of the market, with the remainder shared among niche players. When we take a look at the enterprise cloud operating system or management platform market, we don’t see why it would be any different here, though we’re obviously still a long, long way from the end game.
The critical question in our mind is: Is a cloud platform market shakeout required for enterprise adoption of IaaS to accelerate and hit the tipping point? If so, we could be waiting a long time.
The advent of cloud computing has brought many exciting changes to companies’ IT strategies. One aspect of the cloud that is frequently overlooked, however, is energy efficiency. On the face of it, one might expect cloud computing to be more energy efficient than the alternative. But is it really?
Let’s take a quick look at the three drivers behind increased energy efficiency in cloud environments.
First and most obvious is economies of scale. It’s not rocket science to understand that fixed costs are best allocated among a greater quantity to bring down per-unit cost. Similarly, conducting a benchmarking exercise to measure Power Usage Effectiveness entails significant fixed costs in devoting resources to counting equipment and measuring individual devices’ power consumption. There are certainly economies of scale to be gained in doing this for a larger datacenter than for a smaller one.
The second driver of energy efficiency in cloud environments results from the abstraction of the physical and virtual layers in the cloud. A single physical server running multiple server images will obviate the additional power load from purchasing additional physical servers. Also, if a virtualized environment incorporates redundant server images on different physical boxes, then individual boxes do not need multiple power supplies. The failure of one machine becomes a non-issue when redundancy is built in.
Finally, a datacenter serving cloud clients will have more users from more disparate places, each with different needs. This means that system loads will be more evenly spread throughout each day (and night), which enables the datacenter to average higher system loads and thus more efficient utilization of equipment. Everest Group research shows that individual servers in a cloud datacenter experience three to four times the average load of those in an in-house datacenter.
By now it should be clear that a large cloud datacenter has distinct energy efficiency advantages over a smaller, in-house datacenter. But there are corresponding energy drawbacks to cloud migration that may not be immediately apparent. First, as processing and storage shift to the cloud, energy usage increases. This is primarily from the routers transporting the data over the public Internet; their power use increases with throughput and frequency of accessing remotely stored data.
Also, in a SaaS, PaaS, or simple cloud storage scenario, frequent data access can cause data transport alone to account for around 60 percent of the overall power used in storing, retrieving, processing, and displaying information. At this point, the efficiency advantages gained by the three drivers cited above may be lost due to the extra power required to move the data between the user and the cloud datacenter in which it is stored or processed.
It is true that migration to the cloud can yield significant gains in energy efficiency for certain applications. However, for applications involving high transaction volumes, an in-house data center can provide better energy efficiency.
As power prices become increasingly important in determining data center operating costs, energy efficiency will play a greater role in companies’ cloud strategies.
One of the current mantras that many enterprise cloud enthusiasts are chanting is that “it’s not about cost.” Cloud is all about business agility and flexibility with cost being an interesting side benefit, but not necessarily compelling on its own. Focusing on cost efficiency and TCO is indicative of a stodgy, legacy IT mindset that doesn’t understand the true paradigm shift of cloud.
Nothing could be further from the truth. In fact, we’re finding that some of the more interesting cloud enterprise use cases these days involve leveraging cloud agility to aggressively reduce infrastructure and IT costs.
Take a recent client of ours, a Fortune 500 global energy company seeking to reduce corporate IT infrastructure costs. Its focus was on reducing costs across two primary datacenters that delivered HR, finance, accounting, operations and other applications to business operations across 30 countries. Understanding cloud options for migrating its SAP deployment was a central focus of their effort.
Facing an imminent and significant hardware upgrade cycle, it was more interested in exploring opportunities to reduce costs through traditional IT outsourcing (ITO) vehicles, as well as next generation, cloud-enabled delivery models. Critical objectives included:
Reducing asset ownership
“Variabilizing” its IT cost structure
Outsourcing commodity IT skills
Based on these requirements, our client evaluated potential solution options from nearly 20 service providers, including traditional enterprise IT service providers, cloud service providers (CSPs), offshore ITO vendors and telcos/carriers.
Our client narrowed the field to three potential solution providers, each with different recommendations on where to migrate existing applications and workloads (which were largely in dedicated and virtualized models). Recommended solutions varied not just across cloud delivery model (public vs. private), but also across asset ownership (on-prem private vs. hosted and virtual private):
And what did the client find? As shown below, leveraging a mix of virtual private and public cloud models offered the opportunity to reduce its annual infrastructure costs by over 30 percent! “Provider A,” which suggested migrating approximately 30 percent of the clients’ workloads to public cloud environments ended up with the most compelling business case. While they recommended migrating 80 percent of the workload portfolio to cloud-enabled models, they did recommend keeping the client SAP instances in a traditional, dedicated model.
Some additional observations:
Costs reflect all required migration and replatforming investments
Public cloud costs were indicative of current market pricing generally at the same unit price levels across the period. As shown by the recent AWS price drop of up to 37 percent on reserved instances, this is a very conservative assumption
Efficiencies do not reflect additional potential opportunities from active workload management
So where did the savings come from? Our client found that the savings were driven by four primary levers:
Consolidation and rationalization of underutilized servers
Migration of unpredictable and “spiky” workloads to public cloud models with consumption-based billing
Reduced IT operations and management costs
Defacto outsourcing of maintenance and support to CSPs
We’re seeing similar results across our other clients, who are finding that cloud-enabled delivery models, leveraged correctly, can drive substantial and lasting reduction in IT infrastructure costs.
Maybe cloud and cost efficiency aren’t so boring after all…
Cloud skills are different than traditional IT skills. At the recent CloudConnect conference in Santa Clara, we had a panel discussion on how cloud is changing the CIO’s wish list for new hires.
In this last video blog of the series, Clayton Pippenger, Applications Development Manager at Quest, shares his thoughts on the outlook for hiring cloud skills in the coming years.
In case you are just tuning in, this is the fourth video interview of a series we taped at CloudConnect 2012 in Santa Clara. Everest Group’s Scott Bils chaired the Organizational Readiness track and enlisted an impressive lineup of speakers.
Watch the first video, featuring Francesco Paola of Cloudscaling.
Watch the second video, featuring Simon Wardley of the Leading Edge Forum.
Watch the third video, featuring Erik Sebesta of CloudTP.
At CloudConnect 2012, Everest Group’s Marvin Newell moderated a lively panel discussion on next generation IT governance. The panelists included Thomas Barton, Global Enterprise Architect at Novartis Pharmaceuticals; Jeromy Carriere, Chief Architect at X.commerce; and Erik Sebesta, Chief Architect and Technology Officer at CloudTP.
The panel focused on the governance concept of holding on loosely but not letting go. Though executive buy-in is important for cost-efficient and holistic migration to the cloud, the business unit knows operations and needs the best.
In the third CloudConnect video interview of the series, Erik Sebesta answers the question: How does one balance the decision-making between the executive team and the business unit?
In case you missed the first blog, this is the second video interview of a series we taped at CloudConnect 2012 in Santa Clara. Everest Group’s Scott Bils chaired the Organizational Readiness track and enlisted an impressive lineup of speakers.
Watch the first video, featuring Francesco Paola of Cloudscaling.
Watch the second video, featuring Simon Wardley of the Leading Edge Forum.
Watch the last video, featuring Clayton Pippenger of Quest.