duquesne
Research Duquesne Advisory delivers in-depth analyses of Information and Communications Technologies, their implementations and their markets. Research is based on critical observation of the market by the analysts and their on-going contacts with the vendor community, together with hands-on, practical experience in consulting engagements.

IBM and Public Clouds - The Empire Strikes Back

Cloud Strategies Series – IaaS


With IaaS becoming a mainstream market, the cloud is clearly established as the defining model for 21st century information systems architectures, even though conventional systems and infrastructures will not fade away anytime soon. With this evolution, enter the heavyweights of IT, and in particular IBM.


IBM and Public Clouds - The Empire Strikes Back
Cloud adoption continues to expand into the enterprise. Following Software-as-a-service (SaaS) and private clouds, public Infrastructure-as-a-Service (IaaS), is now seriously gaining momentum with mainstream customers: public IaaS is no longer solely relegated to test and development uses, but now rather routinely considered and evaluated, if not always adopted, for core information system applications.

Public IaaS enables customers to reap the full benefits of the cloud model, particularly in terms of cost-effectiveness, well beyond what can be achieved with private clouds. Until 2012, however, corporate customers – mostly still in a learning and skills acquisition process - had privileged private clouds, notably due to security considerations, for the greater benefit of infrastructure component vendors (including IBM).

IBM: The Cloud as a key Business Transformation enabler

Of course, IBM and most of its usual competitors (be they in the equipment or the services businesses) had long identified the cloud’s long-term potential and already invested in cloud technologies, services and offerings.

IBM, in particular, has long been heralding “Cloud” as one of the four key “rupture technologies” (together with Social Media, Big Data Analytics and Mobility) that enable “smart” reinvention of the ways businesses operate and grow. “Business Transformation” has become the key theme of IBM’s communication and positioning. Such a positioning is in line with IBM’s focus on value-added services, and requires that the company be able to provide all relevant types of cloud infrastructure services as customers adopt them, including public IaaS.

With the rise of IaaS, competition intensifies and vendors seek differentiation to address increasingly distinct market segments

Until 2H2012, most large players, including IBM, were still testing and evaluating the public IaaS market, largely leaving this space to innovative early entrants, first and foremost Amazon Web Services (AWS). Since then, all other major players are, to some extent, playing catch-up with AWS. To this effect, they have either introduced or expanded their IaaS offerings, resulting in exacerbated competition as well as increased complexity for users attempting to compare solutions.

Nevertheless, offerings remain significantly different between vendors in terms of technology choices and “packaging”, pricing and service level models, as well as the combination of IaaS, PaaS (Platform-as-a-Service), and SaaS.

Indeed, while IaaS and SaaS appear as key strategic segments due to their size, growth rates, and potential, the smaller PaaS market will also be a crucial battlefield. PaaS will play an essential role to enable application development and operations, as well as proper integration of cloud infrastructures with the existing information systems. It will be instrumental in making tomorrow’s hybrid architectures possible and effective.

As the market expands, further segmentation has taken place. IaaS offerings now range from “pure commodity” clouds, which are characterized by “shared responsibility” (the customer has sole responsibility for the use of shared infrastructure above the most basic layers) to “enterprise-class” clouds. In the latter case, particular attention is paid to business-relevant features (e.g. the ability to accommodate existing virtualized applications “as they are”) and services (e.g. migration services or operations and management services). The infrastructure supplier may then assume a definable, larger part of responsibility, possibly up to applications operations and management.

Cloud IaaS can therefore no longer be simplistically reduced to self-service offerings. As cloud solutions become integral part of corporate information systems, customers expect to choose from a full range of sourcing options, from bare self-service to managed infrastructure with extended services and support.

While cloud solutions enable substantial gains in cost-effectiveness, choosing such solutions does not eliminate demand for services – quite the opposite, as many customers continue to also seek efficiencies via outsourcing and at the same time contemplate digital transformation projects.

Indeed, our research indicates that service-intensive IaaS offerings, although usually positioned at substantially higher price points than commodity IaaS, can offer a better “Total Cost of Solution” when all related costs are accounted for. In particular, operations and management can be optimized by IaaS vendors through automation and industrialization, extending the cloud benefits into IT operations. This applies not only to IBM’s “managed IaaS” offering but also to comparable, service-intensive IaaS offerings from providers such as CSC or Orange.

In a typical public IaaS project, spending on “raw infrastructure” usually never exceeds 30% of total project costs. Personnel costs and/or services account for the most substantial part of the remaining budget, together with telecom expenses and additional infrastructure spending (e.g. software and/or complementary appliances). This “70% +” part of project spending can easily explode if improper choices are made (e.g. wrong architecture choices, poor infrastructure and application management, etc.).

Enterprise-class public cloud offerings can become truly cost-effective solutions when tightly coupled with value-added, optimized services, a characteristic that makes such a market segment extremely attractive to vendors like IBM.

IBM 2013: Building long-term cloud presence – all segments, and a focus on services

IBM today has a very broad cloud offering, encompassing almost every segment of IaaS, PaaS, and SaaS as well as hardware (e.g. Pure Systems) and software cloud solutions, even though part of the effort has included “cloud-painting” of existing products. The portfolio even includes duplicate, potentially competing solutions, such as the SmartCloud Enterprise and SoftLayer IaaS offerings (see below).

Overall, IBM executed on an aggressive plan to build cloud presence and market share. This seems well-suited to IBM’s competitive positioning. As the leading global IT services provider, it has a relatively unique capability to support a broad portfolio of technologies, solutions, and services.

IBM’s offering covers the three cloud layers, but is not monolithic. The company claims more than 100 different SaaS offerings, but also knows it has to operate in a broader ecosystem encompassing several large “coopetitors” such as SAP, Oracle or Microsoft as well as myriads of cloud-native solutions. It has a substantial PaaS offering but also needs to co-exist with third party solutions. As for IaaS, IBM’s focus is on enterprise-class solutions with a variety of service levels and (since the acquisition of SoftLayer) a broad range of contractual formats, from credit-card subscription to more “IBM-like” formal agreements.

Beyond public IaaS now being an attractive market in its own right, IBM needs this extensive presence in cloud-enabling technologies and services to pursue its previously mentioned broader vision of being a driving force in the “digital transformation” of enterprises, as well as governments and ultimately the entire society. This represents a huge, long-term potential for high value solutions and services, which are the true bread and butter of IBM’s business model. Mastering the infrastructure remains a critical pre-requisite for this strategy – at IBM’s scale.

In line with this, investments have been substantial. The SoftLayer acquisition (discussed below) is estimated around US$ 2 billion, while each of the 10 current SmartCLoud Enterprise datacenters has an estimated unit cost of about US$ 350 million. The financial stakes are significant since IBM is aiming for an aggregate cloud revenue target of US$ 7 billion by the end of 2015, even though public IaaS will only represent a fraction of this revenue.

Logically, the company is also deploying substantial efforts towards relevant indirect channels. Its large solutions portfolio enables IBM to address the needs of a variety of customers, but also to cooperate with a broad array of partners, including “channel” partners, i.e. distributors, resellers or service companies who can address mid-market and SMB customers.

With IBM having focused its direct sales force on a relatively narrow band of very large, usually global accounts, this indirect channel is crucial to its market presence. Many of these partners are already involved with IBM and have expertise on at least some part of its conventional offerings, hence sharing common interests with the company. However, the proper “indirect channel cloud business model” remains largely to be invented, while new channels are emerging (e.g., cloud services brokers).

IBM’s IaaS offering: first addressing existing customers

IBM’s two main traditional strategy “invariants” are a strong investment in true R&D, leading to technology control, and a very strong focus on customer base intimacy. These two aspects clearly appear in the company’s IaaS offerings.

IBM is unique in the market place for supporting non x86 based IaaS solutions, namely solutions using its own Power AIX/Linux and z/Series mainframe platforms. The latter is only proposed to existing mainframe customers; however, IBM is also likely to offer a significant portfolio of mainframe-based, SaaS-like cloud services. The first significant appearance of this type of offering is a set of mainframe-based IT security services, and IBM is particularly likely to expand its cloud offering of mainframe-based big data processing services. The purchase of CSL, an Israel-based vendor offering virtualization and cloud technologies for the zSeries environments, further demonstrates IBM’s determination to ensure that its mainframe technology has a role to play (beyond running “legacy” applications) in future cloud-based, hybrid architectures. However, proprietary technologies support in IBM cloud offerings will mainly result in protecting as much as possible IBM’s installed base revenues and account control.

In the x86 space, after the “experimental” phase involving a KVM-based, development-oriented offering, IBM’s approach to public IaaS has been clearly centered on enterprise needs, with a portfolio of solutions based on VMware and on its own virtualisation platforms (PowerVM and zVM), and two main classes of service. While the “SmartCloud Enterprise” offering is primarily still targeting development and web applications operations, “SmartCLoud Enterprise +” is designed for business-critical applications and provides a large spectrum of management services. In both cases, customers still need to enter a formal contractual agreement with IBM prior to any deployment – thus excluding the ease-of-use of credit-card infrastructure subscription (à la Amazon), but also reassuring customers on the “select” nature of the SmartCloud customer base.

The introduction of these offerings has indeed substantially contributed to credibilize enterprise-class, managed IaaS as a major option for customers looking at the cloud for mission critical workloads.

These offerings enable IBM to cope with almost any cloud and IaaS use case. With their services components, they are particularly suited to IBM’s customer base, which will be able to integrate cloud initiatives and existing systems… with (substantial) help from their favorite supplier.

However, the downside of the current offering is that users may find it complex and even confusing, hence difficult to understand and choose from. Integrating IBM’s various cloud components may also appear as (and actually be) complex and costly, likely requiring substantial involvement from IBM’s own integration services units.

Price competitiveness will be a major requirement across the board, together with service cost clarity (e.g., the service-intensive SmartCLoud Enterprise + does not have a clear, public price list) and predictability (due to the complexity of possible combinations). a major concern in the two “upper” layers.

These constraints represent a particularly significant issue when attempting to attract out-of-base customers, or even expand towards new customer layers inside existing customer organizations, especially non-IT entities seeking rapid answers to their immediate business needs.

The SoftLayer acquisition: a growth accelerator

In 1H2013, IBM acquired SoftLayer, a well-known privately owned IaaS provider with innovative technology and a substantial customer base. This move brings IBM an IaaS offering that is better suited to “cloud natives” (e.g., via the possibility to buy resources on the fly with a credit card), as well as strong infrastructure management automation technology, which is a critical element to economically manage large, even heterogeneous cloud infrastructures.

Despite the capabilities of its own Tivoli division and its own SmartCloud management solutions, IBM recognized a need to acquire such technologies – time-to-market considerations are likely to have played a key role here. IBM had already implemented a similar move in the PaaS-related software continuous release and deployment space, where the acquisition of Urban Code has brought complementary solutions and culture to its internal resources (mainly organized in the Rational business unit), substantially expanding IBM’s “DevOps” offering.

With SoftLayer, the company definitely “imports” a different cloud culture centered on “cloud native” customers as well as SMBs. SoftLayer also brings significant sensitivity to, and understanding of, cloud developers expectations.

The SoftLayer acquisition does nicely complement IBM’s existing capabilities; however, it further complexifies the perception of the offering – at least in a transition phase. In addition, while IBM has been a long-time proponent of OpenStack, an open source cloud framework, SoftLayer was a supporter and user of the competing CloudStack initiative. Even though IBM has the capabilities to either port SoftLayer’s key technologies to OpenStack or to maintain two technology stacks, the company does not yet present a clear perspective or roadmap for customers seeking to define their cloud and hybrid architecture specifications or orientations.

Its acquisitions demonstrate that IBM wants to rapidly impose itself as a major cloud player, with significant, if not dominant, presence in the public cloud segments. This “preemptive strategy” makes sense, since in this fiercely competitive business, market share acquisition is likely to become increasingly costly. It reflects a realistic analysis of the urgent need to add muscle to IBM’s cloud capabilities by importing missing expertise, experience and culture from the outside.

From a sales channel point of view, the SoftLayer acquisition brings IBM a significant SMB cloud customer base, quite different from its traditional markets. Indeed, the company will end-up with three major sales and delivery models: direct with its very large, global customers, via a partner ecosystem for its “general business” clients, and an electronic-commerce (almost BtoC-like, close to Amazon’s) model for SMBs. In the latter, SoftLayer has a key role to play, but IBM’s ability (and, maybe, determination) to succeed in this type of market/channel remains to be seen. Competition will be different, albeit always fierce, in these three spaces.

Challenges ahead

All in all, IBM has a very rich IaaS offering, backed by significant resources, but complex and multifaceted. With its broad set of capabilities, products, and services, IBM does not need overwhelming dominance in a given cloud segment (à-la-Amazon in commodity IaaS) to establish itself as a significant cloud player. Instead, its broad portfolio of products and “base services” will provide a solid basis for value-added services, in conventional domains such as IaaS management and operations, innovative fields such as big data, and more broadly in enterprise digital transformation.

Nevertheless, IBM remains confronted with several significant challenges, among which:

  • Organization and culture

IBM’s biggest challenge will not be in the possible conversion of SoftLayer’s tools to an OpenStack environment, or their application to the SmartCLoud infrastructures. Indeed, it will not have much to do with any technological evolution.

The true touchstone will be in IBM’s ability, or not, to capitalize on its acquisition’s “pure cloud culture” while integrating the SoftLayer and internal IBM IaaS teams together. This “merge of cultures” will also facilitate the development of a successful partner ecosystem, including cloud-native partners. That synergistic integration will be a key challenge for the newly-formed IBM “cloud services division” (operational since the completion of the Softlayer acquisition), which has now sufficient muscle to aggressively develop its strategy alongside IBM’s traditional businesses.

  • Legibility of the offering

IBM’s offering is particularly rich but also potentially confusing. Some rationalization will be needed to achieve simplification and formulate a clear roadmap to make it more legible for customers and partners, both in IaaS and (even more) in PaaS.
In the latter space, the recent announcement that IBM will collaborate with Pivotal (a VMware spinoff) on CloudFoundry, an open-source PaaS solution, confirms IBM’s long-lasting commitment to open source, but does not entirely clarify its immediate PaaS roadmap.

On a more general note, with IBM strongly promoting open-source cloud technologies, SmartCLoud Enterprise users (and the IBM sales force) may wonder whether IBM’s current commitment to VMware’s platforms will continue in its current format, especially at the vCLoud director layer. This is a classic post-M&A question: what should users buy and what should the sales force sell?
IBM has no choice, like most of its competitors in the enterprise cloud space, but to support the virtualisation engine, ESX, due to its market share. However, from an architecture standpoint, substituting a different underlying cloud stack would not be neutral for all customers.

The combination of SmartCloud/SoftLayer IaaS offerings into a highly industrialized offering (including IaaS operations and management services) will determine its cost-competitiveness on a “Total Cost of Solution” (TCS) basis; to this effect, as already mentioned, IBM will also need to make its service costs more transparent and predictable.

Overall, offering legibility must improve for IBM to achieve leadership in public enterprise IaaS. Customers will need a clearer perspective and roadmap on IBM’s offering.

Indeed, with cloud becoming essential to information system “hybrid” architectures, customers need to be able to plan ahead their technology orientations (notably on IaaS and PaaS) together with their sourcing/procurement strategies. They also need solutions they can rapidly adopt to timely address their business needs, without necessarily engaging into ambitious digital transformation projects, but with a reasonable degree of certainty and visibility on the future of the products and services they select. IBM’s current offerings demonstrate the quality of the company’s vision, but do not yet offer customers an equally clear path moving forward.

  • Infrastructure management

As customers purchasing service-intensive IaaS cloud offerings will increasingly require business-oriented, highly tailored, end-to-end Service Level Agreements (SLA), IBM is also likely to strengthen its partnership ecosystem with complementary players, even though they may also be competitors.

Most notably, after having designed a broad cooperation deal with AT&T, the company is likely to seek additional (but not necessarily similar, and on a more tactical, or ad-hoc, basis) agreements with other carriers, who control the key missing piece in IBM’s cloud portfolio: the wide-area network.

Indeed, the importance of the network in cloud performance and even reliability will continue to grow in importance as business continuity and disaster recovery become both a growing IaaS application (even for non cloud based infrastructures) and a key feature of almost any cloud-based IaaS project. Similarly, from a more technical standpoint, latency re-emerges as a key concern for many applications, and the underlying network performance therefore gets paramount importance, irrespective of additional efforts on higher layers such as WAN acceleration.

Conclusion

Cloud markets will not be for the faint-hearted in the coming years, with competition intensifying between the major ICT players. IBM’s moves and large scale investments are consistent with its ambitious business objectives. As always, execution will remain key, and “soft”, cultural aspects will be essential.

IBM’s rich portfolio of cloud technologies, solutions and services enables the company to potentially address any customer requirement.

On the downside, IBM now has two public cloud offerings built on different technologies, and has yet to explain how it intends to reconcile its apparently conflicting orientations. In addition, given the focus on large scale business transformation, IBM’s responsiveness to the needs of smaller customers and to more “IT-centric” projects remains to be demonstrated. Finally, the company has only begun to address the fundamental issue of end to end infrastructure management, including both disaster recovery and the wide area network.

In this strategic public cloud segment, IBM is not likely to be a low-cost, commodity-oriented provider despite being very aggressive in some areas of fierce competition such as public cloud storage. Its offerings will be worth considering from a “Total Cost of Solution” standpoint and in terms of broader business benefits, especially those related to innovation and digital transformation. However, the future evolution of the currently somewhat nebulous spectrum of offerings requires clarification and a clear roadmap.

Wednesday, August 28th 2013
Donald Callahan
Newsletter To subscribe to the Duquesne Advisory Newsletter, please enter your e-mail address.

Duquesne Advisory

Duquesne Advisory is a European firm, dedicated to researching, understanding and advising clients worldwide on opportunities and trends in Information and Communications technology.

Research

Duquesne Advisory delivers in-depth analyses of Information and Communications Technologies, their implementations and their markets. Research is based on critical observation of the market by the analysts and their on-going contacts with the vendor community, together with hands-on, practical experience in consulting engagements.

Consulting

The analysts of Duquesne Advisory leverage the Firm’s ongoing market and technology research to undertake high added value consulting engagements for both ICT users and ICT providers. Focused on client service, their approach is rigorous and methodical, and at the same time pragmatic and operational.