Cloud orchestration

Download for Kindle

Cloud orchestration

Cloud orchestration

Tech Trends 2014

CIOs should be making deliberate investments in developing advanced integration and data management capabilities to support cloud-to-cloud and cloud-to-core models.

Cloud adoption across the enterprise is a growing reality, but much of the usage is in addition to on-premises systems—not in replacement. As cloud services continue to expand, companies are increasingly connecting cloud-to-cloud and cloud-to-core systems—in strings, clusters, storms, and more—cobbling together discrete services for an end-to-end business process. Tactical adoption of cloud is giving way to the need for a coordinated, orchestrated strategy—and for a new class of cloud offerings built around business outcomes.

From cloud to clouds (to core)

Cloud adoption across the enterprise is a growing reality. Forrester predicted that “by the end of 2013, enterprises will use an average of 9.6 software-as-a-service (SaaS) applications.”1 Yet much of the cloud usage is not in lieu of on-premises enterprise systems. Forrester also found that “only 18 percent of the enterprises that were first-wave adopters and less than 9 percent of the second-wave adopters have used SaaS as a full replacement.”2 As a result, these cloud services increasingly require integration back to core internal systems—linking edge offerings to legacy financials, order management, inventory, HR, manufacturing, and other enterprise systems. Companies are connecting clouds—in strings, clusters, storms, and more—and cobbling together discrete services to create end-to-end business processes. Tactical adoption of cloud is giving way to the need for a coordinated, orchestrated strategy.

As cloud services continue to expand in number and sophistication, gaps in managing cloud-to-cloud and cloud-to-core portfolios are beginning to appear, leading to new and smarter ways to operate in this hyper-hybrid3 IT environment. It is also opening the door for a new category of offerings: pre-integrated and orchestrated cloud offerings delivering higher-order business outcomes-as-a-service.

Learn more on Deloitte.com

Learn more about cloud orchestration.

All together now

Integration, data management, and enterprise architecture have long been aspirations for IT. With cloud, these practices have become more complex. And they’ve shifted from leading practices to critical core disciplines. Integration stability and reliability was the number two challenge in a recent survey on cloud adoption, trailing only security concerns.4 Virtually every enterprise should be developing a strategy on how to integrate, aggregate, and orchestrate its collection of cloud and on-premises assets. Understanding the extensibility, portability, and reliability of a cloud service should begin at the sourcing stage.

  • Extensibility refers to the ability to get information into and out of the service—the availability of data and transactions to be invoked by other parties and the ability to trigger external events from within the cloud service. Many cloud providers offer lightweight web services and RESTful5 interfaces, but it’s important to review the assets around APIs and data structures—documentation, toolkits, testing harnesses, backward compatibility, and deprecation policies.
  • Portability represents the ease of migrating your business from the cloud service. Can data be exported? What about customized business logic? Are there contractual terms associated with intellectual property ownership?
  • Reliability addresses performance of the service—not just the core cloud offering, but the surrounding stack. For an orchestrated process, the integration layer and dependencies on legacy systems should be able to scale dynamically to take advantage of the elasticity of cloud services. The end-to-end business process is only as strong as its weakest link.

Cloud orchestration can build from a mature enterprise integration and architecture footprint. The underlying tenets are familiar: service orientation, data correlation, security services (especially authentication, entitlement management, and encryption), and a separation of business logic. Several integration platforms have emerged from the cloud, offering cloud-based deployment options as well as preconfigured connectors and integration patterns for popular cloud services. Providers include Boomi, CastIron, MuleSoft, and TIBCO’s Cloud Bus.

New beginnings

The cloud provider market is starting to address the desire for higher-level, pre-integrated cloud orchestration services. For example, consider the example of a health plan’s recruiting and HR service. Today, health plans contract with separate cloud providers for résumé sourcing, background checks, on-boarding, benefits, payroll, and performance management—which means they need to develop and maintain point-to-point interfaces between the various players to enable the full prospect-to-employee lifecycle. They are waiting for an end-to-end “hire to retire” service to emerge, which could provide contracting, configuration, and handoffs across various systems. The enterprise could subscribe to a single service, priced based on usage or, in an ideal world, on outcomes.

Traditional ERP players are acquiring and integrating cloud applications to supplement core offerings. Established cloud providers are creating storefronts of complementary cloud solutions, which make choosing and buying an expanding inventory of services easier. But we are still in the early days of this expansion, and integration often remains the buyer’s problem. Over time, technical compatibility within a vendor’s stack should become less challenging. ERP and cloud providers are also planning improved interoperability between their products6—an encouraging development, to be sure, but of little help in the immediate term.

Others may yet enter the cloud orchestration market. Systems integrators and professional services firms that specialize in integrating diverse systems could expand and formalize their roles by pre-integrating the components of an end-to-end bundle. For such organizations, this may offer a way to monetize intellectual property around industry and process experience while diversifying from consulting to a product revenue stream. Several high-tech players looking to expand their offerings could emerge, such as Amazon, Google, HP, and Microsoft.

Cloud Orchestration

A brave new world

The initial market for effective cloud orchestration is likely to be startups and small- to medium-sized businesses. They could receive the benefits of one-stop access without the hassle of navigating vendor contracts, integrating systems, and managing data. Larger businesses in emerging markets are also natural targets. Like startups, their circumstances may not justify a full enterprise solution. Finally, serial acquirers could gain agility and advantage from being able to integrate diverse platforms more efficiently. In each case, IT’s mission should be to create integration, data management, and security services to guide cloud adoption.

But the majority of the Fortune 1000 will be living with the reality of a mix of cloud and core offerings, even as sophisticated cloud orchestration emerges. IT’s charter to own cloud integration, data, and security is even more important in this case—especially as businesses are increasingly dependent on hybrid operating environments. Build the components to orchestrate the cloud today, and you’ll be ready to adopt more compelling services as the market develops.

Lessons from the front lines

Linking the network7

LinkedIn, a social networking website, has three main enterprise lines of business: talent solutions, marketing solutions, and sales solutions. As Andres Bang, LinkedIn’s head of global sales and operations systems, described recently: The company adopted cloud services to support sales and CRM functions, but found that its business was outgrowing standard out-of-the-box capabilities and that its processes increasingly required integration to ERP and proprietary systems for generating sales leads.

To address its immediate lead-to-cash process requirements, and to build a scalable solution for future orchestration, LinkedIn adopted a cloud-based integration platform. Bang explains that by using the integration platform, LinkedIn was able to connect multiple systems, including its lead generation tool, CRM system, financial system, data warehouse, and proprietary applications. Integrating both its cloud-based and on-premises systems created a “single pane of glass” for the company’s salespeople to access the information they need to perform their jobs.

Orchestrated banking8

SunTrust Banks, a leading US financial services holding company, found that its relationship managers were encountering issues with accessing customer information in a timely manner, threatening their ability to provide quality customer service. The root of the issue was the company’s reliance on an assortment of back-end systems for loan origination, underwriting, servicing, and CRM. SunTrust’s architecture was a mix of cloud services, on-premises packaged software, and on-premises custom solutions. The company sought an integrated, scalable solution to expedite the delivery of services to customers—and pave the way for future cloud adoption.

The bank decided to adopt a cloud-based integration platform to address these challenges. By connecting SunTrust’s back-end enterprise application and shared services to the cloud, SunTrust was able to eliminate its complex back-end business processes. Furthermore, the cloud enabled seamless integration with the bank’s enterprise service bus and provided preconfigured connectors to cloud services.

Today, SunTrust maintains a scalable solution supporting its broader business process transformation. Furthermore, relationship managers are empowered with the tools and resources to access important customer information in a timely manner, reducing the time it takes to provide service to customers.

Hybrid high tech

A global hardware and software company was undergoing rapid change stemming from acquisitions, organic growth, and divestitures. The company’s goal? To maintain its core hardware and product businesses while expanding its software and services offerings. The company’s expansion introduced complexity in many areas, such as marketing, sales and incentive management, product configuration, pricing, and project and workflow management. Speed to market was a driving force, since the organization wanted to engage with customers from dozens of countries in a consistent and coordinated manner. The company also recognized that its strategy was built around continual transformation of its offerings—and that required flexibility and agility in the enabling systems.

The organization was vexed by decades of what it called “lumpy” expenditures—costly IT infrastructure refresh cycles, with a history of overspending for capacity because of unpredictable demand. But the concern was about more than cost and scale. The company also sought shorter time to market and the ability to more efficiently assimilate new ventures. This was important, given its recent wave of acquisitions.

The company’s vision is to move to a 100 percent cloud-based infrastructure for the enterprise. As a first step in fulfilling this vision, and to continue to provide seamless, end-to-end business processes, the organization orchestrated a complex integration between multiple cloud services and its on-premises systems. A new sale requires smooth interaction between separate cloud systems for many processes: calendaring and messaging; materials development; lead and campaign management; opportunity, sales, and support management; configuration, pricing, and quoting services; sales and support management; and compensation and incentives. The integration enabled these systems to communicate with each other, and it also included hooks into on-premises systems for human resources and order and billing management. Recognizing that the glue to bring together the various services was as important as the individual functionality being delivered, the company created disciplines around cloud-to-cloud and cloud-to-core integration: tools, architectural standards, and a dedicated team to drive growth and adoption.

Through the company’s efforts, maintenance costs have gone down: Instead of heavily funding incremental software improvements, the company is taking advantage of enhancements being rolled out by the cloud services. System performance has improved; outages have become shorter and less frequent. The company’s global teams have enjoyed greater browser and device compatibility, as the cloud offerings have a wider footprint than was historically allowed. And the business feels better served by IT: IT’s responsiveness has improved, as has the business’s understanding of associated costs. Finally, the company has started to take the next step toward the overall vision by shifting to cloud hosting of traditional ERP to “rightsize” the underlying infrastructure—a solution that can scale up (or down) based on the company’s circumstances.

Espresso with a shot of cloud9

Online distribution channels have transformed Nestlé Nespresso S. A. from a traditional, coffee-shop-and-boutique-store business model to a household brand in the single-serving coffee machine category. But in order to meet growing global customer demand, Nespresso needed to replace its home-grown, complex ERP system with a more scalable architecture and integrated cloud solution.

The business began enhancing its enterprise architecture by launching the Nespresso Open Architecture (NesOA) platform, a tool designed from service-oriented architecture (SOA) principles. With NesOA, Nespresso’s IT department could support new distribution channels, manage increased consumer traffic, and introduce new applications and services to the business. Furthermore, by using a cloud-based integration platform, Nespresso could easily integrate a variety of systems, including the Nessoft ERP system, an interactive voice response system, an automated warehouse management system, and an emergency ordering tool.

As a result, Nespresso’s NesOA transformed its home-grown enterprise into a scalable, automated, and more efficient solution to meet business needs. Furthermore, it mitigated the risk of disruption from a single point of failure with a solution based on clustering and redundancies. Nespresso is now poised to leverage cloud and traditional solution offerings to support future growth of its IT system landscape.

My take

Dounia Lievan, former banking executive
Director, Deloitte Consulting LLP

I formerly worked for a regional bank that generated the majority of its revenue from mortgage banking. The bank looked to diversify by focusing on both the retail and commercial banking lines of business. We recognized that we could drive immediate revenue in retail by elevating the customer experience at the point of sale. The longer-term goal was an integrated omnichannel experience driven by online and mobile capabilities, but initially the case for change was to better serve our customers in the branch—knowing that the technology we implemented could provide a solid foundation for our “connected customer” vision.

DUP571_Cloud orchestration_960x550Previously, the process for opening an account with related services was lengthy and inefficient. Employees accessed multiple systems and entered duplicative data. In addition, we didn’t provide bankers with tools or insight to identify customer needs, and we lacked an automated way to manage the ongoing customer relationship. To improve the experience for both the customer and banker, and with an eye toward the broader vision, we invested in a customizable cloud solution with native customer relationship management (CRM) capabilities. It provided one delivery system for branch bankers with increased flexibility, support for process improvement, and the option for future expansion across channels.

We chose to go to the cloud for several reasons: to generate revenue, to increase efficiency, to be agile enough to respond to changes in the marketplace, and to differentiate ourselves from our competitors—while reducing the burden on the IT organization. Cost factored into, but didn’t dominate, the investment decision.

Our legacy systems didn’t provide the functionality needed to solve the business problem, so we used the cloud and SaaS to connect multiple core and ancillary systems. Integration isn’t a new concept, but with the cloud, a balance has to be struck between traditional methodologies and the flexibility that the cloud can introduce. Looking back, there are a few things we might have done differently, such as creating more real-time APIs versus batched transfers. But, overall, the project laid the groundwork for the longer journey.

During the course of the project, and especially after go live, the cloud changed how IT and the business work together. It drove collaboration and created a new team with an enhanced skill set and a different mindset. They’re no longer completing a stand-alone project and moving on, but dedicated to driving continuous improvement and evolving the platform to deliver business results.

As you take the cloud integration journey, executive sponsorship and building enterprise support are key success factors. Create a strategic roadmap, and articulate your plans two to three years out. Upon completing the first phase, showcase the solution and use the roadmap to sell the vision to the C-Suite and across business lines. Provide regular updates on adoption metrics, user feedback, progress toward change, and—more importantly—return on investment. This keeps the platform top of mind and makes it easier to gain support to grow the platform and enhance its value.

As business leaders, we look for ways to drive revenue and efficiency, continuously creating value. At the heart of the banking business is the relationship we have with our customers, and building trust is the foundation of that relationship. What I love about this technology is that we used it to remove compliance and operational obstacles and gave bankers the tools they need to be effective and efficient. We used technology to enable bankers to be present, listen, ask questions, and help people—to make a stronger human connection.

Where do you start?

Even with the more sophisticated cloud offerings that span end-to-end processes, the challenge of integrating cloud-to-core remains. How does the CIO manage the definition of standards for cloud adoption? Establish architecture to support integration? Handle data correlation, retention, and migration? These are important questions to answer now—and they’ll be even more important as cloud services spread across the enterprise. CIOs should be making deliberate investments in developing advanced integration and data management capabilities to support a cloud-to-cloud-to-core model.

  • Petition for a new cloud business model. Many companies could save money if cloud pricing was based on usage and outcomes rather than licensing fees. If this is true for your organization, let the cloud providers know. And if your company is ready for an orchestrated cloud option now, connect with others who share your need. Let your voices be heard by the software vendor community.
  • Build an integration foundation. Even if your organization doesn’t operate in a cloud-to-core environment, it’s likely you eventually will. Laying the groundwork now will make integration easier later. If you’ve already invested in middleware to link legacy systems, build from there. However, you may find that a cloud-based model requires new approaches.
  • Connect the dots. Definitions of customer, product, employee, and other data elements vary from one cloud solution to another—and need to be mapped to your business’s semantics and taxonomy. Understand how each application defines its dataset, and develop a strategy for funneling data from various cloud systems to support your organization’s reporting and analytic objectives.
  • Read the fine print. Develop a healthy skepticism of cloud provider contracts. Understand your rights to data ownership, portability, and migration. If you change providers, can you be confident that your data is protected? Negotiate terms where possible to maintain your flexibility.
  • Build a strong chain. Overall business performance is limited by the weakest cloud solution in the process chain. Understand the performance variability your business will tolerate, and weigh whether each individual cloud service can meet those demands. And remember: The scalability and performance of the interconnected whole is only as strong as its weakest link. Cloud’s elasticity could stress (and break) legacy solutions built around more modest, predictable requirements. Cloud-based integration platforms ramp up (or down) to meet your needs—similar to the cloud offerings you are looking to orchestrate.
  • Explore edge architecture. Borrowing from the days of SOA, consider describing business capabilities and processes as services. The goal is to connect enterprise core, private, and public cloud offerings—which can be broken into a common set of services used to deliver on business needs. This will lead to deliberate identification and management of business rules, APIs, identities and personas, entitlements, workflow items, and interfaces. The goal is to promote reuse, standards adoption, and architectural integrity—from a business-driven mindset. A revamped IT delivery model will likely be needed, as will support from both IT and business executives for a new governance mindset.

Bottom line

As enterprises use disparate cloud offerings to handle critical business processes, the desire to link these offerings to core legacy systems and data grows. IT organizations will be asked to provide that orchestration. A recent Gartner survey shows that “over 70 percent of organizations that are using or planning to use cloud services expect internal IT organizations to assume the role of cloud services broker.”10 That need has generated challenges that extend beyond integration to include security, data integrity and reliability, and business rules for managing a hybrid state. It is also creating demand for cloud orchestration to link multiple cloud services to each other—and to the core. CIOs who have the disciplines of data management and integration architecture in place will be positioned to create harmony out of the existing landscape and to leverage orchestration services when they arrive.

 

Endnotes

View all endnotes
  1. Stefan Ried, The Hybrid² integration challenge, Forrester Research, Inc., May 1, 2013, http://www.forrester.com/The+Hybrid178+Integration+Challenge/fulltext/-/E-RES94881, accessed December 31, 2013.
  2. Ibid.
  3. Deloitte Consulting LLP, Tech Trends 2012: Elevate IT for digital business, 2012, chapter 5.
  4. IDG Enterprise, Cloud computing: Key trends and future effects, June 27, 2013, http://www.idgenterprise.com/report/cloud-research-2013, accessed December 31, 2013.
  5. Representational State Transfer (REST) is an architectural style for designing networked systems using web protocols and technologies such as HTTP (Hypertext Transfer Protocol) and XML. See Roy Fielding, Architectural styles and the design of network-based software architectures, 2000, http://www.ics.uci.edu/~fielding/pubs/dissertation/top.htm, chapter 5, accessed January 8, 2014.
  6. Larry Dignan, “As the cloud turns, Salesforce, Workday integrate more,” ZDNet, September 18, 2013, http://www.zdnet.com/as-the-cloud-turns-salesforce-workday-integrate-more-7000020841/, accessed December 19, 2013; Oracle, “Oracle, NetSuite, and Deloitte announce plans to deliver integrated HCM and ERP cloud services for mid-size customers,” press release, http://www.oracle.com/us/corporate/press/1966087, accessed December 31, 2013.
  7. Dell Boomi, Inc., “LinkedIn’s use of Dell Boomi (video),” http://www.boomi.com/customers/success/linkedins-use-dell-boomi, accessed January 7, 2014.
  8. IBM, “SunTrust Banks, Inc. improves access to accurate information,” June 11, 2013, http://www-01.ibm.com/software/success/cssdb.nsf/CS/RNAE-98JKWH?OpenDocument&Site=default&cty=en_us, accessed January 7, 2014.
  9. MuleSoft, “Nespresso brews up growth, profits with new SOA-based strategy,” http://www.mulesoft.com/case-study/nespresso#download, accessed January 7, 2014.
  10. Michele Cantara, Hype cycle for cloud services brokerage 2013, Gartner, Inc., July 31, 2013 (revised October 28, 2013).

About The Authors

Andy Main

Andy Main is Deloitte’s National Customer Solutions service line leader and the Global Customer Solutions leader. In his national leadership role, Main is charting new courses in cloud computing, providing clients with more flexibility than before.

John Peto

As the US lead for the salesforce.com global alliance and a leader in Deloitte Consulting LLP’s Customer Solutions practice, John Peto focuses on helping companies transform the effectiveness of their sales, service, and marketing functions through the implementation of new processes and supporting application, integration, and data architectures.

Tech Trends 2014, Cloud orchestration
Cover Image by Celia Johnson