Your SlideShare is downloading. ×
SaaS and Integration Best Practices
Upcoming SlideShare
Loading in...5

Thanks for flagging this SlideShare!

Oops! An error has occurred.

Saving this for later? Get the SlideShare app to save on your phone or tablet. Read anywhere, anytime – even offline.
Text the download link to your phone
Standard text messaging rates apply

SaaS and Integration Best Practices


Published on

  • Be the first to comment

  • Be the first to like this

No Downloads
Total Views
On Slideshare
From Embeds
Number of Embeds
Embeds 0
No embeds

Report content
Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

No notes for slide


  • 1. SaaS and Integration Best Practices  Henry Hai  Seitaro Sakoda (Manuscript Received January 23, 2009) The rising adoption of software-as-a-service (SaaS) applications by enterprise organizations has been driven by deep dissatisfaction with on-premise applications, which require organizations to purchase and deploy infrastructure, overstock on licenses, and pay for expensive resources for customizations, upgrades, and on-going maintenance. The large upfront investments combined with unpredictable costs and an immeasurable return on investment have prompted organizations to seek cheaper less-risky alternatives. Many have found that SaaS applications, which require minimal or no infrastructure and maintenance, can be deployed quickly and have a predictable cost model representing less risk and a faster return on investment. The new demand has led to rapid innovation in SaaS applications, SaaS platforms, third-party SaaS add-ons, and SaaS integration tools. However, enterprise organizations still have the burden of integrating these applications with their back-office systems and on-premise applications, without which the SaaS applications have little to no value. Complex enterprise integration requirements challenge even the best SaaS solution providers today; there are still limitations and pitfalls to be wary of. In this paper, we describe some SaaS integration best practices, present a case study, and highlight emerging integration technologies that can help ease the burden of integrating SaaS applications. 1. Introduction to offer few economic benefits over their locally Today’s businesses need to be agile. installed counterparts and had limited ability for Constant change comes from every direction: customization, data sharing, and integration with mergers and acquisitions, re-organizations, new other applications. Customers were still required regulations, economic pressures, competitive to pay for perpetual licensing, application pressures, and fluctuations in customer demand. upgrades, security, performance, availability, Unfortunately, most applications are not designed and unused capacity. Any customizations to the for flexibility, and the unpredictable costs and applications were costly and reserved only for resources needed to maintain the infrastructure larger customers. Customizations also required and extend these applications are constraining the ASPs to manage the development, testing, the business instead of enabling it. and deployment cycles differently for every These problems are not new. The application customer. service providers (ASPs) of the 1990s, which Today, with advances in infrastructure and provided shrink-wrapped applications to business software technology, a new breed of application users over the Internet, made early attempts is being produced: robust Web-based applications to offer robust packaged applications more designed as centralized, shared-instance, multi- quickly and easily. Unfortunately, they tended tenant applications called software as a service FUJITSU Sci. Tech. J., Vol. 45, No. 3, pp. 257–264 (July 2009) 257
  • 2. H. Hai et al.: SaaS and Integration Best Practices (SaaS). Like ASPs, up-and-coming SaaS solution with the IT department having little or no providers take responsibility for managing control over them and they need consolidation the service including security, performance, and integration. As SaaS solution providers gain availability, reliability, and scalability. However, strength and begin to offer better customization the similarities stop there. In mature SaaS and integration capabilities, the business is offerings, multi-tenancy applies to all tiers of beginning to call for IT department involvement. the software architecture: all tenants share the On the flip side, IT managers who are facing same codebase and instances of the application, recent failures of large-scale on-premise packaged which enables large economies of scale. Given application implementations and tight budgets that the underlying application codebase cannot are beginning to recognize the low-risk and quick be changed for each tenant’s customizations, the time-to-market advantages of the SaaS delivery application is abstracted into loosely coupled model. This coalescence is producing some of the fine-grained configurable components. This lets largest enterprise SaaS implementations to date. customers quickly change the presentation, logic, SaaS is growing by leaps and bounds. and database layers of the application through Large independent software vendors that have on-screen clicks without code modification, traditionally offered on-premise applications compilation, or deployment. Moreover, the are creating new SaaS offerings in response to loose coupling allows SaaS providers to market demand. Some of the larger traditional release frequent application upgrades, giving on-premise players offering SaaS applications all customers seamless access to the latest are SAP in enterprise resource planning (ERP); version of the service bundled with their unique Informatica in extract-transform-load (ETL) customizations. The economies-of-scale savings processes, data quality, and enterprise application are passed on to customers though a low-cost integration; Business Objects in reporting pay-as-you-go subscription model. This model and business intelligence; Oracle in customer also has the benefit of permitting customers to relationship management (CRM); and Microsoft regard the service cost as an operating expense in business productivity tools and CRM. Many rather than a capital expense, which allows all of these traditional vendors are migrating their the financial risk to be taken over an extended on-premise offerings to a SaaS model, but they period of time instead of up front. Small and face the prospect of cannibalizing their existing medium businesses, in particular, favor SaaS market and the challenge of transitioning their because of their lack of capital, infrastructure, legacy architectures to a multi-tenant model. and the resources typically required to run Most are currently offering both while they build packaged applications on their premises (known up their SaaS capabilities. Some of the pure as on-premise applications). SaaS players leading the market are Salesforce. Enterprise organizations are beginning to com in CRM and application platforms, Netsuite adopt SaaS. Pockets of enterprise business users in ERP, Google in business productivity tools and are purchasing independent SaaS applications application platforms, and Workday in human without the consent of their overworked capital management. counterparts in the information technology SaaS has driven a new implementation (IT) department. They are taking on the methodology. With no infrastructure to be responsibilities of customizing and administering purchased, set up, or maintained, a SaaS the applications themselves. Organizations are application is immediately available for also experiencing multiple instances of SaaS customization. A declarative development applications among their various departments model enables changes to be quickly designed, 258 FUJITSU Sci. Tech. J., Vol. 45, No. 3, (July 2009)
  • 3. H. Hai et al.: SaaS and Integration Best Practices Release n can integrate anywhere in the stack. In many Release 2 Release 1 cases, this integration is across layers in a Discover Define Design manner not intended by the application’s vendor Develop Project scope Iteration 1 Iteration 1 but performed nevertheless because interfaces and planning are unavailable, because organizations do not Iteration 2 Iteration 2 Iteration n Release design Deploy have the necessary skill set and do what they and construction know how to do, or simply because it is faster and Release implementation cheaper. Customers may also decide to purchase Continuous project facilitation and improvements bolt-on or point solutions from third-party vendors that cannot be upgraded or supported over the long term. These reactions may be convenient Figure 1 SaaS implementation stages. in the short term, but they raise issues for both the application vendor and customer by creating brittle connections that are easily broken by implemented, and shown to users for feedback. software upgrades, additional integrations, or Releases can be smaller and more frequent as customizations applied to the application. The a result of logical partitioning of functionalities SaaS approach to integration leverages a set of and the decreased overhead for testing and standard Web service application programming deployment. SaaS has enabled a shift from interfaces (APIs) published by the SaaS solution traditional waterfall-like development models provider. All data integration is executed consisting of single-phase sequential steps to a though these APIs over the Internet, enabling more agile methodology consisting of multiple SaaS solution providers to continuously provide phases and multiple iterations within each upgrades to functionality without breaking phase. Iterative methodologies help maximize existing integrations. Complex enterprise user adoption by incorporating user feedback integration requirements challenge even the early in the discovery, design, and development best SaaS applications today: there are still stages of the implementation. The SaaS Practice limitations and pitfalls that organizations must of Fujitsu Consulting has developed a five-stage be wary of. The cost of integration can be 30–45% methodology for SaaS implementation (define, of the overall SaaS implementation, which often discover, design, develop, and deploy), as shown increases project durations. While organizations in Figure 1. It is part of Fujitsu Consulting’s will encounter some new integration challenges Macroscope, which is an integrated suite of with SaaS, they are still faced with many of business and IT methods that help Fujitsu’s the same challenges of traditional application clients gain maximum value from their business integration. transformation initiatives. 2.1 Upfront considerations 2. Integration best practices Enterprise organizations in industries such Integration is critical to the success of as financial services, health care, and life sciences any application, especially within enterprise are subject to extensive government, industry, organizations. Its prerequisites are a bridge and internal IT data security regulations that between SaaS applications and data warehouses, are more challenging to address with SaaS entitlement, authentication, billing systems, a solutions but must be considered early in the system-of-record, and on-premise applications. process. For example, some government data With many on-premise applications, customers privacy laws ban any information about its FUJITSU Sci. Tech. J., Vol. 45, No. 3, (July 2009) 259
  • 4. H. Hai et al.: SaaS and Integration Best Practices citizens deemed confidential from being stored taking advantage of any specialized functionality outside the country. If the SaaS application offered by the SaaS solution provider. In some solution provider’s data center is located outside cases, these connectors cannot be used because the country, then it cannot be used. In the of incompatible integration software versions, so financial services, industry organizations are some of these integration vendors offer stopgap required to provide detailed audit trails of all measures through standalone appliances or correspondence with customers and prospective subscription-based pricing until the organization customers. These audit trails are typically has caught up and upgraded to the appropriate required to be archived by the organization for version. Some SaaS solution providers provide a certain period of time and in such a manner integration as part of their service offerings by that they cannot be tampered with. If an providing an integration server with prebuilt organization’s E-mails are sent through a SaaS connectors to common on-premise applications, solution provider, the organization may not which further reduces the integration effort and have records of these messages on their internal cost. Lastly, there are new emerging offerings systems and may have little or no control of dubbed integration as a service, which are the archival process or access control for them. analogous to SaaS applications in integration We must also not forget organizational IT space. They can be used for Cloud-to-Cloud, data security policies, which may dictate what Cloud-to-on-premise, or on-premise-to-on- types of data cannot be replicated outside the premise integrations. We discuss integration as organization’s four walls, different levels of audit a service in more detail in the next section. trail that must be maintained for changes to that data, or data that must always be encrypted 2.2 Integration architecture and design when stored. Lastly, there are the business Once the fundamental integration requirements of the application that may impose requirements have been established, the process complex data visibility rules that are unique to of designing the integration can begin. Given the organization’s business process and cannot that SaaS application integrations typically occur be changed. SaaS applications that have rigid over the Internet, the integration architecture visibility models may require additional work to must consider the locations of the different shoe-horn these rules into the application, which on-premise source and target systems within often compromises the integrity of the data or the the organization’s network. Understanding the application. locations, connections, and protocols between The ease of use of Web service integration these components and the other systems that available in most development tools may entice must be traversed for Internet traffic will organizations to build their own point-to-point offer immediate insight into interoperability, integration, but unless a common data integration security, scalability, and performance concerns. framework is used, this is not recommended. Most Furthermore, defining performance metrics for enterprise organizations will have some internal each integration upfront is paramount to the standard application for data integration. Many success of the project and can help define how of the larger traditional on-premise integration the integration is designed and implemented. vendors such as IBM DataStage, Informatica, In the world of SaaS integration, Web service and Ab Initio already have Web service capability integration and mashups are common, but and are also providing specialized connectors they introduce factors that an organization has for SaaS applications. The connectors can little control over such as the Internet and the significantly reduce the integration effort while different service levels of the SaaS solution 260 FUJITSU Sci. Tech. J., Vol. 45, No. 3, (July 2009)
  • 5. H. Hai et al.: SaaS and Integration Best Practices providers. Initiating a proof of concept early in to ensure that the organization will not exceed the integration discovery period or even as early its data storage limits and incur additional costs. as the sales cycle will yield valuable data about Some SaaS solution providers offer data archiving the approach and performance when combined services that can reduce the storage overhead but with measurable success metrics. may change the functionality of the application. There are a few integration patterns that Organizations should check that the volume of are commonly applied to SaaS implementations. data will not detrimentally affect application The functionality offered by SaaS applications performance, stability, and functionality. Since can be leveraged only if the data to be acted all data integrations for the SaaS application upon is stored within the SaaS application’s occur through the API and many SaaS solution data tier. This is fundamental to most SaaS providers impose rate limits on the number of applications and often leads to data replication API calls that an organization can perform within and synchronization between the on-premise a certain period of time, it may be necessary for system of record and the SaaS application. A the SaaS solution provider to raise those limits. clear understanding of the desired functionality Some SaaS solution providers support different will yield the minimum data elements requiring modes of operation for large organizations and synchronization and replication. Most SaaS have enhanced methods for expediting the load solution providers offer APIs that can facilitate of large data volumes. This can be critical if batch synchronization of data through simple an organization has a small time window for Create, Read, Update, and Delete (known as execution because of dependent integration jobs CRUD) statements. More sophisticated SaaS or global time constraints. solution providers offer event-driven near- realtime data synchronization capabilities. 2.4 Extension of service oriented Mashups are another common integration architecture governance pattern used to support information sharing, As organizations create more business information enrichment, and collaboration in functionality and integrations with SaaS SaaS applications. In particular, mashups applications they must also extend and adapt their are used to enrich SaaS application data with service oriented architecture (SOA) governance realtime back-office data presented in concert practices accordingly. SaaS solution providers seamlessly to the user. This is an alternative are continually upgrading their capabilities and to the approach of replicating back-office data creating more sophisticated APIs. Some even in the SaaS application. It is essential to match offer organizations the ability to create their own the integration capabilities of a SaaS solution custom business services. Governance must be provider with the integration requirements and extended to cover matters such as ensuring that with the pros and cons of each approach. the SaaS solution provider maintains backward compatibility in their APIs and continues to 2.3 Working with large volumes of data support any custom-built business services There are often cases in which enterprise through change management processes and organizations have large volumes of data testing; to validate compliance with IT standards, required in the SaaS application. It is important especially around auditing and security; and to work with the SaaS solution provider early to monitor and track the quality of the SaaS in the process to understand any potential services to make certain that integrations ramifications. Foresight into initial sizing and do not suffer faults or failures. There are a growth of the SaaS application data is necessary number of different tools available in the SOA FUJITSU Sci. Tech. J., Vol. 45, No. 3, (July 2009) 261
  • 6. H. Hai et al.: SaaS and Integration Best Practices industry, and there are many emerging ones for any necessary connectors or software updates. SaaS applications that can help facilitate SOA The agent provides regular reports on the health governance. and integration status to the integration-as- a-service solution provider, which can then be 3. Integration on demand monitored and managed by the organization. As more organizations adopt SaaS However, the integration itself is communicated applications with greater interoperability, an only among the on-premise agent, the source emerging integration-on-demand market has on-premise system, and either a target SaaS been created, which calls for an integration service application or another target on-premise system. that provides all the benefits and architecture Some solution providers offer virtual agents of SaaS. These services allow non-technical to organizations that do not wish to deploy any end users to customize and deploy integrations infrastructure but require access to on-premise through menu-driven wizards configuring data systems for on-premise-to-Cloud integrations. sources and targets, mappings, transformations, Some of the different integration scenarios are integration processes, and the scheduling of illustrated in Figure 2. integration jobs. The service provides a natural The pricing model for integration-on-demand model for Cloud-to-Cloud integration, a simplified offerings is much more flexible and affordable model for on-premise-to-Cloud integration, and a than that for their on-premise counterparts. conventional model for on-premise-to-on-premise Organizations are generally charged a flat integration. It also presents a platform for easily monthly fee for the connectors that they require incorporating new connectors and value-added regardless of the usage, data volumes, or deployed services, such as data cleansing and validation, agents. With some solution providers, this flat into organizations for a significantly reduced fee also covers unlimited use of all connectors, cost compared with on-premise products in a which makes the cost much more predictable and pay-as-you-go fashion. The value proposition significantly reduces the deployment risk. for integration as a service is high for small In enterprise space, we can expect and medium businesses that lack integration organizations to continue to leverage their capabilities, resources, and tools. corporate integration tools. Some will maintain These services come with a browser-based a hybrid model of an existing on-premise integration design and with management tools integration system combined with an integration having pre-built connectors to SaaS applications, service because it can offer a broader spectrum databases, and on-premise applications. They of services and connectors more quickly and offer services such as data replication, data cheaply than their on-premise counterparts. synchronization, data quality assessment, On the other hand, the on-premise integration and custom integration. For some on-premise systems will continue to offer complex integration integrations, an agent must be deployed within capabilities that their on-demand counterparts an organization’s infrastructure with access to cannot provide. Moreover, as with the adoption the source or target systems. These agents are pattern of SaaS applications in enterprises, self-contained integration servers that possess savvy business users are using these on-demand the components required to execute integration integration tools sometimes below the radar of the processes from end to end. The agent receives IT department, which raises significant security its instructions from the integration-as-a- concerns about unauthorized client applications service solution provider on the basis of the accessing corporate data sources. Eventually, organization’s integration processes as well as as these tools become more sophisticated and 262 FUJITSU Sci. Tech. J., Vol. 45, No. 3, (July 2009)
  • 7. H. Hai et al.: SaaS and Integration Best Practices Integration-as-a-service provider SaaS solution providers Integration Hosted SaaS SaaS service agent application application SaaS application Cloud to Cloud Internet Deploy, schedule, give status Design, On-premise to Cloud implement, manage On-premise to Agent on-premise Application Application Database Database Browser On-premise systems Figure 2 Integration scenarios. organizations move more applications to the was an organization-wide initiative to leverage Cloud, SaaS application solution providers will a newly created enterprise data warehouse begin to incorporate these services into their own to drive a singular view of the client across all offerings and supply out-of-the-box integration applications. between different SaaS applications at the click The goals for using the new CRM application of a button. were to improve relationship management, streamline the creation and distribution of 4. Enterprise integration case opportunities, enable the prospecting of new study business, increase the sharing of information A large retail banking firm replaced among business units, and permit a 360° view multiple on-premise CRM applications with a of a customer by any sales representative. At single SaaS CRM package. There were several the same time, the firm did not want to repeat different reasons that drove the firm to replace past mistakes related to customization and costly their existing CRM applications. Some of the maintenance. The benefits of the SaaS model and applications had been so severely customized the robust integration capabilities of the CRM that they could no longer be easily supported package played a key factor in the successful or upgraded to take advantage of new releases. realization of the stated goals. These applications were from different vendors The integration with the data warehouse and had been implemented by different required batch synchronization of visibility, client, departments, which created partitioned IT and financial account data, which consisted of knowledge, business processes, rules, and data. A millions of records. Dynamic data such as account merger with another large financial organization balances was available on demand through a drove the need to share client information easily realtime interface to the data warehouse from across different business units. Finally, there the CRM application. Opportunities and alerts FUJITSU Sci. Tech. J., Vol. 45, No. 3, (July 2009) 263
  • 8. H. Hai et al.: SaaS and Integration Best Practices were fed to the CRM system in real time and then challenges comparable to traditional application routed to the appropriate sales representative in integration. the nearest retail branch for up-selling and cross- selling opportunities. Sales representatives were 5. Conclusion able to generate prospecting lists by querying Although SaaS has dramatically transformed their data warehouse in real time through the application lifecycle, it has not significantly the CRM application, and these lists were reduced the complexities of integration with then brought into the CRM system so that the SaaS applications. Organizations are still representatives could continue to manage them. faced with vendor-specific APIs and varying The sheer volume of data and complexity of these levels of integration capabilities from different integrations required the use of an on-premise SaaS solution providers. Since most enterprise ETL tool capable of pushing millions of records to organizations require integrations to the back the SaaS CRM provider over the Internet, which office, many of the same traditional integration was probably the most challenging element of challenges are encountered. Integration-as- the project. a-service solutions are beginning to simplify Overall, the project was a success and integrations, especially in the Cloud-to-Cloud the firm was able to accomplish its business space, but they do not yet offer the same level goals through the SaaS CRM package and of capabilities as their on-premise counterparts. it significantly reduced its maintenance SaaS solution providers are starting to focus more costs. However, the complexity of the on pre-built integrations, which will help reduce integrations required contributions from many the cost and complexity of integration. However, different teams, required an open channel of system integrators will continue to play a key communications with the SaaS solution provider, role in preserving the SaaS value proposition and involved development efforts, tools, and while creating durable value-added integrations. Henry Hai Seitaro Sakoda Fujitsu America Inc. Fujitsu America Inc. Mr. Hai received a B.S. degree in Mr. Sakoda received a B.S. degree in Computer Science and Philosophy Physics from Kyoto University, Kyoto, from Rensselaer Polytechnic Institute, Japan in 1998 and an M.S. degree in New York, USA in 1996. He joined the Physics from the University of Tokyo, SaaS Practice of Fujitsu Consulting Inc. Tokyo, Japan in 2000. He joined Fujitsu in October 2003 and has been engaged Ltd., Tokyo, Japan in 2001 and engaged in implementing and integrating SaaS in system development and technology applications for enterprise customers. support. He moved to Fujitsu Consulting He has been with Fujitsu America Inc. Inc., New Jersey, USA in 2004 and has since its merger in April 2009. been engaged in technical consulting and system development. He then joined the SaaS Practice of Fujitsu Consulting as a technical leader. He has been with Fujitsu America Inc. since its merger in April 2009. 264 FUJITSU Sci. Tech. J., Vol. 45, No. 3, (July 2009)