The worlds of SOA, SaaS, and Web 2.0 have been swirling around each other for a couple of years now and in 2008 we'll finally see these gel into a practical, modern vision of next generation enterprises. And a variety of forces are coming together to make 2008 the year that enterprises refit themselves for the 21st century.
The driving forces for change this year will be the aging of existing IT systems, the rise of up-and-coming new approaches such as highly capable new Web-based applications, mashups, collective intelligence powered business software, Web-oriented architectures, and last but certainly not least, social software. These are providing the raw materials to use upon the freshly cleared canvases many organizations are readying for themselves as many organizations begin to retool and upgrade. Even the IT foundations we've come to get so used to, such as the operating systems we've used for years, have recently evolved and not always in the direction we're going. If nothing else, the ever-advancing computing environments of the workplace and the Web are encouraging us to move to newer and better models out of sheer momentum.
1. SOA finally goes pragmatic, Web-oriented, and lightweight. We've heard this prediction before but in 2008 it will be one of the items front and center for IT departments for a variety of reasons. Many of the ponderous, heavyweight SOA initiatives still in existence will finally refactor their design principles and then their architectures to be much more lightweight and RESTful. The classic SOA principles will still apply but changes in how they are realized inside organizations will just reach the tipping point in 2008. One key driver is that organizations are increasingly tired of waiting for ROI on their SOA investments and the demand for change is pushing IT leaders to search for new, more effective approaches. Web-orientation has enabled SOA on the greater Web on a vast scale and gained credence for a critical mass of the SOA community.
Read about ongoing story of Web 2.0 and SOA convergence, which will be well under way by the end of 2008.
The bottom line: If a Web service/open API can't be consumed in the browser, it will find itself relegated to the deep end of the back office, if not retired outright. That's not to say that the infrastructure for SOAs is getting simplier or that browser consumption is the ultimate litmus test, it's not. And as we'll see below, high velocity, large scale governance will be required to get any use out of these new highly distributable models for projecting content and functionality to any point in the enterprise.
Pervasive syndication for enterprise data, particularly with ATOM and sometimes two way, will also be a bright spot this year but will remain a largely emerging story until 2009.
2. Enterprise search will remain broken or highly limited in most organizations. I've covered previously the many reasons why search can't work in the typical enterprise without enormous effort and consequently this won't be fixed for most organizations this year. However, good enterprise search is necessary to leverage the fast growing and woefully under-leveraged information warehoused in the vast acreages of most enterprise data centers. Workers are still left with literally no choice but to pull their information from the Web or sequentially rummage through various silos to piece together what they need instead of putting a few keywords in an enterprise search engine and scanning the results. The unfortunate news: The penetration of local search engines into enterprise data will only improve a handful of percentage points this year.
3. Security will become a major concern as Web 2.0 apps and SaaS make the edge of enterprises increasingly porous. In 2008 users will self-provision themselves with consumer Web applications across the firewall, more and more business information will be found out in the open in enterprise wikis, workers will spend more time in public social networking sites, and the very pliability of mashup-based applications will make it unclear where data comes from and where it's going. This will make security around next generation platforms become a fast growing concern and the demand for solutions will grow throughout the year.
4. Unstructured information from blogs and wikis will proliferate, driving demand for solutions to extract and consolidate business information. The growth of social media in the enterprise will generate mountains of wiki pages and blog posts with valuable information that range from project status reports to list of products, acronyms, departments, teams and everything other imaginable type of enterprise information. While this will build much needed link structures between information in the organization and will enable effective search to find much of it, there will also be a growing demand to "see" all of this information at a high level and for it to be centralized, normalized, and aggregated. Tools that can effectively data mine social media applications will see adoption from organizations that have been early adopters of Enterprise 2.0 for example. Read Fracine Barone's catalog of the various ways of mining blogs or read Peter Bochner's overview of what IBM is doing in this space as good examples of what is yet to come.
5. A large number of enterprise intranets will get social networking capabilities. Popular intranet solutions like Microsoft Sharepoint have long had social features, but expect them to move to the forefront as organizations look to improve expertise location, relationship management, collaboration, and innovation. The connections workers have between each other are improved when they have easy access to the right tools that effectively help them manage their relationships. Look for the social graph to start to move to the center of attention -- replacing or reinventing the traditional address book -- and become the organizing structure that communication and collaboration revolves around, for both traditional channels such as e-mail and IM, as well as newer social media/Enterprise 2.0 applications. Companies such as KickApps are making it possible for insert access to social networks by the use of widgets wherever needed on any Web page or applications.
6. Enterprise mashups will make their appearance in a widespread way but won't take off big, yet. I wrote a list of reasons exploring why end-user mashups wouldn't happen in 2007. Since then a couple of those reasons were partially addressed and a few more will be by the end of 2008. Despite some shortcomings in capability and the early state of the industry, we'll see mashup tools being deployed this year by early adopter enterprises and used by IT departments and power users to deliver applications via visual assembly and little coding, leveraging the resources of the local SOA and global SOA to build composite applications with real business value in hours instead of weeks. The average business worker will need time, possibly several years, to absorb the possibilities of this new model for personal business application development but it's highly likely we'll see the first significant deployment of end-user assembled Web applications in 2008. Expect 2009 to be the year enterprise mashups begin hitting the tipping point of adoption.
7. Budget for next generation governance infrastructure will be demanded by IT to manage the proliferation of SaaS, new SOA, and mashups. Governance was another big enterprise watchword in 2007, but there was still too much steering committee and not enough operating governance infrastructure. A deep seated need for effective, rapid, decentralized, and pervasive governance with declarative, run-time control will develop as SaaS and mashups (and the services need to fuel them both inside and outside the enterprise) proliferate. Policy-based solutions that set SOA and SaaS users free but keep them within reasonable boundaries, like those from AmberPoint and Layer 7 will go a long way towards solving these problems. I put this prediction in budget terms here since it will likely be widespread grassroots behavior outside of IT that will drive the need for these solutions within IT departments.
8. Collective intelligence platforms and enterprise decision management support applications will see significant early adoption this year. Prediction markets, such as those from Inkling Markets and Consensus Point, began to get on business leader's radar in 2007 and are poised for early growth this year and are ostensibly the end game of "real" Web 2.0 applications: harnessing the intelligence on the edge of our networks. Predictions from companies like Gartner are also putting a lot of stock in enterprise decision management (EDM) this year but both will see only early adoption this year with 2009 being the primary growth year.
9. New rich user experience platforms such as Adobe AIR and Microsoft's Silverlight will see widespread use in the enterprise in 2008. Ajax remains a popular approach in terms of mindshare for a variety of reasons but well-designed rich user interface platforms created from the outset to bring compelling rich Internet application development capabilities to enterprises will see broad adoption in 2008. Ajax will remain too complex and difficult to develop with for most corporate development shops beyond simple eye candy.
10. Mobile devices, such as the iPhone and especially the BlackBerry, will see significant uptake for embedding workers into business processes and Web-based collaboration. Mobile business applications, already growing in popularity, will literally explode this year as businesses find they can easily improve the pace of work and keep far-flung and increasingly virtual organizations both effective and cohesive. Approval processes and workflow applications in particular will be exposed as mobile devices early on in many organizations. Freeform business information retrieval applications similar to ChaCha are also likely to become popular because of their friendliness for mobile use. Mobile social networking for business will also be a significant growth area in 2008.
11. Web 2.0 versions of traditional enterprise applications will struggle for marketshare amongst the software giants. The sales, marketing, branding, budget, and marketshare advantages of the large entrenched software vendors such as Oracle, Microsoft, and SAP will make it a tough go indeed for Web 2.0 enterprise apps of the CRM, HRM, and ERP variety. These types of applications are often the least subversive of Web apps as well. Despite this, there is a lot of room for growth in the small and medium-sized businesses (SMB) and most significant adoption of line of business SaaS and Web 2.0 apps will occur there.
12. Enterprise 2.0 will become a standard feature in most organizations, while views on the outcomes of deployments in organizations will be viewed through the lens of predisposition. Last week I wrote that one of the big Enterprise Web 2.0 stories of 2007 was that E2.0 had arrived. This will only be truer in 2008 but the opinion on the results of Enterprise Web 2.0 will be colored by the preconceptions of those that observe them. The flattening of organizations, the non-hierarchical communication, the free and open exchange of information encouraged by Enterprise 2.0 platforms will be looked on as potentially disruptive by some and as a breath of freedom by others. Neither will be entirely right but the apps will stay in place with tweaks and controls added along the way to address the inevitable and unfortunate misuses that will be encountered during their adoption.
Be sure to catch some of the other excellent prediction lists including Phil Wainwright's terrific Eight reasons SaaS will surge in 2008 and David Chappell's insightful predictions around the 2.0 era of architecture.
For a broader, non-enterprise perspective, read my general predictions for Web 2.0 in 2008.
What did I miss? That most enterprises didn't actually open their SOAs up to the Web? How about that, as Vinnie Mirchandani says, Web 2.0 apps are still stuck in between the margins of "real" business apps? Please leave your comments in TalkBack below.