Articles

The State of BPM: Perspectives of an Industry Insider

The BPM industry is awash in hype right now.  Industry trade shows draw larger and larger attendance.  Media coverage of the space is greater than ever.  Major analysts continue to expand their BPM-related practice groups.  An entire BPM professional ecosystem has developed on a large scale in just the last three or four years.  What was previously an IT backwater is not only mainstream but sexy.

Even so, overall market growth for BPM software, measured by the larger public BPM companies that come primarily from the integration side of the market seems to be lagging.  The relatively inexpensive acquisition of Fuego by BEA indicates that Fuego’s revenues and/or growth may not have been particularly impressive.  A closer look at private BPM companies may reveal a similar story – growth is there, but rocket ship growth that one would expect to see given the market hype has not yet materialized.  BPM still hasn’t broken out.

Business Model Issues
Part of the problem is that BPM as an industry is burdened by dated license and revenue models that are more mid-nineties than late two-thousands.  BPM continues to be sold in large, lumpy perpetual license deals by expensive direct sales forces, while the rest of the software industry has moved to software-as-a-service, pay-as-you-go models.  A typical BPM implementation is less expensive and less time consuming in terms of both human and hardware resources than traditional enterprise applications like ERP or CRM, yet ease of implementation and time to value lags far behind models like Salesforce.com which have now become mainstream.

Business model issues are not solely responsible for the lag in adoption.  Simply delivering a hosted, on-demand BPM variant is not a panacea for solving BPM’s adoption issues.  On the contrary, every effort to deliver SaaS-based BPM to date has failed because BPM as a technology where the value is inextricably linked to integration services is ill-suited to a SaaS model.

What’s the use?
To truly understand the state of the BPM industry today and the reasons why BPM continues to have such a high hype-to-reality ratio, it’s critical to look beyond the marketing to understand how BPM is actually being used in the real world.  The state of the market reveals itself when the practical use cases of BPM are examined.  Today’s cases also tell us a great deal about how the technology needs to evolve going forward.

The big marketing myth of BPM is that it is all about process improvement.  The core value proposition articulated by every marketing department of every mainstream BPM company includes continuous process improvement, process flexibility, and enterprise agility.  Yet the promise of process improvement does not seem to be the sales driver for most real BPM deals.  Here the evidence suggests that companies aren’t buying BPM for its process improvement benefits.  If they were, more organizations would be interested in simulation functionality, which is essential to optimizing the behavior of modeled processes.  Integrated simulation remains the functional basement of the BPM suite market.  BPM deals are rarely impacted by a BPM suite’s integrated simulation capability, and BPM vendors have very rationally responded to the general lack of interest in this area by de-prioritizing simulation investment.  

If BPM isn’t really all about process improvement, why do organizations purchase it? A review of actual, practical usage of BPM suites today suggests an interesting reality.  BPM suites are used primarily for two purposes: application integration and application development.  Companies choose to go with BPM when it is the most cost-effective option in one of these areas.  Process improvement and optimization is at best considered an ancillary benefit.

Notably, Forrester breaks up the market in this way, featuring two different Wave evaluations for “integration-centric” BPM and for “human-centric” BPM.  The latter category represents the application development use case, which typically requires a higher- level support for human-process interaction.  This segmentation of the market is becoming increasingly artificial as the two different foci of BPM vendors continue to converge in integrated suites.  Nevertheless, Forrester’s bifurcation implicitly recognizes that these two use cases are really what BPM is all about.

Application Integration with BPM
The integration-centric use case represents the majority of BPM market activity by dollar volume.  However, it’s easy to overestimate the real growth in this market.  Sales of BPM technology for integration purposes are dominated by existing integration vendors such as TIBCO, WebMethods (now Software AG), IBM, and BEA Systems.  As a result, it’s difficult to quantify what represents organic growth of the market in this area versus simple substitution away from older integration technologies towards BPM.  In some cases, the actual integration technology delivered by these vendors in a BPM sale has changed very little – adding focus on a graphical modeling tool and support for BPEL does not make a BPM suite – but their marketing has shifted to take advantage of the BPM hype.  Those integration vendors that have done acquisitions (TIBCO plus Staffware; BEA plus Fuego) are most likely deriving the bulk of their “BPM” revenue from sales of pre-acquisition integration technology, not from sales of technology from the acquired companies.

The benefits of BPM for application integration are clear and have been well articulated elsewhere.  The big opportunity for BPM is not to supersede existing integration technologies.  Given BPM’s inherent ease-of-use advantage (with a graphical model based on a known standard that is portable across different BPM engines), the fact that BPM is a natural compliment to a SOA, and the fact that by now all the market leaders have transformed themselves into BPM vendors, it is inevitable that BPM-type approaches will dominate this market.  The true upside opportunity for BPM is to evolve into a platform that supports rapid application development, change, and integration with a visual model-centric paradigm that represents a clear advantage over previous application development approaches.

Application Development with BPM
On the application development side, BPM is being applied primarily in custom application development.  The potential market for custom applications is a large one, but the penetration of that market to date has been lackluster.  Most organizations remain very much in the “learning about” BPM phase; some have crossed into “experimenting with” BPM; only a very few have widely deployed it in this application development context.

There are a number of common use cases typically seen in this market, all of which feature prominently in the announced deals of most BPM vendors.  These use cases include loan origination, case management, claims management, and other types of exception management.  It’s interesting to note that these are generally document-centric processes that can be supported by the traditional human interfaces of BPMSs – task lists and forms.  This seems to signal that the technology in this area is still relatively immature.

Some BPM vendors have started to evolve their application development capabilities to the next level.  Lombardi has delivered greater sophistication in terms of its support for extensible and easily maintainable data models.  Appian has delivered user interface innovations that enable development of BPM-based applications that look and function like traditional purpose-built enterprise applications.  Breaking the “task list and forms” paradigm is an exciting development that could substantially expand the list of application development use cases for BPM into more transactional processes typically owned by ERP, CRM, and SCM.

If hosted BPM is ever good for anything, it will most likely make its mark based on its potential for application development – specifically, development of standalone apps that only require relatively simple, web-service based integration.  Unfortunately, completely codeless application development is still a ways off for even the most advanced BPM suites.  Given the state of the art in BPM today, if hosted BPM takes off at all it will need to be driven by largely pre-built application templates and a delivery model similar to Salesforce’s AppExchange, but with more powerful process flexibility for end-user customers.

People Don’t Process Model
The major problem the BPM industry confronts is process modeler adoption.  People simply are not adopting graphical process modeling as a business methodology en masse.  Modeling takes training.  IT users that can readily grasp mapping data flows and transformations using visual tools must be trained in how these techniques extend to describe human-to-system interactions.  And business users that are trained to model cannot develop executable models in BPMSs because executable models require further skill in data modeling and integration technologies to create.

Couple these gaps to the inevitable challenges organizations have always had in bringing together members from disparate organizational groups to define cross-functional business processes, and it’s easy to see why adoption of process modeling is nascent.  In the real world, “process analysts” titles are about as rare as hens’ teeth, and where they do exist it is in the relatively rarified air of the world’s largest financial and manufacturing corporations.

The BPM industry has only just begun to address this critical hype-to-reality gap.  There are a number of notable approaches here to date:

• Standards.  In this area, one standard rises above all others: BPMN.  The emergence of BPMN as the widely accepted lingua franca for process notation is a critical step.  However, BPMN notation has to date gained little traction outside the relatively small BPM community.  Business users are highly unlikely to spontaneously adopt BPMN notation the way they spontaneously adopted the spreadsheet almost thirty years ago.  XPDL, the XML Process Definition Language developed by the Workflow Management Coalition, provides serialization of BPMN process models, and interchange format transferring models between applications.  Yet it similarly suffers from a lack of visibility outside of the BPM community.  The BPEL specification is incomplete, failing to encompass human-to-process interaction, and is primarily a specialized tool for IT users in an application integration context.

• Free Modelers.  There are a few BPM vendors that are now offering free process modelers and related tools in an effort to drive adoption and encourage sales leads.  The early innovator in this area was Savvion, with its free downloadable modeler.  More recently, Lombardi Software has released an innovative new hosted tool, called Blueprint.  Blueprint is not actually a modeler per se, but rather a process capture tool.  Lombardi’s marketing says it best: “Most process analysis and modeling tools make process discovery very difficult.  They are complex or unstructured or do not foster collaboration.” In effect, Lombardi Blueprint takes the modeling out of process modeling.  Simplified interfaces capture information such as process participants, inputs and outputs, forms, and metrics.  Once captured, these are automatically consolidated into a BPMN-graphical model.  The idea is to drastically simplify the challenges of process documentation.  If successful, the concept would seem to eliminate the need for basic business user training in process modeling.  But the strategy completely ignores the integration and data model complexity in developing executable process models.  Only time will tell if Blueprint makes the transition from ineffectual toy to valuable tool.

• Communities and Social Network-based Approaches.  Again, Savvion has taken the lead in this area, with significantly less success than their free modeler download.  Savvion’s ProcessXChange is an interesting concept, but for all intents and purposes it was stillborn.  A visit to the site a full year after its launch reveals almost no use or activity.  There is very little activity on the discussion forums and no strong indication of significant user interest in the process models that Savvion has pre-loaded into the site.

Until the modeling adoption problem is solved – and until modeling itself becomes much easier – BPM adoption will continue to lag, the bulk of BPM implementations will continue to focus on integration, and the BPM market (human-centric BPM in particular) will never achieve explosive growth.

Conclusions
For companies considering a BPM investment, IT should lead the charge.  Vendor hype has created market confusion about the proper role of IT with respect to BPM.  BPM is not a threat to IT.  BPM will never enable model-savvy line of business owners to displace IT from its role in developing, integrating, and maintaining enterprise applications.  On the contrary, BPM is a technology that should enable IT to engage more proactively with line of business operations.  The process model-as-executable will inevitably be owned by IT, but the process model-as-business description can easily be a shared, visual communication mechanism that brings business and IT closer together.  IT should be proactive in engaging and training business users in how to at least understand modeling notation and the benefits derived from taking a model-centric approach to application development and integration.  By taking an active role in education and thought leadership, IT will also firmly position itself as the gatekeeper between an increasingly process-aware enterprise and the process-based applications that form its backbone.  

Vendors ought to be concerned that companies aren’t buying BPM for its marketed benefits.  Vendors should also see this as an opportunity.  Vendors should recognize that their customer is the organization responsible for building and integrating applications, typically in IT.  Pure play BPM companies need to become more aggressive about selling and marketing directly to IT and stop believing their own rhetoric about line of business customers driving sales.  With its familiarity with code and visual tools for development, the IT audience is far more likely to programmatically adopt process notation than is the business audience.  If the pure-play BPM vendors fail to focus on the IT buyer, integration-centric BPM competitors will ultimately end up being the big market winners as the two approaches converge and their offerings become better suited for application development use cases.

 

 

About The Author: Kevin Spurway

Following his distinguish tenure in the BPM software sector, we invited Kevin Spurway to share is candid assessment about the state of the BPM. Previously he had held senior management positions in marketing, product management, and business development at technology companies including Rave Wireless, WebGen Systems, Appian, and Parametric Technologies (PTC). As Appian's VP of Marketing, he helped grow the company to a recognized leader in the business process management (BPM) space. At PTC, K