Articles

The Hyping of BPM

BPM is growing rapidly as a discipline and technology for modeling, optimizing, and automating business processes across industries. Numerous analysts’ reports and customer case studies have shown that BPM yields impressive ROI and provides many significant benefits to organizations trying to compete in a global knowledge economy.

These reports and case studies provide ample justifications to support the continued and rapid adoption of BPM. The BPM market growth rate is pegged at 15%-20% by several research organizations including Gartner and IDC which far outpaces the sub-5% growth rate of the ERP and CRM markets. BPM has gained momentum by providing tangible reductions in lag time, cycle time, and rework, as well as measurable improvements in productivity.  

Not surprisingly, the success of BPM has attracted attention including a certain amount of industry hype.  This hype detracts from the practical value that BPM generates and creates unrealistic expectations around advanced functionalities that are appealing on the surface but fall apart under close scrutiny.  It flourishes because most users of BPM are new with little or no prior BPM experience. Therefore, BPM participants are unable to discriminate the facts from the fad, and the practical from the theoretical. 

Foremost among the promoters of hype are software vendors eagerly trying to differentiate their offering in a crowded market. Many of them are relatively small companies aggressively trying to make inroads in anticipation of bigger players jumping into the market.  The software industry media also plays an important role as they seek new and “sexy” things to talk about. The mundane reality and grunt work of automating business processes is no longer newsworthy. Even though BPM results may be impressive, they have already been reported. The media must find the next big thing to talk about; especially, when it comes to a new technology such as BPM.  Finally, analysts promote and add credibility to the hype since it is their business to talk about trends that portend the future. The more seasoned analysts do add a word of caution when they talk about future trends, but the vendors, the media and the novice consumers seem always eager to talk more about future possibilities than the practical realities of today.

These days BPM hype is centered on the following five areas:

i. BPM standards
ii. Process templates
iii. Round-trip optimization
iv. Dynamic optimization
v. BAM and BI

There has been much hype about BPM standards ever since the formation of the WfMC. Since then, the industry has seen many standards come and go, including WfMC, SWAP, BPML, and BPEL among others. The pundits of standards often say that the BPM industry will never flourish without standards. Yet, BPM continues to grow fast without any signs of slowing down.  No analyst examines why the ERP and CRM markets became some of the largest in the software industry without any standards! The latest hype is centered on BPEL since major companies in the industry, including IBM, Microsoft, Oracle, SAP and others, all support BPEL on paper if not in practice. Yet, no one, not even the major software companies with deep R&D pockets, has ever demonstrated the benefit of these standards for customers. In 12 years in the BPM industry, this author has never heard of a single case where a customer has effectively ported processes from one BPM system to another using a standard such as BPEL. Standards do not become standards simply because a group of vendors submit some arcane specification to OASIS or W3C. They become standards only when they are adopted by customers on a large scale. Finally, modern BPM systems already use numerous industry standards which include: the HTML browser for user interfaces, SQL databases for store, Web Services for integration and access, SMTP/POP for messaging, and TCP/IP for connectivity, etc.  This author’s position on standards is elaborated in a separate paper which can be found at www.ultimus.com/standards.htm

Process templates are a mantra for analysts preaching rapid BPM deployment and for some vendors who have slapped together a collection of mostly simplistic process flows. They are appealing to new customers because they promise to reduce the amount of services needed to deploy BPM; therefore, reduce the cost of deployment. Why re-invent the wheel when you can get a proven process template which incorporates best practices verified in other deployments?

The case for process templates is highly appealing on the surface until one realizes some basic facts about processes.  The most challenging and costly undertaking of BPM projects are: (a) the design of the user interfaces, (b) integration with existing systems, (c) defining the organization structures and roles, (d) the database structure and integration, and (e) business rules and exceptions handling. None of these are standardized across companies; not even among companies in the same industry!

Therefore, any process template will have to be substantively customized if it is to be adapted to a company.  In the majority of cases, the cost of customization and deciding what to change and what not to change will be more than the cost of starting from scratch. Then the only reason to use a template is that it will enable the company to benefit from the best practices developed and proven by others. This sounds reasonable. What does not get mentioned in the hype is that no company which values its processes and considers them to be a competitive assets, is prepared to share its best practices with others who may potentially be their competitors! Ask your organization if it will be prepared to share its process best practices with others. The answer you will get from your organization is the same as you should expect from other good organizations who value their processes.

When organizations spend significant sums of money on BPM systems, they do so with the expectation of automating their important, mission-critical processes. The character and caliber of a company is defined by these processes. Every company believes that its processes are unique and a source of competitive advantage. They jealously guard their processes and do not want others to know what they are; especially, if they are good enough to be classified as “best practices.” Selling process templates to such companies would be akin to selling them a pre-fabricated building to house their corporate offices. The benefits of both are the same: they incorporate “best practices,” are quicker to deploy, and are lower in cost. Process templates are as unattractive for processes as pre-fabricated buildings are for the corporate office!

In recent months there is yet another new buzzword floating in the BPM market. It is variously called “round-trip optimization” or “round tripping.” The notion of round-trip optimization is conceptually very simple.  BPM systems automate business processes and can capture process metrics such as time, cost and volume in real time.  BPM systems also have process modeling tools that enable business analysts to model and optimize processes. Thus, BPM systems can “close-the-loop” by taking process metrics and feeding them directly into process models and enable business analysts to optimize the processes.  It should be noted that optimization is not automatic and does not happen by itself.  It still involves business analysts who take the process metrics, run simulations, analyze and optimize the results by using their judgment and experience of knowing what is practical and what is not. Finally, if the optimizations result in changes to the process or the resources used, the updated process has to be deployed and the resources provisioned or decommissioned.

Some analysts take this hyperbole a step further. They talk about the concept of dynamic optimization which means that the process is optimized by itself without involving simulations and optimization by a business analyst.  Round-trip optimization and dynamic optimization are myths for very practical reasons which even the BPM novice will easily understand. One has simply to look at these claims a bit more carefully in light of what is practical to understand that it does not make technical or business sense.

Consider the following simple fact. In a human-centric process, the single most important metric needed to optimize a process is task time. Task time is what it takes to complete a task in contrast to elapsed time which is the total time the task waits at a particular step. If the task time is unknown, it is not possible to optimize a process.  People are the single most important and costly resources in most processes. The time people take to perform a task varies considerably. Systems, on the other hand, perform tasks in more or less a constant time. Even if they do not, one cannot simply add or remove systems in order to optimize a process on the fly. So the bottom line is that one needs to know the task time of human steps in a process in order to optimize it. The other important thing which one has to know for optimization is the probability of the conditional paths in a process. In real-life this probability is a function of the business rules for the step.

No BPM system that we know of can capture the actual task time of human steps since there is no electronic way to measure task time. Therefore, the notion of round-trip optimization becomes a myth by simply this factor. There is no way to capture the most important data needed for optimization! The round-trip becomes one-way as soon as one starts making assumptions about the task time for each step, which is what IDS Scheer, Savvion, iGrafx, and other process modeling tools do.

The second reason why round-trip optimization is hype is because it is not easy and simple to do. The sales pitch is that customers simply have to buy the software and round-trip optimization will follow and life will be good. Round-trip optimization is not simple because changing processes involves many steps that need consensus and approval:

i. If new resources have to be added to the process, the resources have to be provisioned; people have to be hired or transferred, systems have to be brought on line.

ii. If resources have to be reassigned, then they must be trained for the new responsibilities.

iii. If business rules are changed, management, and in some cases regulatory approval, has to be secured.

iv. If the level of customer service will be impacted, then it is important to verify its impact before changes takes effect.

v. If workload on systems will change, it is necessary to get approvals from those who own these systems and see what impact it will have on other parts of the business.

All the above activities involve people, time and discussions. These decisions cannot be made dynamically and in real time. For these reasons dynamic optimization, while great to talk about in theory, is impractical in reality.

Finally, there is much talk about Business Activity Monitoring (BAM) and the convergence of BPM and BI. This is one area where technology is readily available, the potential need is real, and the benefits are obvious. So the problem with BAM is not that it is hyped as a technology of the future. The problem is that the consumer is given the impression that if they do not have BAM today, they are way behind. The fact of the matter is that the BPM industry is still in its infancy. Very few processes have been automated.

A recent Gartner research note concludes that “By 2009, 20 percent of business processes of Global 2000 companies will be supported on BPMSs (0.7 probability). By 2012, the proportion of business processes in Global 2000 companies supported by BPMSs will have reached 40 percent (0.7 probability).” [Gartner Research ID Number: G00136533,  February 2006]  If one extrapolates this backwards to 2006, the inference is that only about 2% of processes are automated today. This is very much in line with this author’s own assessment. At this early stage, the BPM industry is at the same stage as the auto industry was when the Ford Model T was introduced. The primary focus of customers and organizations should be the automation of processes which has proven to provide significant tangible and intangible benefits. When most of the processes are not automated, it does not make any business sense to engage in BAM since there is not much to measure. Asking for BAM today is like asking for a digital dashboard in the Model T! It looks cool, but is it really needed or are companies better off using their resources for automating other processes?

Hype is bad for customers and for the BPM industry. Customers should focus on automating processes in order to produce significant ROI and intangible benefits. They should not be distracted by an undue emphasis on advanced capabilities or impractical theories that are unlikely to add value. In BPM, like in other examples of automation, “better is the enemy of good.” The BPM industry will suffer if the practical benefits of new features do not live up to the hype.