By Dennis D. McDonaldThis is a second in a series of articles discussing Enterprise Web 2.0 project cost analysis. The first was How Much Will Your Enterprise Web 2.0 Project Cost?
Introduction
In How Much Will Your Enterprise Web 2.0 Project Cost I introduced system integration costs this way:
How many corporate systems will the new system have to interact with? Is the Enterprise 2.0 system a completely standalone system? Or will it require some sort of interface with an existing system (say, a database of employees and their network rights and privileges?) If it will require an interface with an existing system, how much will it cost to manage that interface over time?
Technology related costs
When considering the potential system integration costs of an enterprise web 2.0 project, one of the first things to do is to identify potential system-related “ripple effects” (if any) of introducing an enterprise web 2.0 system into the organization. We also need to consider the cost implications of addressing and managing these ripple effects over time.
Traditional cost categories include “fixed” and “variable,” “one-time” and “recurring,” “direct” and “indirect.” To this list we may also need to add categories such as “additional” or “replacement” if we are considering changes to existing systems or procedures, say, through the addition of collaboration or improved user interface features to an existing client server system.
Whatever the architecture of the system being introduced, we may need to understand the impacts on what we currently refer to as data, applications, and infrastructure, as well as the impacts on development, maintenance, and support. This analysis might start with reviewing the existing application portfolio and identifying which applications or data stores will need to send or receive data from the new system. A related question will be whether a new interface system or process will be needed to support the new system, e.g., to aid in managing identities or user privileges.
I’ve put the term “currently” in italics above since such traditional distinctions can become clouded when in the Web 2.0 world we put capabilities in the hands of users that were traditionally kept in the hands of IT staff. Wherever the costs occur, though, the costs will be real and we need to decide how — and if — the costs are relevant to the decisions or projects being considered. For example, if we place more responsibility for content creation into the hands of users through the enterprise wide introduction of blogs or wikis, we may still need to consider the cost of content creation as part of overall system costs (assuming those using the system are in fact employees).
System to system communication is what Coach Wei refers to as “integration,” as distinct from the “interaction” that take place between system components and communities such as employees, suppliers, partners, and customers. Technologies relevant to integration include WOA, mashups, REST, and SOA (Service Oriented Architecture). Those especially relevant to “interaction,” according to Wei, include Ajax, RIA (Rich Internet Application), situational applications, blogs, and wikis.
As Hinchcliffe suggests in his contrived or converging post, the distinctions may not be so clearcut, but for purposes of discussing costs, traditional “bean counter” types questions are still relevant, for example:
If I have to create a new set of tools to support this new functionality, should I relate my one-time costs only to this new functionality or should I be able to “spread” these costs as an investment over other applications or functions I develop in the future?
If I am developing capabilities that overlap with existing applications, am I able to replace an existing or older application?
If I can’t replace the application, how do I treat the “sunk costs” of the existing application?
Are these sunk costs even relevant to the current project I’m considering?
Business process related costs
In addition to the system related ripple effects and the ripple effects associated with development, maintenance, and support — which have traditionally been costs incurred by the IT department — another consideration will be the impacts on business processes and associated issues of change management and user acceptance.
Given that essential functions for web 2.0 applications include collaboration and relationship management or enhancement, it is likely that a “web 2.0 project” will involve multiple systems. This could be due to the need of the new system to employ or operate on data shared by multiple systems. If this is the case, then business process changes associated with maintaining or using the shared data may also have to be modified.
When we talk about “business process related costs” we have to also consider questions of business ownership and responsibility. Business process related costs are usually heavily weighted towards staff costs associated with using a system that is maintained and supported by IT. In the enterprise web 2.0 world, even if the organizational distinctions between business units and IT are maintained, responsibilities can be blurred and this blurring may result in some blurring of responsibility for maintaining and controlling costs, for example:
If a business unit bypasses the IT department and contracts with an external vendor for the maintenance and support of a system that, due to certain Rich Internet Application (RIA) features, requires the upgrading of desktop and laptop computers to generate acceptable performance so that a business process can be performed efficiently, who is responsible for paying for memory and computer upgrades?
When business process changes require staff retraining, who pays for the training?
If newly enabled relationship management features make it easier for employees in one department to locate and communicate with experts in another department, and those experts become “popular” to the point of needing to regularly field questions generated by other departments, how should the costs of this new business process be tracked? How should it impact departmental budgets and staffing?
Business goals and business functions
Some of the questions raised above are not questions about how costs are tracked and accounted for but rather are questions related to how costs are assigned and managed in light of a company’s business goals (what the company is attempting to do) and business functions (how the company organizes itself to accomplish these goals).Just as we saw that enterprise web 2.0 systems can blur traditional distinctions between common IT concepts such as “applications” and “infrastructure,” we also may see that enterprise web 2.0 systems can blur, through their usage and operation, traditional organizational hierarchies, boundaries, and responsibilities.
If we enable employees within an organization to form relationships, communicate, and create and tag content in ways that bypass existing organizational boundaries, we still need to track the costs of what the employees are doing. We also need to understand the technology related costs of the systems that support the work these employees are doing.
But we need to understand these costs in light of the company’s strategic objectives. And this returns us to Professor Andrew McAfee’s discussion of the difficulties associated with calculating benefits of IT that I referred to in an earlier post. I maintain that, even if it is impossible to accurately or convincingly calculate benefits, at minimum we should be able to calculate costs — and cost increments and cost savings — in relationship to all the other work the company does. And we need to be able to relate these costs to business goals and business functions.
We need to ask the tough questions, questions like,
If we invest $50,000, $500,000, or $5,000,000 in system X, will it measurably improve our ability to accomplish Strategic Goal !? Or Strategic Goal 2? Or Strategic Goal 3? If the answer is “yes,” what is the basis for my belief? Are there other (more or less expensive) ways to accomplish these goals?
If we invest $50,000, $500,000, or $5,000,000 in system X, will it measurably improve (e.g., reduce the cost of) the way we recruit customers? service customers? manufacture green widgets? respond to government requests for information? train new employees? If the answer is “yes,” what is the basis for my belief? Are there other (more or less expensive) ways to accomplish these improvements?
Now, I’m not saying that answering these types of questions is easy. But these are the types of questions that companies and organizations have to make all the time. Given scarce resources, allocation decisions need to be made. The point here is that such decisions cannot be made rationally if we lack the most fundamental type of data of all — cost data. And there is no reason why we shouldn’t be able to measure and analyze costs for enterprise web 2.0 projects in ways that make business sense.