This is the first in a series. The second is System Integration Costs and Enterprise Web 2.0 Projects.
Professor Andrew McAfee has an excellent series of posts related to the application of “Return On Investment” (ROI) calculations to enterprise IT. His posts are here and here and deserve a close reading. He gets some flack for seeming to argue against measurement, but as he explains in his second post, that’s not his point; he’s basically saying that too many business cases that rely on a haphazard or incomplete calculation of IT benefits are flawed, incorrect, incomplete, or self serving.
I agree with many of his points. As a consultant I know firsthand how difficult it is to quantify IT related benefits. Benefits can be qualitative or impossible to measure, they can be difficult but not impossible to measure, they can be easy to measure (an example being cost savings), but they frequently involve estimates that are, well, estimates.
How does such a basic framework impact the costs of implementing an “enterprise 2.0” type of technology-supported project? Much has been said about the low barrier of entry to implementing web 2.0 technologies. Does this carry over to doing a corporate implementation of, say, a standardized blogging system that will support internal and external users? How about an expertise management system that automatically pulls and tags employee subject competence based on an analysis of email traffic? How about a wiki that is used by supply chain staff around the world to document current policies and procedures?
Here are a couple of questions I think are useful to ask about these types of situations.
- What are the costs being compared to? Do you know what the current costs are for the processes that the Enterprise 2.0 system supports? Can you estimate costs over time with and without the Enterprise 2.0 system?
- 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?
- How well can you predict? If you start small and depend on “viral effects” to lead to a gradual adoption of the new system on a wider scale, can you track the costs of the introduction in order to be able to better estimate what future costs will be as the system becomes more widely adopted and used? (Here I’d opt for an unobtrusive system that automatically tracks usage time.)
- Will the new systems add new or change business processes? The ongoing cost of supporting a business process — whose impacts should be ultimately traceable to a strategic business goal — is a mix of the costs of people time and other resources including technology. If the business process is new that is supported by the Enterprise 2.0 system, can you estimate what the new costs will be? What will be the impacts of supporting the new process on other processes? (A simple case: to maintain a corporate blog will you be able to re-allocate staff or will you have to hire new staff? If you reallocate staff, what happens to the business processes they currently support?)
- Will the new system(s) require changes in infrastructure? If the new system is 100% remotely hosted, it will still need to be managed. If its development and support requires IT staff to learn new skills, build new libraries of code, or learn new languages, what are those ongoing costs in relationship to existing maintenance and support costs?(Here it may help to consider fixed versus variable costs especially if new technology is required to be adopted that may also be adapted to support other systems.) Also, can the current population of desktop PC’s and laptops manage the local processing that the new system might require?
The most important question about Enterprise Web 2.0 project costs is being able to answer the question, “Cost compared to what?” You need to understand the baseline (or current costs), and you need to be able to estimate what the changes (or deltas) will be. Only then can you begin to answer the legitimate questions about benefits raised by McAfee in the posts linked above.