It’s happened three separate times this month. In hindsight it’s obvious but it amazes me that large organizations whose staff should really know better continue to bite off larger enterprise-wide implementations than they’re capable of delivering. The problem is, to be sure, insidious. After all with “enterprise-ready” software hype available in every publication, you might be lulled into thinking that just choosing the right software means that everyone in your organization will instantly become an integrated machine with all parts harmoniously moving in the same direction. The problem is, implementing enterprise-wide software of any category is much, much more than the purchase of the package itself.
I’m sure no implementation starts out this way but, for the three companies who I spoke to about it this month, they certainly ended there. I’ll not name these firms but, you’d recognize each of them instantly. One, an insurance company, the second a financial organization and the third a telecommunications company. In each, the issues were similar yet distinct. Each of these firms has an interest in making themselves more efficient by implementing project-oriented software. Sounds simple enough but even in this sentence the mischief starts. “Has an interest”. Well, firms don’t really have an interest, people do and if a firm is large enough, it is represented by many, many, many people. Part of the problem with implementing any kind of solution that will affect everyone in the enterprise is finding out first, how the enterprise will determine its wants and needs. Should it use the democratic method? Everyone puts in a vote and the solution with the highest number wins? Should it manage by consensus? Dissenting voices are managed one by one until there is some kind of movement in a single direction? Should the decision be made hierarchically? The executive at the top of the food chain decides and everyone else follows the lead? The answer to all of these may be yes… or no. Each firm’s management, for each situation must determine what degree of seriousness to allocate to an enterprise-wide plan.
I/S managers who might have spent years cajoling, teasing, arguing, forcing, enrolling and trading with their fellow managers in an attempt to generate enough consensus to move an enterprise-wide solution forward are often given the full authority of management to implement such solutions. In the past year I’ve seen Boards of Directors actually make operational decisions on software and on implementation schedules, something that in the 5 years previous would have been unheard of. With that kind of authority, anything can be implemented.
Yet, not every solution carries that much weight behind it. How do other enterprise-wide solutions get chosen and implemented?
In the case of the insurance firm, a committee has been struck with a mandate from the highest levels of management. Yet the mandate wasn’t clear in how it should be accomplished. In this firm, a couple of mainframe-based applications run high-level project tracking for the finance group. The goal is to eliminate these cumbersome systems and replace them with more modern reporting systems. The first idea was to look for a commercial system which closely matched the existing functionality. Surprise, surprise, no one has exactly the functionality required in a single system. While the requirement specification was being created to even evaluate systems however, the project increased in scope. Why look for obsolete functionality, the argument went, when we could create a state-of-the-art enterprise-wide project control environment? It sounded wonderful and the team was inspired. Now the search was on for multi-project, enterprise-wide systems which could integrate the work of thousands of employees and hundreds of supervisors, project schedulers and managers.
It sounds wonderful, but there was a small fly in the ointment. The majority of end-users are already using desktop project scheduling software from a funny little company based in Redmond, WA. While very popular among end users, this software was deemed insufficient for managing the centralized, cross-enterprise needs of the firm. “Did the committee,” I asked, “have the mandate to replace this software across the board?” “Well… no.” was the answer. But, some committee members thought it was such a good idea that they were determined to lobby for the notion. It sounded like fighting the good fight but without the backing of the highest levels of management, this committee was destined for failure. Trying to replace the desktop product would ultimately be so contentious, raise so many issues and concerns, step on so many toes that the committee would be handcuffed in implementing anything.
At the telecommunications firm, the news was similar. Here, one division has successfully implemented a project management system across their entire group and have come to the conclusion that if only the other 30,000 employees of this firm would see things they way the 1,000 of them do, everyone’s life would be easier. Nice plan, tough execution. This firm, like most high-tech telecommunications firms these days is enjoying explosive growth and changing directions and goals faster than can be imagined. The only way these firms survive is to compartmentalize their operations, leaving each group somewhat autonomous with significant discretionary power. The individual groups agree on virtually nothing. It is, in some ways a problem but it comes with the territory. Implementing an enterprise system that could only be used successfully if everyone agreed on how it would be used, how data would be analyzed, what quality of data would be included and last, but not least that all … every bit of data would be entered into this particular system is going to be very difficult. The original idea was valiant but in this case, even if the CEO himself thinks it’s a good idea, deploying such a system is going to be tremendously difficult if it’s even possible. This presumes of course that the product(s) that are selected and that they attempt to implement can even carry the load across such a large number of users with so many diverse interests.
In the case of the financial company, reason has somehow prevailed. Here, the original notions of an enterprise-wide planning system have been toned down. Technicolor plans of real-time resource management with instant management oversight have resolved to a more modest first phase approach. This firm has elected to implement only one of a wide range of elements that may one day make up the final solution. In this case, the first element is an activity-based timekeeping system which will, of course, touch every employee in the firm, yet is already somewhat culturally accepted. This first phase, will deliver some of the results management has been looking for with an ability to closely approximate some others without having to deploy tools or collect data from every user.
The lessons seem to stay the same. If you are contemplating implementing an enterprise-wide solution consider some of the following elements:
Ø Do you have the authority to do so? Has someone high enough up in management; someone with the authority to impose this solution if need be, requested that this solution be selected and implemented and; do they understand what kind of commitment will be required of them if the plan moves forward?
Ø Is there a need? If end-users are already satisfied with whatever solution they may be using at the moment, how will you convince them to switch? What benefits will this solution provide to each level of the enterprise?
Ø Finally, consider a phased-in approach. After all, if it’s all or nothing, then you’ll need to wait until the entire system is selected, installed, configured, loaded with existing data, tested, people trained etc until the first byte of useful return will come out. This may take months years or… forever. A phased-in approach lets you get small victories with benefits along the way. It’s true that the final dream system may stay forever outside your grasp this way but… was it ever really within your reach anyway?