From the consolidation in the telecom and fiber sector over the last four years came fewer and larger companies and difficult integrations. Two of the most aggressive rollup artists were Level 3 Communications (LVLT) in the first half and Zayo in the second, each assembling a vast array of mixed assets. The two are not really comparable in size or geography of course, yet the systems they are unifying are similar and their approaches to patching all those different processes together have been and remain fundamentally different and interesting.
At Level 3, the focus has been on “Unity”, a total reformulation of their internal processes onto a common new platform. It is a top-down approach, where one specifies the larger components first, then later details subsystems in greater detail at each successive level. That the subsystems fit into the whole is perhaps more important than how well well designed they are on their own. While some external building blocks are used, this is essentially a proprietary approach which is dependent on internal expertise and hopefully results in a long term competitive advantage. As that ‘hopefully’ implies, a top-down approach carries the risk of failure, that one could expend vast resources and not achieve any advantage or worse. In the case of Level 3’s Unity project, for what it’s worth, the grapevine has both very positive and very negative views but I think the jury (properly) remains out on the results until completion in 2009. But it is no surprise Level 3 chose this path, from the beginning they have looked to internal development for their company processes.
At Zayo, the focus has been rather different. From BearonBusiness.com, we have learned that they use Salesforce.com heavily to track both sales and now installs, Google Groups for self-organization, and yes blogs for both public relations and intra-company communication. This is a bottom-up approach, piecing together an solution from smaller focused pieces. The design of those underlying pieces is perhaps more important than the fit with the whole. The pieces are ones anyone can use, and are customized only at the highest level by Zayo itself. Most of the underlying technology is designed, written, and maintained externally. And most of the functionality is designed by those external creators to use the internet to empower the individual or the small group. The uniqueness of Zayo’s result depends on whether the value of the whole exceeds the sum of the parts. The hope is that by supplying a framework of tools and encouraging people to exploit them, an empowered workforce will produce an emergent company. The risk of course is that nothing is proprietary, everything might theoretically be copied if it is shown to work – though of course this is harder than it looks.
Which approach is best? As in software, both have their uses. I have chosen Level 3 and Zayo only because they both bought many others and are so wildly different in outlook, there are others out there in both camps and plenty in between. And it’s not really fair to say that either Zayo or Level 3 ought to do what the other does, because they *aren’t* the same. Zayo’s smaller size gives it more flexibility and less fear of organizational chaos, for example, and they should enjoy such benefits of youth while they can.
Personally, I prefer the bottom-up empowerment solution simply because in a top-down approach there are too few brains making the decisions to ever hope to come up with an optimal solution, no matter how intelligent they are. Internally developed software invariably takes more resources to produce and maintain than people think it will when they start out. At the same time, I think the bottom-up alternative is very difficult to do right, it takes an energetic, motivational leadership style such as that of Zayo’s Dan Caruso which is actually quite rare. I also suspect it becomes much, much more difficult to manage as a company gets larger and more complex. But of course, both approaches are superior to the Worldcom method, i.e. do nothing, wait a couple years, and call it done.
If you haven't already, please take our Reader Survey! Just 3 questions to help us better understand who is reading Telecom Ramblings so we can serve you better!
Categories: Internet Backbones · Mergers and Acquisitions
Join Our LinkedIn Group
Thanks Rob for the post and the kind words. Time will tell. You say “I also suspect it becomes much, much more difficult to manage as a company gets larger and more complex.” I disagree completely–in fact, just the opposite will prove to be true in my opinion.
I would welcome the opportunity to show you what we are doing. Seeing is believing–and you know enough about the guts of telecom to appreciate what we can show you. As an enticement: I used the word “show”, not the work “tell”…this is a key attribute to how we are going about the difficult task…it will make sense if you take me up on my offer.
I think I’m going to agree with Dan here. A bottom-up, loosely coupled architecture tends to “give” more elegantly during even the most subtle shifts in requirements dictated by the larger environment than a top-down, tightly-coupled one. As the network scales, however, there is also a danger with the bottom-up approach if the number of one-off designs get out of control. Within reason, point-solution design templates can help to limit the number of variations from one site to the next in a bottom-up design, or from one general application area to the next, thus reducing the total number of skills to be mastered by operating staff.