Getting to the Point of Point Solutions

I attended a presentation at this fall's VON discussing the road from Quality of Service (QoS) to Quality of Experience (QoE). As I was expecting, it was fluff designed to justify a layer of management above yet another perfectly sound technical concept that does not need managers who fundamentally don't understand the underlying technology making policy decisions that influence the deployment of said technology.

Last year's big topic at VON was IP Multimedia Subsystem (IMS). This is another perfect example of adding a management layer to an already existing technology concept - convergence. Of course, we can't use the word "convergence" anymore. It lacks pizazz. We need to call it "unified communications". We need to implement a framework. We need IMS. In reality, we don't, but pundits and managers alike need it because some smart technical people were getting too much recognition inventing new concepts and getting work done. They needed to be stopped - or at least slowed down.

And now it's everywhere. You need Unified Threat Management - which of course is an overarching concept that covers vulnerability asessment, intrusion detection/prevention and firewall functionality. We have Enterprise Management, which is a layer on top of the classics (like HPOV) to provide dashboards and metrics. And why are the metrics needed? To justify the existence of all the managers who took the place of all the IT people whose jobs were outsourced to save money.

When I first starting consulting, a client would give us a business requirement - "I need to connect my two remote sites to share data." We would give them (what is now called a point solution) a router. Configured, tested, job done.

A few years later, a client would have the same requirement, but now we needed - and for some reason the client wanted - a Plan, Design, Implement and Operate lifecycle. This of course involved planning why they needed the router, designing the deployment of the router, configuring and implementing the router and providing knowledge transfer to their staff for ongoing support of the router.

A year or two later, the same requirement demanded a "solution". A strategic relationship with a router vendor needed to be formed. A strategic relationship with a carrier to provide the circuit needed to be executed. A strategic relationship with the implementation resource had to be established. A holistic solution was provided that included a router and a complete industry standards based, best practices configuration for said router.

And now clients want services - so integrators offer that. They own the router. They manage the router. If the client wants a tiny new function supported, they charge large amounts of cash to type a few configuration commands into the router. Pundits and integrators have done a great job convincing enterprises that the original router they needed (and that worked fine 10 years ago) is no longer adequate. It must be bundled with services and business justifications that deliver value adds.

I'm not down on the whole services thing. Unified communications is a wonderful new suite of collaboration tools that will enable distributed users to communicate seamlessly and effectively. However, no one can tell anyone what "unified communications" actually is. Cisco has their idea. Microsoft theirs. Some interoperate. Some don't. Some promise the world, but deliver nothing more than beta products with a vague "roadmap" to panacea. So outsourcers base their delivery on the visions (or lack thereof) offered by the vendors. <sarcasm>And enterprises reap the wonderful benefits of outsourcing.</sarcasm>

Reduced to it's most common denominator, if the enterprise needs a router, give them a router. The managed service router is great for the outsourcer who reaps the monthly charges from the enterprise. The enterprise is now paying a monthly recurring charge for something that could have been done as a two week one-time cost. And if done correctly, still be working 10 years later.