Over the years, there has been a shift away from certification being considered relevant as a measure of admins' skills. The proliferation in the late 1990s of test-only admins led to the term "paper MCSE," referring to people who passed the required exams to gain the Microsoft Certified Systems Engineer but had never done actual engineering work.
The use of multiple-choice questions that often test only the most granular level of knowledge (say, a PowerShell cmdlet you might use one time in your entire career) has continued to diminish the respect accorded of admin certifications.
Should admins even bother with these certifications today?
Although there is certainly a need for changes to the existing certification path, I believe the quest for certification should be encouraged, even incentivized by management. I see the tangible value of certification at a company I work with, which has pushed for its employees to seek certification — specifically for Exchange 2013 and Office 365.
That company initially recommended certifications because Microsoft's partner programs require having certified staff members. For example, to be a gold partner, you might need four staff members who have passed a certain technical certification. If you’re short on certs, the best approach is to encourage staff to get certified, with incentives tied to completion — quid pro quo.
I'm working as the training provider, and nearly a quarter of all employees have jumped into the program. Not everyone will take the exams, and among those who take them, not all will pass.
But one certainty prevails: Since the program has begun, the level of Exchange knowledge among those employees has increased dramatically. Studying for an exam improves your skill level, confidence level, and technology grasp. The residual benefits of seeking certification for the partner program requirements reconvinced me of the value and relevance of certification.
But there continues to be a disconnect between the real work of an admin and the questions they get in an exam. This has led to the proliferation of sites where the questions and answers are posted online for all to debate and study because the possibility of passing the exam based on real-world knowledge is nil. In a perverse response to these sites' publication of questions and answers, exam creators have made the questions even more obscure, thus propagating the cycle.
If you want to test a person’s ability to paint a picture, play a sonnet, drive a truck,, or admin a server, you can't realistically use a multiple-choice test. You can use a multiple-choice test to ensure the person has a basic grasp of painting, music, driving, and admin concepts and issues. That's what these certification exams should be: a way to check a person's basic knowledge for a product, especially when there have been changes from one version to the next.
Instead of calling such exam-passers "engineers," it'd be more accurate to call them "solutions consultants" because they’ve proven they understand what products Microsoft offers and, based on scenarios, when to offer one over another. When taking such exams, I grin and think, “Well, if you’re not up to date on the latest features, you’re not going to get this one right.” That should be the point.
However, there will always be people who want to prove themselves on a mock battlefield, to show they have the ability to truly perform the tasks of an admin. For that need, there's long been talk of simulation exams where a person has to perform the install, configure the solution, and prove it works — then to break it, so the admin has to troubleshoot and fix it. Such as exam trial would truly hold value.
The problem? It’s not a cost-effective, scalable method for administering exams globally — yet. It would be a huge financial undertaking for Microsoft with very little payoff. That kind of ROI thinking is what led to the demise of the Masters (or "Ranger") program, in which people would travel to Microsoft, study for weeks, and go through the trials of setting up products and troubleshooting them. The cost of running the program made it impossible to scale.
Perhaps it will take a CompTIA or other exam provider to come up with a better model that truly allows admins to show they have the right stuff.
Whether or not this happens, I am increasingly confident that certification exams have value for an organization. They should be encouraged, incentivized, and valued for what they are: a good way to assess how aware people are of key concepts and issues.