...and what can Microsoft do about it?

Interesting how these things dovetail together. After my rant yesterday, I read this post from Viral Tarpara, where he addresses the issue of 'dumbed-down' marketing because it may be what some customers respond best to. But along the way he addresses the issue of "IT Pros" who don't have a deployment strategy for their next Windows rollout. This ties back into the topic of retention of talent in desktop engineering.

I think a big part of the problem is that there's no way to get recognised for desktop expertise beyond the bog-standard 'desktop support' qualifications which are considered (rightly or wrongly) to be pretty entry-level. The assumption, as I said yesterday, is that anyone who is any good will have moved on to server technologies after a year or two, and only those not good enough for 'promotion' will still be at the desktop. (Not to mention that 'server' products are where the money is.) And I believe the lack of a good desktop certification track only strengthens that view.

MCSE has been widely recognised for a long time as being a qualification that requires a certain amount of expertise and investment of time, but the "supporting Windows desktop" tracks have never gone much deeper than knowing how to uninstall printer drivers and a basic understanding of the results of IPCONFIG.

Some of us in large enterprises don't suffer as badly; we have tiered hierarchies where there's a reasonable chance a hiring manager will understand the value of a skilled desktop engineer, and might have a clue of how to recognise one. A typical hiring manager at an SMB would probably have wider responsibilities so might not be aware of the value of such a person, or be confident that they could pick one out from a lineup, so they rely on tangible things like exam passes and certifications. And if there are no 'engineering' qualifications for the desktop, then surely that means there's nothing to engineer? Just do a vanilla install (or even easier let your hardware supplier start shipping the new OS as your kit gets replaced) and let the support guys figure it out how to make it fit your environment as they go along. Those who hire a firm to do major deployments are likely to go with the cheapest contractor who "gives them what they want". Engineered equals fancy and expensive, they just want something simple, right?

It's going to be the exceptional individuals who don't mind swimming against the tide who also have a passion for desktop technology and don't actively dislike dealing with end-users (can you see how the field is starting to narrow?) who stay the course and become truly expert to a point where they can, for example, create complex automated deployment solutions that scale to thousands of users or are simple and robust enough to be passed on cheaply to multiple SMB customers.

Perhaps Microsoft could help things along by promoting something similar to the BDD MCTS (does anyone outside a very small group of people even know it exists?) and including it in a wider track including e.g. elements of desktop management through SCCM (rather than setting up SCCM architecture), GPO (rather than setting up AD), topics around security that relate to the desktop (rather than focusing on perimeter security), etc. And call it an MCSE so people know it requires the same level of expertise as supporting an Exchange environment or an AD domain! If our skills were better recognised, perhaps more technicians in their first few years working with the desktop would be encouraged not to jump ship to an entry-level 'server' job kicking off VM installs or swapping out disks in a datacentre.

If I sound at all bitter, I have no reason to be. I'm lucky to have worked in some great companies where the end-user experience is prioritised and the investment in desktop engineering has reflected that. I've worked on teams with highly talented people where we've built some truly best-in-class deployment and management solutions for the desktop, and I've done pretty well out of it, thanks for asking.

But I'm only here because I'm a stubborn creature who stuck with something when everyone else told me it was a bad idea. It would have been so much easier to accept the status quo and make that move to setting up file shares and print queues and long for the day I was allowed to configure something with CISCO written on it.

I wonder how many potentially great desktop solutions engineers we lost along the way...