At these volumes, spam isn't merely an annoyance -- it constitutes a legitimate reduction in available services to an organization, whether that be reduced bandwidth due to inbound spam, increased costs due to additional servers or services required to contend with the deluge, or simply the time lost when legitimate emails wind up buried in a junk mail box or lost forever.
The unfortunate reality is that methods to reduce or eliminate spam have been around for a while, such as whitelisting or ISPs charging a small cost per email, but they're so Draconian they would all but destroy the concept of email. We don't need to throw the baby out with the bathwater, but we can't keep putting our fingers in the dyke and shaking our heads sadly.
A bazillion different "solutions" to this problem revolve around email filtering. For example, greylisting institutes a delay period for unknown senders and causes spam blasts to miss their targets. Meanwhile, good ol' whitelisting and blacklisting add plenty of manual effort and can cause problems with reliable email delivery. But none of those solutions do anything about the vast herds of spam flitting around the Internet, chewing up bandwidth and computing resources the world over. If they work at all, they merely prevent spam from hitting our inboxes, which is a Band-Aid, not a fix.
IT fix No. 7: Virtualized application appliancesInstalling a new and expensive line-of-business server application shouldn't require two weeks of training. It should be delivered ready to go, with all the requisite dependencies, patches, and other detritus that commonly accompanies these massive collections of code.
Rather than be saddled with an install DVD and a future filled with hours of watching progress bars scrape their way across a screen, we need a virtual machine that can be imported and fired up immediately. In many cases, those dreary installation procedures are occurring on VMs anyway, so let's skip the middleman. Instead, designate the virtual machine as the default application delivery mechanism rather than a Windows installer or a tarball.
The time, effort, complexity, and support costs that could be saved by more companies taking this approach is significant. That's not to say that there shouldn't be a way to also do a standard installation, but as a default, go with the VM.
IT fix No. 8: IPv6It's not lost on me that if I'd written this article five years ago, this would definitely have made the list, yet we're not any closer to widespread IPv6 adoption.