kid on a lawn in a Charlie Brown ghost costume (white sheet with a lot of holes inked on in black ink)
PHOTO: Steven Saus

Ghosts. Goblins. Horror movies. Massive piles of sugar at every turn (that you have no choice but to sample). There are plenty of good reasons to be scared this time of year, but the existence of technology silos isn’t one of them.

Now, you might be thinking, “Wait a minute, haven’t folks like you been telling us for years that silos are bad? Now you’re changing your mind and telling us not to worry about them?”

Well, yes and (in certain instances) yes.

People like me have indeed long perpetuated the notion that technology silos are a detriment to your business, an evil worth rooting out and breaking down. It was sound advice at the time we started dispensing it, but times change and so do best practices. After all, one of the things that makes the conjoined worlds of business and technology so interesting is their dynamic nature. Nothing is set in stone. Nothing is forever. What was good for your business yesterday might be detrimental today, and vice versa. Such is the case with silos.

With that in mind, let’s take a closer look at three specific instances in which technology silos might now actually be good for your business.

Legacy Platform Support

For years now, tech vendors have encouraged (and in some cases, attempted to outright force) users to move off legacy platforms and onto their platform du jour. Use the newest browser. Install the newest operating system. Upgrade to the newest device. As such, it seemingly made sense for organizations to stamp out teams still working to test and support these legacy platforms and rally all resources around the latest and greatest.

But then a funny thing happened. Users didn’t migrate, at least, not en masse. Many of the end-users for whom most organizations develop and deliver products and services still rely on legacy web browsers, operating systems and mobile devices. And if they’re still using those older browsers, operating systems and devices, then you still need to support them by developing, testing and delivering apps that work flawlessly across them. Remember, your customers aren’t about to let you off the hook for a poor user experience just because they happen to be using an older browser or mobile device. Legacy platform or not, the user experience is still your responsibility.

So, if there’s a small (dare I say, siloed?) group in your organization still quietly working to support these platforms, well, that’s a good thing. Sometimes being progressive means looking backward as well as forward. Users will eventually move on to the next thing, but until they do, it’s imperative you continue supporting them. Silos can help. 

Related Article: What SharePoint 2010's End-of-Life Means for Your Organization

Data Analysis

As an industry, we’ve spent a great deal of time talking about the need to break down analytics silos. But, to quote Brooks Hatlen, the memorable character from the now-classic movie (and Steven King novel) "The Shawshank Redemption," in the time since we started doing so, “the world went and got itself in a big damn hurry.” In the age of digital transformation, no one has time to wait, especially decision-makers seeking answers to small-but-significant questions. And in those instances, easily accessible data silos can be quite valuable.

Let’s take the example of a marketing leader who, in trying to decide whether to pay for a similar initiative, wants to know how a recent webinar performed in terms of generating pipeline. That’s a simple question. Unless, of course, you’ve managed to break down all of your data silos, in which case, getting the answer will be anything but simple. The analyst you’re working with will likely need to access the data lake or global data warehouse, query the master transactional system, and connect to, among other steps. She’ll need the right permissions and the right licenses. In other words, she’ll need time. All for a simple question that requires only a small subset of data.

Now, if you’re that marketing leader, and the one thing you don’t have is time, you might be inclined to say just say forget it and make a decision based on factors other than data and information. That’s far from ideal.

Certainly, the ability to bring together and analyze large subsets of data to answer big, macro-level business questions is critical. But so is having real-time access to smaller (dare I say, siloed?) subsets of data necessary to answer simple, day-to-day operational questions. As you transform for the digital era, you need to have both.

Related Article: Old MacDonald Met a Data Silo, E-I-E-I-O

Open-Source Exploration

When tech pundits first began advocating the breakdown of silos, open-source wasn’t nearly what it is today, where it's emerged as the driving force behind so much of the innovation emanating from the tech world. The idea behind breaking down silos is to standardize and drive cohesion. But open-source is about flexibility. It’s about giving your best and brightest the freedom to learn and grow. You can’t do that while trying to force standardization. Standardization around open-source only works if it happens organically. Until then, consider yourself lucky to have many different (dare I say, siloed?) groups across your organization engaging with a variety of open-source technologies.

Related Article: Do We Really Need to Get Rid of Data Silos?

Dare I Say, A Silo Can Be a Good Thing?

Just as we might have gone too far in declaring silos an absolute evil, we shouldn’t swing all the way back and advocate for the mass proliferation of silos. There are undoubtedly many instances (such as DevOps and CI/CD, for example) when breaking down silos makes tremendous sense. But many instances doesn’t mean all instances. 

In the right circumstances, silos can help you ensure broad platform support, increase your data agility, and encourage open-source exploration. And there’s nothing scary about any of that.