Struggling with legacy system modernization? In a previous post, I talked about how companies can use business...
process management as a bridge between systems. Today, I'm here to suggest the same sort of use of the cloud.
But first, I want to make two important points. First off, legacy isn't always bad. Second, cloud isn't always the best option.
Point one: Legacy doesn't equal bad
When it comes to legacy system modernization, many people tell me they have to rip out the old and replace it with the new. While this approach sometimes is necessary, it's not a given -- no matter how earnestly a vendor makes the case. Many systems, in fact, chug along just fine for years, and because they're -- usually -- long since paid for, they're nothing but value machines.
Furthermore, there's nothing wrong with managing the information these systems contain in place rather than migrating it to another platform. This is another common point of debate, and I find it as often to be an argument of philosophy as one of technology. The trick is to ensure your legacy products can actually connect to the rest of your infrastructure -- something I'll come back to soon.
Point two: The cloud doesn't equal salvation
Although no longer new to the computing scene, many of my clients and students still speak of the cloud in reverent tones. While it's true that cloud-based environments often prove to be enormously effective, even the best of them did not experience an automatic curing of all ills; nay, these outcomes were the result of the intelligent application of carefully considered solutions to a particular set of problems.
As often as not, cloud-related successes involve the rehosting of existing applications to make them more straightforward to manage and maintain and more accessible to more people regardless of location. But there's more to the cloud story than just these aspects, and one of the most intriguing is to use the cloud for legacy system modernization.
Find the middle(ware) ground
Think of a cloud-based technology as a piece of macro-middleware. Never mind that conventional use calls for middleware to actually process the information it receives -- let's instead consider it as a piece of connective tissue that simply passes information from a legacy system on one side to an otherwise incompatible application on the other.
The analogy is of an English-speaking person needing to communicate with a French speaker, and doing so by talking through someone who speaks both languages. It may not be ideal, but it can succeed enough to do the job.
The first time I saw this model work well, the system in the middle was SharePoint, which had captured the market's imagination to such a degree that nearly every application worth its salt -- legacy or otherwise -- was being equipped with a SharePoint connector. The remarkable thing at the time was that, in this installation, SharePoint wasn't doing anything except acting as the go-between, rather than being used as a shared repository for the applications on either side. The organization in question already had the necessary licenses in hand, so it was a real win-win.
ERP serves as a starting point
Today, I'm encouraging you to think about using the cloud as the intermediary, for it now has the same high visibility and technical support that SharePoint did then. Better still, the cloud runs on an open standard -- or two, or 12 -- whereas SharePoint is a Microsoft product. So the cloud can be painted as being philosophically "pure" as well.
One place this approach can pay great dividends is the realm of ERP. Vital to companies of any meaningful size, ERP systems tend to be among the first true enterprise applications to be installed, and their cost and complexity means they often hang around for a long time. In 2014, for example, a study by Aberdeen Group found that "the average age of an ERP solution in mid-market organizations is over seven years." A lot can and does change in that span, and it's not uncommon for an executive to look around and say, "This old thing just isn't doing it for us anymore."
I've seen this situation more than a few times among my clients, and it's always fun to show them how the cloud might be used for legacy system modernization back from the technology margins. The effort and economics of the requisite master data and interoperability work don't always lead them down this path, but there's huge value in exploring its potential as a modernization technique and a way to possibly reverse the diminishing returns.