This article originally appeared in betanews*
As more and more applications move to the cloud there's an increasing assumption that it's the right approach for everyone. Indeed we've seen recently that many cloud investments are being made out of fear of missing out. But this one size fits all approach isn't necessarily the right one, and can put companies in a difficult position.
We spoke to Michael Hiskey, Chief Strategy Officer from intelligent data management company Semarchy, to find out what businesses should consider before moving to the cloud and how they can avoid being forced down that route.
BN: Are companies adopting the cloud for the wrong reasons?
MH: Your FOMO article reminded me of many of the conversations I have with clients in both the UK and the US. There is so much momentum for cloud-based software solutions, whether they be Platform-as-a-Service (PaaS), Infrastructure-as-a-Service (IaaS), or the original Software-as-a-Service (SaaS) solutions. This causes some industry decision makers to be blinded by the short-term benefits of trading Capital Expense for Operating Expense, even if the Return on Investment (ROI) is unclear.
Don't get me wrong, many -- perhaps most -- solutions can live in the cloud. At Semarchy, we have the #1 solution on Amazon Web Services for our category, Master Data Management (MDM). We also run on Microsoft Azure and Google Cloud -- and see those platforms growing strongly. I’m not saying the cloud isn't the way to go for new solutions.
Clients I speak with about cloud solutions don't think of the Cloud the way most vendors do. Some are subject to a 'cloud first' strategy mandated from executives, or talk about their organizations 'multi-year journey to move to the Cloud.' Still others refer to it as 'just another data center.' It’s this last category that is thinking most clearly about it. Put solutions in the cloud when they make sense.
BN: Are suppliers actually forcing enterprises to adopt the cloud against their wishes?
MH: It's not exactly forcing. Publicly traded enterprise software companies are under pressure from Wall Street to look like they are a 'cloud company.' They're looking to report as much revenue from the cloud as they possibly can. For this reason, they've gone through machinations to make license revenue look like cloud revenue. This allows them to talk about the cloud trend with financial analysts.
At the same time, the FOMO study shows that many executives feel that it's good for their career, resume, or experience to preside over cloud projects -- but the savvy ones will look at the bigger picture. In some cases, over the useful life of a solution, you could be spending three or four times what the total solution would have cost you, if you’re not careful.
BN: How do businesses gain the confidence to reject cloud FOMO in favour of practices that are best for them?
MH: Technology teams at businesses need to focus on driving business value, not on how the technology is implemented. All too often IT folks get caught up in 'philosophical' debates on these types of decisions.
We think about the business value a client is trying to get out of a solution -- for example 'better customer service,' or 'increase revenue per customer,' etc. That's generally about mastering customer data, data quality, and data enrichment. For the end-user, where that solution sits is completely transparent. They access it via their laptop or mobile device or whatever -- it could be sitting in Amazon's Data Center in Ohio or on Mars, for all they know (or care).
BN: What do companies need to take into account when considering a cloud move?
When debating 'to cloud or not to cloud,' always look back at the original business requirements -- let them be your guide. What data is delivered at what level of service to whom, for example, will guide the path. Often, data that starts in the cloud (think eCommerce) lives in the cloud. Building up piles of servers, in expensive data centers, with power and cooling, adding operating systems, and the like, is certainly the old way of doing it -- but in some cases, it may deliver a stronger return on investment.
BN: Is there a risk of companies losing overall control of systems or data? Is skepticism about data retrieval and recovery from the cloud justified or unfounded?
MH: Largely unfounded. There are a handful of cases you may have seen where there is a 'catastrophic' data loss. These cases are exceedingly rare. The bigger question on cloud back-up is understanding what you’re trying to deliver -- is it speed? Reliability? Peace of mind? How is it that data stored on-premise is safer in the cloud? If speed and accessibility are your key requirements, the cloud is the right choice -- if it's just to be absolutely sure -- I suppose that depends on how much you trust your internal data center staff!
BN: What technology options are there for avoiding a forced move to the cloud?
MH: In those cases where the cloud isn't the right option, the useful life of the solution may already be long term with much of the relevant data living on-premises, tied up in legacy systems. The key here is to watch out for vendors that are creating a forced march to the cloud. In general, these businesses are never acting philanthropically -- they may be railroading you into a solution with limited features, fewer options, or capacity costs, that will become problematic down the road.
BN: Are their any corollaries to this situation with other tech trends?
MH: AI/Machine Learning, in the sense that because it's a hot topic right now (like big data five years ago), customers and clients are asking about it. Companies can succumb to the pressure to say they have AI/ML capabilities, even if they don't, or exaggerate the capabilities of products which are still in the early stages of development. This creates a false narrative/concept bubble that can misrepresent products and mislead investors/shareholders.
BN: How do you see attitudes towards cloud use shifting over the next few years?
MH: More businesses will continue to have a 'cloud first' mentality. Legacy vendors who have 'cloud-washed' solutions (those that are on-premises designed software with a weak veneer for the cloud) will not thrive in this environment. At the same time, solutions from every vendor need to work just as well in the cloud as they do in private data centers, and be completely transparent to the business end-users. At the same time, hybrid cloud environments and those that can 'burst' into the cloud (predominantly on-premises with back-up, additional capacity and processing power in the cloud) will become a more normative fashion.