As someone who has the privilege of regularly engaging with IT executives and architects, it’s clear to me that the push to public cloud continues. For your elastic consumption or infinite scale requirements, it just makes good sense. And if I was starting a new company today, I’d likely go all in on public cloud.
The reality for enterprise leaders today, though, is existing data center investments are a mix of legacy and new applications. I’ve also heard that “Cloud First” or “Cloud Only” mandates in some organizations have in some cases caused unintended consequences. Then it occurred to me that I’ve seen this pattern before. It brought me back to my days as an engineering leader, and an analogous push to outsource or offshore R&D. Let me explain.
I’ve been directly involved in a number of these R&D outsourcing initiatives. Some were indeed successful, while others not so much. To make them work, you must properly consider a number of factors. These could include project management challenges when working with teams across the globe, or the relevant skillsets of the outsourced organization. These all factor into time-to-market, and so if takes you six months longer to deliver, the savings may no longer be so black-and-white.
In my early days I recall a meeting with the execs where our outsourced project was one of the three (of ten) that was deemed successful, and my director in fact received an award. I was somewhat aghast, since everyone in my team pretty much considered it a failure. It was months late, over budget, and did not come close to meeting the requirements. I mentioned this to my Director, and he said, “You should have seen the others”—meaning that the projects that were deemed failures were in fact much worse off. From his perspective, since we did eventually make use of the outsourced technology, it was mission accomplished. The following week, our team got a free pizza lunch to celebrate our “success”.
My point here is that making a comprehensive change because of management dogma, and then simplistically expecting it to save costs, is not a recipe for success. It needs to be a much more nuanced decision, factoring in all of the relevant considerations. The reality is that there is likely a proper mix of insourced and outsourced R&D for your organization that optimizes all of these factors.
I’ve seen a similar pattern emerging with public cloud consumption.
Some organizations adopt a “Cloud First” policy, based on the belief that it will save them money. Then they get their first monthly bill, and realize the planned savings are just not there. I continue to hear this story regularly from the IT folks that I talk to. Now there are certainly ways to mitigate public cloud costs, but that’s a blog for another day.
Very often the choice of moving to public cloud isn’t about cost, but instead about business agility. IT is not delivering services fast enough to their internal customers, and so those customers bypass IT and go directly to public cloud. These pockets of shadow IT throughout the organization have a number of unintended consequences. Nobody now knows the true IT spend, since it’s split up and hidden across the company. Moreover, IT standards for patching, backups, and security are no longer consistent organizationally, which can cause a range of problems. Often these pockets of shadow IT eventually end up being consolidated by IT, who now have to manage it.
Again here, my point is not to discredit public cloud consumption as an IT strategy, but to properly look at all of the factors before jumping in.
Deciding whether capex or opex savings are more important to you can be a key consideration in choosing private vs. public cloud. If your problem is that IT can’t deliver services fast enough, then consider adopting private cloud automation in order to make optimal use of your existing on-premises infrastructure. If your developers are using the latest and greatest AWS services, or moving to serverless architectures, then those applications will be public cloud native. IT will have to enable and support them in order to make the business successful.
Many effective organizations have in fact adopted a cloud brokering mentality, choosing a hybrid cloud approach. Again, it’s about the proper mix that optimizes all of the relevant factors for your organization.
Whatever private and public cloud combination you choose, it’s critically important to measure their successes honestly, considering all relevant criteria. And when you achieve those successes, your reward should be more than just pizza.