It’s pretty significant when the Wall Street Journal talks about cloud issues, and this story (behind a paywall) is no different. The gist is that as enterprises support a mostly remote workforce with cloud computing, they are, of course, seeing rapid growth in the monthly public cloud bills.
Although a 20 percent expansion in dollars burned each month is average, I’ve seen expenses go as much as 50 percent higher in month-to-month growth. This is without expanding the number of applications or data—just how the clouds are now being used.
It’s expected to pay more for expanded use of cloud-based resources, but enterprises are making mistakes that lead to higher costs than necessary, such as:
- Not planning in advance for cloud use, which would gain the use of reserved instances at a much-discounted price, as well as other benefits. While many enterprises are responding rapidly to the changing needs of IT due to the pandemic, others are still not cost planning, which could significantly optimize their cloud spending.
- Lacking cloud cost governance systems. These tools not only monitor existing usage costs which are tracked down to the applications, databases, and users, but assist in demand planning as well. They are invaluable.
- Tossing money at the problems. In short, we’re forced into stress testing cloud-based systems perhaps prematurely. Pushing applications quickly into production means that the number of issues needing fixes ramps up 500 percent in the short term. Enterprises are forced to hire some very pricey people to fix the issues, using whack-a-mole tactics which are hugely expensive.
The key phrase here is cloud cost optimization, meaning the ability to get the most out of public clouds for the least amount of money. Not surprisingly the pandemic panic has caused enterprise IT to act reactively not proactively. This leads to cost considerations and planning being tossed out the window.
Now that we’re months in, it’s time to give this cloud cost stuff a bit more thought. Some have cut their monthly bill in half and still use the same amount of cloud resources. I’ve seen it done.