5 Ways Companies Can Cut Cloud Costs
Cutting cloud costs effectively involves examining vendors’ proposals, balancing high-level goals and budget limitations, and finding the best fit.
How much is your cloud costing you?
It’s a fair question for business leaders who have to fit all of their dreams and desires into a concrete budget each year. Cloud can enhance our operations in so many ways — it takes critical data, from payroll to customer histories to product details — out of the "box" of a company’s on-premises servers, and distributes it for on-demand access anywhere in the world. But it also costs money!
As to the question of exactly how much cloud costs, you're hardly going to be able to give a company a blanket estimate, since every company's IT is inherently different and has different pricing. In fact, the issue of cloud costs is often talked about as one of the “great unknowns” in the business world. (To learn more, see 5 Things To Know About Cloud Pricing.)
For instance, take this article from CloudTech. Here, almost half of all survey respondents in a 2015 study said that they have trouble identifying public cloud costs. The study also revealed that almost half use Amazon cloud services for enterprise, up from about 38% the year before, showing just how dominant AWS services have become.
There's also the indication that some of the “cloud rush” is abating. Over 40% of respondents said they did not increase their cloud resources during the year. Experts are looking at that as a possible sign that the massive drive toward public cloud has in some ways come to an end. Still, as a source contends at the end of the article, cloud cost is still important, because cloud is not going away, and because it’s still vital to modern business.
Even though cloud costs money, it's still often massively more efficient than in-house systems.
For example, take this specific small business use case from BetaNews, where a business sets up Office365, Lync and email functionality for ten users. Comparative costs were estimated at around $22,000 for an in-house system, versus $9,000 for "software as a service" or a cloud-provided system. Then the costs for cloud leveled off to about $2,700 per year, while the annual costs for the in-house system weighed in at over $4,000.
Case studies like this can help show companies a little more about their ballpark numbers for cloud use — and as for Amazon, this handy Amazon calculator can help enterprise leaders assess likely costs for S3, EC2 and SQS. Other kinds of cloud calculators are also available online, but again, they're somewhat generic, and not really tailored to an individual company’s bottom line.
So with all of the vagueness around cloud costs, how do you get your company’s costs down? What do you do to really save money in getting the cloud resources you need to effectively run a business?
Private Cloud to Public Cloud
One very common way that companies save money on cloud is to go to a "multitenant" or "public" system where they share resources with other clients.
The benefits and disadvantages of public versus private cloud have been trotted out so many times that most chief executive officers are at least passingly familiar with them. The biggest problem is that since data in a private cloud is isolated or walled off from other client data, it’s generally less vulnerable to security problems. But that said, if you have the right public cloud provider, you may be able to save a bundle of money and still have some decent reassurance that your data is safe in its own separate container, if not in its own server facility. (To learn more about different types of cloud services, see Public, Private and Hybrid Clouds: What's the Difference?)
Decrease Data Use in the Cloud
Another way to save money is to decrease the amount of data that goes to cloud servers, not to mention the amount that gets funneled back into the enterprise office. For example, some experts show how outgoing bandwidth is expensive — in other words, you pay more for data coming out of the cloud servers than you pay for data coming in. Since that's the case, it makes sense to throttle the data that you want to get back out of the system. Put some analysts to work on which data sets will be most valuable, and you may be able to leave the rest in deep storage, and help to contain your costs.
Another idea here is what's called fog computing — rather than sweeping all of their data into the cloud server, companies hold back a whole bunch of data and keep it at the edges of their own networks. You'll see a lot less go through the cloud gateway, and because the meter’s moving a whole lot less, your overall bill is going to be a lot lower. A 2014 piece by Christopher Mims in the Wall Street Journal gets to the heart of this well, as Mims talks about how most companies have “too much data,” and suggests that sending 100% to the cloud is, well, a little indulgent.
Pick Core Vital Services
Another notion about cloud costs is that companies too often go for pricey “buffet” options — instead of migrating services individually, they go out and sign on across the board for payroll, customer relationship management, product development, inventory and even other extras that they don't need. As a result, when they get their cloud bill, they have budget problems. Instead, a lot of experts would recommend picking and choosing specific cloud services that come at a discount or reduced price, or those that will provide the best functionality for business, and leave the bells and whistles where they belong — on the table, where they're not costing extra money.
Improve Cloud Databases – On Their Way Out
Lots of companies also see problems with raw or unstructured useless data, duplicated data or bloated databases that keep running through a cloud system and racking up associated costs. It's become extremely important for many companies to have individuals looking at database quality. Instead of just suctioning up things like partial addresses and irrelevant or unformatted purchase histories, these experts make sure that everything is culled and formatted correctly, and that it's the right caliber of data to be shipped out into the vendor's cloud servers.
Look Carefully at Competitive SLAs
Another key tip is to look at different service level agreements for cloud services. What are the differences in price? And what are the differences in what's being offered? Will a particular vendor try to hold data hostage and jack up prices for extra services? In these ways, purchasing for the cloud is much the same as it is for other complex transactions, such as buying a major item like a car, or buying any kind of professional service. You have to be able to trust the vendor to some extent, and you have to know as much as possible about what you're getting for that price. One issue that often comes up is downtime — because downtime can cost businesses a lot of money even in small doses, experts keep hammering home the point that cloud buyers should really look at uptime and downtime provisions at a detailed level.
Any of these strategies can save a company money, but in the end, you have to look at every vendor’s proposal separately, keep your high-level goals and budget limitations in mind, and learn how to make the best decision that will result in the right fit.