Standardized data the key to making cloud storage a commodity

Standardization is the key

Commodities rely on standardization. There isn’t a standard API for cloud storage yet, though Amazon’s S3 API is clearly at the head of the pack. Wouldn’t it be nice if you could move your storage between cloud storage providers like Amazon to Google or Microsoft without changing one line of application code? Someday, I believe, this will be the case. API standardization is key to driving down price because it reduces or eliminates vendor lock-in. Once storage is truly portable, vendors will not be able to get away with locking you into their proprietary storage.

The importance of price

Price is probably the biggest determinant of success in a commodity market. Take a look at gasoline, for example. It’s a classic commodity product: gas is gas. All regular gas is 87 octane. Any car can fill up at any service station.

Similarly, when it comes to data storage, it will just come down to price. If all cloud storage was fast, super-reliable, and secure, then the only thing that would matter is price. With commodities, standardization drives down price.

Two factors are at play. First, as cloud storage prices drop, the migration from on-prem storage to the cloud becomes more compelling. With cloud storage prices from Amazon, Google, and Microsoft in the $.02-.03 per GB per month range, there is no significant savings versus the total cost of ownership of on-prem storage. However, cut the cost of cloud storage to a conceivable $.0039/GB/Mo and savings from migration to cloud storage can be 75% or more.

The second factor concerning price is that as storage costs drop, many new things become economically feasible that were not previously. For example, suppose you want to create the next Instagram or Pinterest – free apps that contain a lot of photos or videos. Your biggest cost will be storage. Your revenue will likely come from advertising. Your business model may never make money if you have to pay $.02/GB/month for storage, but it might be highly profitable if storage were at one-fifth that price.

You can’t fight economies of scale, right? Not necessarily.

Common wisdom says that a startup should not be able to store data as cheaply as an industry giant like Amazon. Yet history is replete with examples of innovation overcoming scale.

For nearly 100 years, US Steel was the largest steel company in the world. Then in 1968, a little company called Nucor invented the highly efficient steel mini-mill which used the new technology of electric arc furnaces. Today, Nucor is the largest steel company in the U.S., and in May 2014, US Steel was removed from the S&P 500 index due to its declining market capitalization.

There are many other examples: AT&T vs. MCI. IBM vs. EMC, and so on.

How will we think about cloud storage 10 years from now?

My bet is that cloud storage will become just another piece of the infrastructure, like bandwidth or electricity. Almost every application needs storage. Nobody should be locked into proprietary vendor solutions. You should be able to unplug any cloud storage vendor and plug in some other vendor if you want to do that - with no price hit or service lock in. This is my dream: cloud storage that is so cheap, so fast, and so reliable that it works for almost any storage need.