Bits Also Want To Be Free

By now you’ve probably checked out YouTube and Google Video, where you can search and view thousands of user-generate (mostly silly) videos. But have you considered how these sites let their users upload huge video files for free and stay in business? It’s not clear the online video services do qualify as businesses yet, but they point to a very interesting trend: storage is becoming really, really cheap. This is enabling new innovation and business models.

When Google announced Gmail, email with one gigabyte of free storage, two years ago, it was a big wakeup call to the competition. It signaled Google had figured out how to make storage so cheap it could give it away. That gave them a very real competitive advantage. To neutralize that competitive advantage, others set out to build their own low-cost storage infrastructure. Recently, a major development happened on the (almost) free storage front.

A few weeks ago, Amazon.com made an announcement I think will both greatly benefit the consumer and foster innovation. I’m referring to its new Web service, S3, Simple Storage Services. (Google is rumored to be working on something similar, known as Google Drive.) Amazon’s announcement is not only interesting, it’s going to change the field of storage-intensive computing and services in profound ways.

Consider this: a few weeks ago, I was having a conversation with someone in a research department at one of the largest European telecom operators. We got to discussing a recent market study showing how younger people consider losing their mobile phones as traumatic as losing a family member. Why? They keep so much personal history on their phones, notably text messages and photos. Remember your first love? Well, it seems teenagers keep the history of their first romantic relationships on their mobile phones. Loosing that history is more than a little traumatic.

So we began talking about how interesting it would be to have an Internet service that would automatically make collections of a mobile user’s SMS (define) messages, photos taken with the mobile phone, even their voice mail. The service would provide backup as well as anywhere access to text messages, photos, voice mail, even video.

“Those are really fun ideas,” said the researcher, “but we could never afford to store all that data.”

“Store all that data?” I was unsure what he was referring to. He tells me their users send hundreds of millions of messages every day and that there was no way his company would be able to keep all that data around in its internal storage servers. And that was just text. Forget about voice, photos, and video.

I realized we live in parallel universes. SMS text messages are really short. My guess is the average message is far fewer than 100 characters. But let’s be really conservative and say 100 characters per message. Let’s then be really aggressive and say 500 million messages are sent by the telco’s 100 million or so mobile customers every day. These messages would take up 50 billion bytes (characters) of data, requiring 50 gigabytes of storage per day.

In a month, the telco would require 1.5 terabytes of storage to save every single text message sent by the users. Sound expensive? Well, if you’re a big telco with a big legacy infrastructure, yes. But if you use Amazon’s S3 service, this storage would cost less than $250 per month! That, of course, doesn’t include bandwidth (i.e., network traffic) costs, so let me quadruple the number, and we’re now up $1,000 per month. If online collections (and sharing) of all your SMS messages was offered to mobile users at a cost of $1 per month, this would be a 1,000 percent gross margin offering!

Simplistic? Sure. But you get the idea. Storage is becoming a commodity, and that’s rapidly changing the rules of the game.

Much initial discussion about S3 has focused on useful yet unimaginative applications, such as hard drive backups and the like. But creating a cheap backup service for your laptop is hardly what’s interesting about what Amazon’s doing. The enormous capital infrastructure investments companies need to make if they want to build services that require mass storage are no longer necessary. If I can store billions of bits for (almost) free, that opens up untold opportunities for innovation. Small companies and large telcos alike can offer value-added applications and services on top of a set of competitive, commodity services.

Information wants to be free, or so goes a popular saying generally credited to Stewart Brand. Free information poses an economic challenge when the underlying infrastructure cost is high. But if the storage of bits is free, it becomes much easier to make information free, too. The impact of storage joining the ranks of commodity services is going to be profound with impacts on all aspects of publishing and information sharing. Stay tuned. The Internet is not going to be your father’s TV.