During a recent CNBC interview, Mark Cuban, owner of the Dallas Mavericks, said:
When I bought the Mavericks, everyone in the organization thought we were in the business of basketball. But we aren't. We're in the business of "Honey, what do you want to do tonight?"
I read about the quote on Yaron Galai's blog. Yaron is a co-founder of Quigo, an ad network. Over the past 6 years, he's seen a few million examples of how sellers sell, and how buyer react. To the Mark Cuban quote he adds that:
It's so easy to describe what business *you* think you're in. The trick is to define the business *your customers* see you in. It's not about what your product does, or which cool features it has. It's all about understanding which piece of attention of your customers you're fighting for, and who are the others trying to get that same slice of attention.
I thought of Yaron's post when I saw MH-One's recent ad in Wired Magazine (pages 130-131 in the November 2006 issue)(MH-One is an utility computing service from Data Return). Its tag line was "Say Goodbye to Dedicated Hosting".
Rackspace's Mosso, too, devotes substantial website real estate to comparing "The System" against dedicated servers.
The assumption in both cases is that providers of grid hosting are fighting for prospects who are interested in dedicated servers.
In contrast, let's take a look at how Amazon describes its Elastic Compute Cloud: it's affordable, scalable, reliable and secure. Which is why developers are lining up for the beta.
So my question is, are Data Return's and Rackspace's customers really so keen on ordering un-dedicated-servers? Or are they just looking for an affordable, scalable, reliable and secure hosting platform? Is the goal to trumpet how far our technologies have come? Or to talk about how close we are to giving customers what they want?
How many web hosting providers today fail to mention that they are affordable, scalable, reliable and secure hosting? Virtually none.
Is the allure of these benefits really the reason Amazon ECC has so much buzz? Unlikely.
People are lining up for the Amazon ECC for other reasons:
1. Amazon ECC is highly differentiated from the hosting space by design, which allows them to get cute with their messaging. ECC is built on some sort of mythical infrastructure that scales without effort and never goes down. No one really knows exactly how scalable or fast it is, but it really doesn't matter. Buyers already have that perception and Amazon isn't about to change that by publishing technical details.
2. The pricing model is unlike anything we've seen in the hosting space. More important, it appeals to startups to rapidly prototype proofs of concepts to demonstrate that their business plans hold water without deploying any capital. The amount of time these startups have to prophetize the virtues of Amazon ECC is inversely proportional to the size of their customer base. All of that adds nicely to buzz creation.
Elastic Computing Cloud is a fascinating service and I think Amazon has a winner on their hands, but I seriously doubt that adopting "affordable, scalable, reliable and secure" messaging is a game changer for hosting companies.
In other words, I don't think EC2's allure is 100% attributable to technology features. The folks at Amazon really gave serious thought to what web app developers are looking for. In contrast, many web hosting execs (including myself, when I worked at EV1Servers) have no clue who their target audience is. "Small business owners", is what most of them say, or "IT professionals". They assume that because dedicated servers is what they've sold, clearly dedicated servers is the right reference point. But is the average "small business owner" or "IT professional" really so attached to the idea of a dedicated server??
"We are in the real estate business. The only reason we sell hamburgers is because they are the greatest producer of revenue from which our tenants can pay us rent."
Questions then for the web hosting cognoscenti: who is -- or will be -- the McDonald's of web hosting? What web product is -- or will be -- the revenue-fat virtual hamburger?
It does drive me crazy that ECC is so nebulous. What does scalable really mean? If an entire rack of servers loses power, will my service still stay up? What if the entire data center goes out?
Maybe I am just not their target customer.