Ok, I’m going to go into rant mode. Why is everyone so enamored with Google these days? I read yesterday about their recent Google App Engine, and didn’t really see anything that particularly turned my crank.
I’ve been bouncing around the web today reading what people are saying, and for the most part, everyone is down in the People’s Temple drinking the Kool-Aid. I don’t think the App Engine is a bad idea, I just don’t really get why you’d even consider it, especially considering AWS is around and has a much more flexible offering.
First, I’m going to start by saying that 99% of web companies never even reach a point in their business where they actually require scaling. I’ve seen this talked about many times. So why, when developing a new web technology or company would you lock yourself into the Google shackles simply on the hope that you hit that 1%? It would be different if they offered a solution that you could migrate at some point, but what they do offer is fairly Google centric.
Also, in terms of the valuation of a company, most of the large acquisitions in the last five years have primarily been in exchange for the company’s user base. That is, if the cost of acquiring a customer is $100 by conventional marketing means, but you can buy a few million of them (by purchasing a company) for around $30/user, that’s a good deal.
But if you use Google (at least based on some sites I’ve been reading), they you are tying your users into the Google infrastructure. Maybe that’s a good thing in your mind, but if you think your user base might be worth something, then I wouldn’t want that intermingling with Google’s world. In fact, I’d actually question if it is your user, or theirs (i.e. read the fine print people). Google has clearly shown that all your data is their data, even inside of your personal emails (ever see the contextually sensitive ads in gmail?) Clearly Google wants to take data mining to another level with this service.
In terms of actually having a legitimate web company (you know that like, makes money and stuff, and doesn’t just talk about that), the majority of your costs these days will most likely be salaries, not hosting (the obvious exception might be a company like YouPorn, but they are the exception, not the rule). Our company at work hosts tons of data files for our customers, and we get hundreds of thousands of downloads a month. You know what that costs? A few hundred dollars in bandwidth per month.
You know, I’m not even sure why so many companies are trying to offer cheap hosting. Who wants cheap? I want good. If I were to start a web hosting company, I would do the opposite — I would offer expensive hosting, and then get behind my product. It would be the hosting equivalent of those gas stations in the 50s where guys would run around and pump up your tires for you. Maybe I’d make those guys really hot girls instead, and have them in cute cheerleader uniforms, but it would be the same concept. What company wouldn’t pay $100 a month to have hosting that was reliable, run by people with a vision and who cared about the success of your product or website? People only go cheap today because in a sea of shitty options, you might as well have one that doesn’t break the bank.
If I had to choose between Amazon’s AWS and Google, I wouldn’t even hesitate to choose AWS. Google is far too restrictive at this point, and I really think it’s a bad idea for a new company.
Let’s talk about Google in general for a moment. They are the clear and obvious leaders in search, hands down. But what about all the other things they’ve worked on? Truthfully, they haven’t really done much that’s exciting. Gmail was a home-run, and most people I know use that. GoogleTalk, not so successful (in fact, compared to Yahoo messenger, the numbers are fairly embarrassing). Part of the reason I think is the whole “perpetual beta” concept that these large web companies seem to be stuck in. I don’t mind a product in beta, but you know, the beta should end at some point — I think two years is enough time to work out the bugs.
Also, as most of you know, Google has a flat structure internally without a ton of management (at least that’s what they claim). I also know that almost every cool project that has come out of Google (GoogleTalk and Gmail I believe) was part of Google’s 20% time projects (every employee is allowed to spend 20% of their time doing the things they want). So I mean, if you’re management at Google, and you see that your best stuff is coming out of the time your employees are spending doing their own things, shouldn’t you, like you know, make the 20% time the 80% time? Maybe tie up some loose ends and release a few solid projects? That’s what I would do.
How about OpenSocial or XMPP? I had a conversation with some friends last year about OpenSocial. I thought it was another interesting idea, but didn’t think it would go anywhere. A year later, I’m still waiting to hear about some huge successes with it but nobody talks about it anymore, at least not in the circle’s I’m in.
XMPP is great technology. Really it is. But where are the wins in that space? GoogleTalk is based on XMPP, but like most things Google releases, it came out and then sort of disappeared. If Google would throw their weight behind XMPP and GoogleTalk, maybe it would turn into something truly revolutionary for the web. But that might require pulling a product out of Beta.
Open standards are great, and that’s the direction everyone needs to move towards. But you can’t have an open standard without the support of the industry — without it, it’s ultimately bound to fail. I’ve seen it time and time again at the IETF and other open-source meetings. The community needs to collectively admit there’s a problem, come up with an open standards approach to fixing the problem, then as a whole, get behind it.