Is Google App Engine a Sign that Google’s Jumped the Shark?

Cyndy Aleo-Carreira,


Google App Engine logo imageLast night, when the news of Google App Engine hit the web, I was hobbled with DNS issues that left me able to watch the Twitter stream but unable to read any of the details. Waking up this morning and finding all the Twitterati still throwing confetti, I'm starting to feel that I'm a lone dissenter when I say that I don't think Google App Engine is going to be good for the Web.

Just yesterday, Hank Williams followed up his “VCs are killing us with free” piece with another that came to the same conclusion I did after writing my original response: Google is to blame when it comes to the “must be free” mantra as well. For years now, they've handed us free applications on a silver platter: free web-based mail and calendaring, as well as a free office suite and free web-based email for businesses, and a host of other services. We've grown accustomed to this business model from them, with free services supported by the behemoth of Google's ad revenue.

I've joked before that any time Google sees something in the Web space that is enjoying any sort of success, they will build it themselves. No company is safe, and this morning, Amazon is realizing that as well. The Google App Engine is such a close copycat of Amazon's Web Services that there can be no doubt that Google's gone gunning for Amazon with their new offering. And unlike Amazon's offerings, Google App Engine is free up to 500 MB of persistent storage and 10 GB of bandwidth per day. I'm sure that Google feels Amazon should be shaking in their boots today.

There are two problems with their launch, in my opinion. The first is the choice of Python for launch. In the Web 2.0 space, Ruby (and specifically its Rails framework) is the “It” language. When you look at the applications currently using S3 (Twitter being one of the biggest), they aren't using Python, and I doubt they plan on recoding just to move to GAE. Google promises other languages will be added, but they would have made a bigger splash and lured more of Amazon's customers away, had they started with Ruby.

But beyond the choice of language for launch is Google's repeated reliance on the free offering to get customers of other services to switch. A pattern has emerged with Google: copy another company's idea, make it bigger if we can, and give it away. Of course this increases the eyeballs on Google properties, which in theory should lead to bigger ad revenue, but the numbers on ad revenue are already slowing down for Google. The U.S. economy is suffering, and ad revenue will surely follow. At what point will Google realize you can't give everything away for free and rely on the ad business to fund it?

Even beyond the ad-supported business model for entry-level apps and the language choice is that Google has ceased to be innovative. Sure, the majority of bloggers this morning are singing Google's praises, but when is the last time that the company who revolutionized search and came up with the model for monetizing content with ads introduced a new idea? Every single offering that Google has brought us over the last few years has been in response to another company introducing a new and innovative service. Google's entire development model at this point appears to be a “me too” cycle, dependent on their size and the weight of their influence to take over in the space they enter.

So while the rest of the tech blogosphere can continue to toss confetti this morning, I'll be over here pondering whether or not Google has reached the point where they are steering their huge ship at the latest Next Big Thing, never realizing that they are slowly running out of gas to keep heading to the next destination.


If you enjoyed this post, make sure you subscribe to profy RSS feed!
7 Comments (Subscribe to rss)
  • You are dumb

  • This offering is missing some of the main advantages of AWS and has some serious drawbacks:

    1st: There’s an API
    You have to develop your app against the Google API to deploy it

    2nd: It’s not just a VM
    You can’t deploy multi-app solutions, integrated solutions, or anything off the shelf on it

    You can’t take your app that you’ve been building for 5 years and has an installed user base, and sell it as a VM image to run on this service

    Who’s going to want to bend over and be Google’s bitch just to run on their servers? You have to rebuild any app to use this service, and it can only be a web app. If you have a problem with Google’s hosting, billing, or other policies, you’re screwed. You can’t pick up your toys and go home, because they own your home and let you sleep in it… for free… for now…

    With AWS you’ve got a Linux install… You can install as many or as few apps on an image as you like, configure them to work together, and, if Amazon pizzes you off, you convert it from an AMI to a VMWare image or just clone it onto a Linux install and you can switch hosting providers in no time.

    With Google? Well… bend over and try to smile… WAAAY too invasive for my tastes, but some of you Web 2.0 people seem to like to be abused…

  • Not to mention Python… Python? Really? Are you kidding me? Why not PHP while we’re being crazy? Jeebus…

  • Your article is food for thought. I have always thought of Google as the “good guys” who break down walls of inefficiency, but you have a point. Not only Google is aiming at Amazon, but think of all the hosting companies that will be impacted by Google App Engine. Maybe the lack of PHP or Ruby support is for the best

  • Alexis, I honestly believe that they stopped being the good guys a long time ago. They got too large, and anything that large ends up working like any other business: cut-throat.

  • I think you have missed the big point that differentiates GAE from S3 - serious scalability. S3 gives you a little linux box on the web, GAE gives you a massively scalable application. So if your S3 app hits the big time, you’ve got to figure out how to replicate databases, load balance and share sessions across several S3 machines. On GAE, you just watch it soak up the traffic, knowing it is built on the same architecture that gives millions of search results a second from the biggest database on earth.

    I also find your preference for Ruby over Python ill-founded. Python is a lot more mature, and I guess it will last a lot longer than Ruby - already some Ruby fanatics are moving on to Lua or whatever the next great language is. And I also expect that Google’s bet on Python for App Engine will be self-fullfilling - it’s suddenly going to get a lot more exposure.

  • @FoxyLad Or, you know, you planned ahead and designed your architecture to do that in the first place. GigaSpaces has a platform to do just that, as does Appistry, and a smart developer would keep that in mind from the very start instead of worrying about it once the need arose.

    I never in a million years said I preferred Ruby to Python. I’m really keeping an eye on Fire Eagle because so far, I haven’t seen any Ruby-coded project scale well. However, based on the number of Ruby apps out there, Google’s Python push isn’t likely to win a whole bunch of new converts who are willing to switch just to use GAE.

Leave a comment (We support avatars from Gravatar, MyBlogLog, and FriendFeed)