Friday, September 02, 2005
Canonical example: Writely.
At a recent presentation to one of the Big Net Players, I got a lot of pushback about the notion of coordination.
It's a pretty simple (and well-trodden) concept in econ: people getting together to do things, by processing information. This is exactly what Writely is.
Here's the thesis I tried to outline to them: presumably, because doing things is costly in the real world, you can create value to the bound of coordination costs saved. Looking at Writely, we can make this simple model a bit more concrete. It saves you the cost of emailing docs back and forth (or running your own wiki, etc).
Now, this opportunity seems tiny to most of the usual suspects. But the point is the return - for a very small investment, you can create an absolutely enormous coordination economy. The market size isn't hundreds of billions - but if you can aggregate enough coordination economies, the market size grows, and you realize strong scope economies as well: the whole opportunity is much greater than the sum of it's parts.
Now, they distinctly didn't get this argument. Why not? I think inertia has a lot to do with it. Let's consider the usual suspects for a sec. Google is busy extending the contextual ad model to new domains (viz, location based) and leveraging competences into new markets (viz, leveraging their new unbundling competence to slice, dice, and resell print ad space); Yahoo continues to try and string Web 2.0 bits and pieces together into some kind of coherent whole, and develop 2.0 competences (which is the wrong approach, IMHO).
This kind of inertia is pretty cool, IMHO - because it leaves huge gaps open, just like I've been predicting, for smart players to engage in coordination arbitrage (viz, JotSpot, TheFaceBook, iMeem, Last.fm...etc).
I am surprised that they didn't get this argument : after all, this is a proven idea in the enterprise space where Groove built its entire platform around simplifying coordination within an enterprise. I have used Groove platform with excellent results to callaborate with my team in b-school. It's really quite powerful, you only have to experience it for the its value in simplifying coordinaiton to become very obvious.
Based upon my experience with Groove as well as wikis and portals, I am not too sure how well Writely will work. Web-based systems are notoriously difficult to build for managing latency and coherence capabilities required for real-time collaboration : P2P is an inherently better technology for that purspose.
I'm one of the Writely engineers.
This is a good point, that it's very hard to do good coordination within a web app. We considered this as well. It goes both ways - P2P means you have less control over the local setup, and it can be harder to deal with, etc.
We've spent an enormous amount of energy so far on "hidden" aspects of this problem. For example, most online editing solutions don't contemplate live co-authoring - we do. We notice this, coordinate and manage merges, let you know who else is editing the document, etc.
It's certainly not perfect, or even where I want it, yet. But I think it's possible to seriously address this kind of issue in a Web based system. Well see, anyway!
Sam, thank you for your respone. I'm sure it'd be a very interesting problem for you to work on - speaking as a Linux geek, I know I would :-). I agree, having no client side set-up is an advantage over P2P. Good luck and best wishes for your venture's success.