I agree, but I want to write down some of the other calculations that go into putting constraints into products.
Two significant projects I lead at Flickr were designed with a major "creative constraints". The first, video on flickr, imposed a 90 second limit for video playback on the site.
The second, Galleries, is an extensible tool to curate photos from other members, and it has an upper limit of *just* 18 photos in each gallery. Galleries is impossible to find on the Flickr homepage, but it's actually pretty cool.
Constraints are simple, elegant and can be revisited after the fact.
The most obvious fact about constraints is that you can always relax them, but you usually can't go back and impose them. So, you can start with a completely nutty 90 second restriction and then gradually relax it to say, 2 and a half minutes. But your 2.5 minute shooting customers will actually flip out if you suddenly tell them that they're no longer welcome because you're lowering the limit to 90 seconds. Duh!
The less obvious calculation about constraints is one that I think is actually worth a good amount of thought: being a considerate storyteller. When someone creates a voluminous bit of user generated content, it can be inconsiderate to their viewers. 45 second videos of my nieces and nephew are a perfect escape from my desk job, but 8 minute videos are kind of a slog (for most people). When we created galleries on Flickr, this was one of the intangible reasons we started with 18-- plenty of room to tell a story, but not an exhaustive repository of every photo on the subject.
Why 90 seconds for video?
Beyond the topics discussed above, I think Stewart and Heather (and others) knew what video on Flickr was long before I arrived on the scene. It was "long photos" and a careful expansion of what photography was.
Beyond this thesis about what photography is, Flickr wanted to manage video costs and to ensure that people didn't put stolen video from around the web and TV onto flickr. The 90 second limit supported those goals too.
Do I wish I could go back and increase the limit a little beyond 90 seconds now that video is well established on Flickr? Yes.
Beyond this thesis about what photography is, Flickr wanted to manage video costs and to ensure that people didn't put stolen video from around the web and TV onto flickr. The 90 second limit supported those goals too.
Do I wish I could go back and increase the limit a little beyond 90 seconds now that video is well established on Flickr? Yes.
Constraint decisions in Galleries.
I got to work on this project with some of the most fun, creative and stubborn people at Flickr: Heather, Aaron, Adrienne and Jude. Galleries had been a long-pondered feature for Flickr, and something Heather and Aaron had explored before Flickr with the Mirror Project.
As I recall, the biggest driving force with this project was to get something out in front of Flickr members, and we made a number of design and policy decisions to just get it out and learn how people would use the service. You could only add public, safe photos to a gallery. Every member could opt their photos out of any gallery, and would be notified whenever their photo was added to one.
And then the limit of only 18 photos.
We argued so much about this, mostly with the extended flickr team and management. But not within the Galleries team so much. The folks who were thinking through the basic uses for the feature were all bought into the 18 limit. And so we just pushed it through, because it seemed the right place to start the conversation with our members.
We argued so much about this, mostly with the extended flickr team and management. But not within the Galleries team so much. The folks who were thinking through the basic uses for the feature were all bought into the 18 limit. And so we just pushed it through, because it seemed the right place to start the conversation with our members.
--
I'm sure there are lots of other flickr examples of constraints, but these two exemplified our philosophy at a time of Flickr greatness.
Constraints are also a result of physical/scaling/ops variables. For example, any social activity extrapolates out the costs exponentially with each additional contact node -- but giving users everything they need to know about their contacts without restraint to time, degrees of separation, etc. detracts not only from usefulness but also the ability of a service to scale.
ReplyDeleteInteresting post, Shanan. In some respect, Facebook destroyed MySpace because of its design customization constraints. Facebook constantly seems to be reducing its features as fast as it ads them, especially on the ad side. That has a real impact.
ReplyDeletevery interesting
ReplyDeletenot sure I agree on the FB premise but food for thought...