June 08, 2005
Designing for the Sandbox
I've been thinking a lot about control and design.
People who design experiences often believe that in order to succeed they must exert complete control. And while in extremely rare instances they might be afforded the opportunity to dictate an entire environment (say, in a casino, or a theme park), when designing for the real world, for the ebb and flow of actual lives, such control is impossible. Often, the designer's response is to exert as much control as possible on their portion of this world. (Jeff found out this sad fact when reviewing submissions to an interactive design competition.) In fact, the best thing a designer can do is dictate *as little as possible.* Because the point isn't to control, it's to connect--to weave your offering into the complexity of people's life experiences, to allow them to figure out how to make sense of your offering within their world.
Photo by Yogi from Flickr
In my head, I've been calling this "designing for the sandbox." This acknowledges a space for content, tools, and people to interact and create their own meaningful experience. This is not a monolithic creation, that dictates how the content, tools, and people best interact. This is instead reminiscent of David Weinberger's phrase "small pieces loosely joined" -- things that connect, but aren't bolted onto one another.
The example I've been thinking of compares Ofoto with Flickr. Ofoto wants nothing more than to exert control. Hell, if I post photos to Ofoto, and want you to see them, you have to register to do so. Flickr wants nothing more to do than provide a space for interaction. Post your photos. Others can see them, no hassles. Connect photos with tags, groups, sets. Connect with people. Flickr has simply provided us a database of photos, people, and tags, with some ways to loosely join them, and let us go to town.
They've also provided an API that lets others dig deep into what Flickr has to offer. This leads to brilliant interfaces like Mappr. Or, in a more mundane, yet more impactful, note, the Flickr uploader plug-in for iPhoto. Before I used that plug-in, uploading photos was enough of a pain that I never got close to the max of the free account. Once I started using it, uploading became so easy that I needed an upgrade to Pro in order to use it how I wanted it.
So some passionate guy creates a free plug-in utilizing Flickr's openness, and Flickr gets my money. *That*'s the beauty of the sandbox. You encourage others to improve it, and you still get the fee. And people love you for it. Meanwhile, the other photo sites devolve into price wars, desperate for your business.
Del.icio.us, Backpack, Upcoming, and others are designing for the sandbox. Hell, Google is designing for the sandbox. Here's their API. Maybe you'd like to put some of their ads on your site? Go ahead. Or use their maps to view rentals on Craigslist? Sure, why not.
Thinking about this, I would say that, while people have done a lot to compare Yahoo and Google of late, this is a key point of distinction. Google designs for the sandbox. Yahoo tries to control the playground (am I overextending the metaphor?). Yahoo is still about drawing you in, and not letting you go. Google has always been willing to let you go, ever since they bucked the then-predominant search engine trend of "stickiness". Maybe Flickr will help the rest of Yahoo get the sandbox religion.
I wonder about how my clients could design for the sandbox. Is there a sandbox for Silicon Valley enterprise IT companies? Could Oracle have a sandbox? Sun? Can you conduct marketing in a sandbox? Is that what the Cluetrain guys were talking about?
And finally, I think about, how do we design for the sandbox? What are the guides? Is there a Tao of the Sandbox? My sandbox will look different than yours. Amazon.com was an early sandbox designer -- customer comments, wishlists, friends, etc. Barnes and Noble made feeble attempts at maintaining feature parity, but it was clear their heart wasn't in it. Could B&N be a sandbox company? How would B&N's sandbox differ from Amazon's? How could they relinquish control in a way that felt true to their approach to business?
Anyway, some things I've been thinking about.
Posted by peterme at June 8, 2005 10:18 PM
Yahoo does in fact have a pretty extensive API; look at the section "What Services does Yahoo offer?" on that page. There's also a list of apps people have built with it.
The GoogleMaps/Craigslist thing, while incredibly cool, isn't using a supported API. That's a straight up hack. Although apparently Google's already made some effort to make these sorts of hacks easier--and there's no way they're not planning a maps API at some point.
I saw that Yahoo had an extensive API. But, for whatever reason, the extensions people have built with it haven't really made impressions. And I think it's because this approach to the sandbox mentality isn't really... authentic for Yahoo. I mean, why don't the non-search properties have APIs? Why can't I tap into Calendar, messenger, maps, photos, etc.? Is Yahoo360 open?
I think the fact that Google is, well, hackable in that way is even more to my point. They built a sandbox application without using an API.
Sandbox design is in large part a mentality, an approach.
Posted by: peterme at June 9, 2005 07:40 AM
There are some rumblings-from-within that Google may be settling on a public API for Google Maps.
....which would be amazing, useful, non-evil, etc.
Posted by: michal migurski at June 9, 2005 08:23 AM
Posted by: peterme at June 9, 2005 10:49 AM
You might be interested in taking a look at my blog - the issue of control in design is a theme I've been exploring for awhile. I first wrote about it here, and in fact happened to write a bit about it last week - one entry where I suggest that one way to understand control in design is in terms of degrees of control i.e. like a scale - and in another, how we can understand it in terms of activity.
There is certainly interest in the "Designing for the Sandbox" end of the scale right now (previously I was looking for a term for this as well and the best I came up with was "Designing for Creativity"), and there are many ideas out there related to this - hacking, remixing, etc.
Posted by: Rick Pan at June 9, 2005 12:41 PM
i agree that the sandbox can be a great money-making tool by providing customers/fans the means to customize the product or service to their liking. However, i think you need to be careful holding up flickr as an example of a truly sustainable business model. if their exit strategy was to be acquired, then they did great. but what does this mean for cusotmers/fans? all the stuff you love about flickr may disappear (or get slowly diluted). why is that?
I think the problem may be twofold: Instead of designing for a sandbox, where play and discovery can occur, these companies build toolboxes that extremely skilled/knowledgeable mircrosubsets of people can use to build applications. Secondly, the commercial model often subsumes the interaction model. When Ofoto first started, it was a brilliantly usable and useful destination. With album sharing providing a collaborative space, Ofoto also enabled analog sharing through its easy ability to deliver physical photos. But business requirements to have more people buy photos instead simply browse them online forced some decision-makers to implement the registration-to-view opt-out. And when Ofoto became Kodakgallery.com (or whatever it is), the commercial stranglehold of control tightened.
As much as I love Flickr (http://www.flickr.com/photos/mojohand), I expect the same thing to happen.
Posted by: jsokohl at June 12, 2005 04:00 AM
I was a sandbox expert as a kid; we had a great big built-in box full of real Lake Superior beach sand, and the whole neighborhood played in it.
Good sandboxes are social spaces above all, it doesn't really matter that much how good the toys are (we had old 2x4 chunks and bits of plywood for building) as long as it's big enough and there are shady places to sit in the hot summer afternoon.
(and someone to scoop out the cat poop)
Posted by: Edward Vielmetti at June 14, 2005 09:28 AM
i completely agree with jsokohl.
ophoto was a *huge* upgrade to the old process of "shoot rolls of film, submit them for development, guess how many copies you want made, send them to friends/family or wait for them to contact you with requests." it also solved the digital film process of "shoot pictures, ftp to server (for the technical people) or email to friends (for the majority of people)."
if anything, ophoto successfully merged these two human behaviors -- pre and post-digital film -- into one useful and usable service. i had no problem providing light registration info (email) to be able to check out photos of a party from the night before. and if i wanted prints, the site leveraged established e-commerce parameters (full registration).
ophoto created a community vibe with the ability to announce that an album was ready to view. the gratification i had after receiving that email from a friend following their latest vacation, the birth of their first child, etc. made me want to do the same for my friends/family.
so here's my question for you: would the concept for flickr ever have come into being if ophoto didn't break ground first?
Posted by: spcoon at June 15, 2005 01:21 PM