July 22nd, 2009

Finding Innovation in Design

by Joshua Porter  |   4 Comments  |  shortlink: http://bokardo.com/p/1327

The best way to innovate in design is by solving today’s problems…if you’re successful then you’ll soon be asked to solve tomorrow’s as well.

In Chapter 2 of my book Designing for the Social Web, I introduce and talk about what I call the AOF method. AOF stands for Activity, Objects, and Features. First you determine and research the activity you’re going to support. This helps you identify the social objects within that activity and the actions people take on those social objects. These objects and actions become your feature set.

The idea is that your features are nothing more than the actions people take on accepted, agreed-upon social objects. They are not pie-in-the-sky features that may or may not be valuable. By modeling your software on existing behavior you can be sure that the features you add will be valuable…after all…people are already doing them.

However, in talking with folks since the book came out, I’ve learned that I didn’t stress enough about how the AOF method is meant to model existing behavior. That is, the activity we model isn’t a future activity, a hypothetical activity that might possibly exist in the future if all goes well. It’s meant to support an activity that people are already doing today.

You’ll notice that the most successful software does this. Skype models making phone calls. Salesforce models the customer relations. Netflix models choosing and recommending movies. In each of these cases the technology is new but the behaviors are not: they have been around for a long time.

There are several benefits to modeling existing activities:

  1. Because they’re current activities, you know they are conducted to solve an existing, real-world problem.
  2. If the activity is non-trivial then you know there is an existing market for your product.
  3. You can reduce feature creep because you prioritize features based on observed need, not hypothetical desire.
  4. Existing activities are much easier to research than hypothetical ones. Your research will be grounded in reality.

One concern I’ve heard about modeling existing behavior is that some designers feel it’s not the way to innovate, to truly come up with something new. This may be true…to start. The AOF method is meant to be a practical way to come up with a valuable feature set for existing activities.

However, it can be extremely difficult to create something innovative until you’ve learned what the current problems really are. Don’t assume you know enough to really solve existing problems if you’ve only identified them from a 50,000 foot view. If it were that easy then the problems would already be solved. There is a world of difference between our initial assumptions about a problem and what we learn once we actually try to solve it.

It is this effort, the effort of trying to solve an existing problem, that leads to innovation. Let’s say you start off modeling existing behavior using the AOF method and you’ve got some people using your software. What happens is that you start talking to your users and can’t help but see how your software can be improved. People will give you feedback and guidance about what they need and want…they will essentially map out the way forward for you.

Now you are ready to really innovate, to push further than the current problem set. And, you won’t even have to come up with the ideas yourselves…because you’re actively supporting existing behavior people will naturally want more and more from you. Innovation is somewhat counter-intuitive in this way…the best path to it is not necessarily the direct path, but an indirect path that comes from nailing the current activity to the ground.

So instead of looking too far forward, design for now. Focus on current activity and an existing problem. You won’t need to find innovation. It will find you.

Check out my latest project: Make them Care!, a book on designing great sign-up experiences. Get reminded when it's published.

Links to this Post

Comments

1.  Dave Malouf 12:29pm, Wed 22nd, 2009

Josh, I have a few issues w/ the above while at the same time acknowledge where your at. I think that for some people what you are professing makes sense. But for many more this is too narrow. I do not think in this case it makes sense to apply the tone of authority you have within this piece as a “global truth”. Instead it is an offering that fits a limited pool of design types.

Here are my issues though:
1. while at some high level the 3 examples you give do address existing activities, by saying that they do so, negates how much all 3 have radically changed the ways these activities are performed and thus the innovation of behavior they created. I.e. to say that netflix addresses the activity of watching movies is a bit off.

2. I also think that you are using the terms behavior and activity equally which I think causes the issue in #1. If you are clear that there is a high level activity but we can innovate the behaviors for achieving or accomplishing that activity then I think it might be more strong/sound.

3. At some point though (to my initial criticism) there are many people who are looking to invent activities and sometimes when your area of design is big enough, addressing one activity type, say transportation (horse > car), may lead to the invention of new activities (filling gas, changing oil, washing car) that in turn create completely new activities around completely new services.

The best example of having to think forward today is how Nissan/Renault is bringing battery powered cars to Israel. They are not just selling the car, but installing a completely new infrastructure. W/o the new infrastructure (battery exchange stations) the entire project would fail.

Now you could say they discovered the need based on modeling existing behavior, but they also had to model future behavior along the way to understand the validity of their answers since you can’t test it until it is built.

– dave

2.  Mike 5:13pm, Wed 22nd, 2009

This raises a lot of interesting questions about what innovation is and whether designers should be doing it. What innovations make Netflix work? DVDs that are small enough to be mailed, cheap internet access and cheap web application infrastructure. What makes Google innovative? PageRank, a more accurate algorithm for generating search results which means faster access to information. Electric cars are innovative because they can run on non-fossil fuel energy sources, making it less environmentally costly. Lower transportation costs lets people go on vacation on the other side of the world, etc.

From these examples, it seems like innovation is just a fancy word for “cheaper”. People have always wanted what these technologies provide, and the innovation is in making it cheap enough that it can be made widely available. As a designer, it’s not clear how I would participate in this. Do designers innovate by reducing the complexity of a product/service, making using it a more efficient use of mental resources? Isn’t the point of user-centered design that many technologies often shift costs on to the users? “We introducing a brand new system to make things much more efficient! We’ll be having a 4-day training program…” In fact, aren’t some of these new user-incurred costs *caused* by the new system’s innovations.

This doesn’t seem to cover all aspects of design — aesthetic, cultural, psychological — and this might mean that the scope of the word “innovation” doesn’t cover design, or at least not completely. I think its extremely important that so many people are obsessed with the next new thing, and yet what designers do, exactly, seems to escape their notice. It’s as if the notion of innovation has stagnated, and what we need today is not new practical innovations, but a new theory of innovation within which we can locate design.

3.  Josh 4:45am, Thu 23rd, 2009

@Dave, I’m not sure I understand your criticism…it sounds like we’re pretty much in agreement. Your example of Renault sounds like a good one…they’re making some existing activity (powering your car) easier. If they didn’t have the confidence that this was going to be successful (say, if they were inventing a new activity) then it would be a fool’s errand.

And, by pointing out how all of these three designs model existing behavior, I’m showing how important that is! Design isn’t about inventing something new for people to do, it’s about inventing better ways to do what we already do…that’s what these services have done.