[youtube http://www.youtube.com/watch?v=XetUkZiNLxI&w=560&h=349]

When people think about and research smart services, they focus on the technical aspects, items that are measurable. There's a false belief that just because a technology is more contextually aware it will, by default, deliver a better experience, says Olof Schybergson (@olof_s), CEO of Fjord (@fjord), a digital service design company. The problem with that theory is that it completely ignores human traits which are often idiosyncratic and rarely rote, observes Schybergson. Digital content is made up of both technical and human elements. Human behavior can be studied, but it's hard to detect. Plus other factors confuse detection, such as group dynamics, mood and just general nuances in behavior. Weighing the Costs and Benefits of Smart Services Another issue: People are wary of the things like loss of control, privacy intrusions and annoyances, says Schybergson. These are some of the barriers to delivering successful smart services. Although once they understand the returned benefit, people are actually happy to pay these personal costs. But be wary of the service you offer and how you try to anticipate behavior. Microsoft's Clippy ("It looks like you're writing a letter") was the most famous case of a smart service that poorly anticipated behavior in an effort to be helpful. To avoid your smart service being a subject of irritation and unease, developers have to be aware of falling into the "uncanny valley," the attempt to replicate human traits and experiences but just falling short which, even though slight, is enough to cause a repulsive response.

What Fjord discovered is there's a high point of experience with a smart service before and after the "uncanny valley." Your goal is to aim for one of those two sweet spots. In a private conversation, Schybergson and I admitted that shooting for a smart service that perfectly matches human traits and experiences is expensive and can take a very long time to achieve. A good example of this is dictation software, like Dragon Dictate, which has gone through decades of iterations and passed through the uncanny valley. In their attempt to be really smart, many services fall into the uncanny valley. And sometimes a small change in an existing application can cause a slide into it. It can be very difficult to figure out how you got there. To design for these sweet spots, you need to understand human elements. Schybergson offers up three areas of opportunities and four insights for developing smart services. Three Takeaways:

1. There’s a great need for service mashups across industries, environments and business models.
2. Given our fragmented lifestyles, everyone could use some type of personal assistant service.
3. Digital will dissolve into the physical environment.

Four Insights:

1. The user is the OS. We really need to understand a significant amount about people to know what spot to shoot for.
2. Privacy is the currency. People will use their personal information as a trading system.
3. Digital becomes physical. It's also an opportunity.
4. The mashup economy needs orchestrators