Design for timely interactions

The Notifaction Model for our wearable world

The new category of devices we call wearables are entering the market for a couple of years now. Smart watches and Smart glasses are the most well-known. The interaction model for services on these devices have a different architecture than that of mobile phones and tablets. Looking into the setup and use of the Google Glass last year I developed a model to describe the new interactions. This model is also very applicable for the smartwatches that come to market. The Notifaction Model try to give insights in the changes.

From destination-based to trigger-based

With the rise of mobile as primary touchpoint for a lot of the products and services we use we see how important the app architecture has become. Despite the discussion on web apps versus native apps, the model of apps has changed a lot in our use of digital services. Functions are split into apps and we relate to our service providers and favourite brands via their apps.

With the app architecture we have to decide ourselves to open an app to start using it. Notifications are added to bring more dynamic interactions with the user. Those notifications are often very one dimensional and not relevant for the user, resulting in a lot of users switching them off.

The traditional app model

With the increase use of mobile also as primary device we see how apps become full functional services. The good services create an ecosystem of native apps and responsive websites, where the latter is the primary for on boarding to a service, and the apps have a role as soon as the user has a longer relation with the service.

Services get more and more relevancy driven. When do you need what information. Very little apps change their entry state depending the moment of use, but I think we will see that more often appear.

Next to the embedded concept of relevancy in service ecosystems we see also a mixture of devices that are touchpoints to the services. Especially the wearables bring a new take because they are still highly intertwined with the smartphones. Most wearables do not work on their own.

So what happens is that we get services that get triggered by the impulses from the context and compose the right information. Profiles are key, both stored in the cloud as ad hoc profiles that emerge from the actual use of a service at that very moment, captured by the sensors.

With the Notifaction Model the notifications orchestrate the divers touchpoints of a service. The notifications are defined by profiled data and real live data provided by sensors.

Profiled impulses

The impulses drive the experience of the service and the orchestration of the touchpoints. What kind of information is offered depends not only on the users active device but also on the state of mind. In this new impulse driven world the context of tasks is the knowledge on the user from his profile. If possible this profile is rich and continuously built during time. But it is also possible to create ad-hoc profiles by referencing behaviour to known patterns that are stored in system. This will become the default practice. Imagine you entering a grocery store and picking a couple of vegetables. From that moment on it is possible to compare the combination of your choice to the stored combinations. From the match emerges possible next steps and will the grocery app be able to give you highly personal recommendations.

The ad-hoc profiling is fed with input from sensors. Cloud based profiles and here and now sensor information is combined into the rules for serving the right content at the right moment.

The well designed services use that knowledge to optimise an experience without dictating fixed routes. These kind of services are rule-based by nature.

Rule-based services

Services and also products are not static anymore in their presence and functions. It is becoming an accepted given that designing for frozen moments in user research is not the way to go anymore. Personas become task-based. Customer journeys are the standard. These are however still not as adaptive as necessary. In the Notifaction Model we think in rule-based services. The principle is inspired by gamedesign where rules are the core of an experience.

The architecture of a rule-based service have a solid core function; what does the product solve in essence.
Next to that a service has an adaptive layer filled with specific instances that is shaped by the use. These are the product service combinations that will be the defacto standard the coming years. The concept of cloud-washing machine of BERG was a good example how this concept can be designed well. It shows that you need to rethink the product from the ground up starting with the core function, and not just add an app to an existing product.

The physical products themself are still evolving into this model. There are some materials that can adapt which is super interesting. Think of a long-sleeve shirt that becomes a sweater as the temperature lowers, and can have different prints depending your location of wearing. We of course won’t speak of shirts and sweaters anymore when that moment arrives…
For the coming years the adaptive layer will be mainly in the service connected to the product. The product will generate specific triggers though thru the integrated sensors or even connections. The internet of things is not literally a network of things connected via the internet, instead it will be the current and new products that have a direct relation to the product cloud. The whole of the product and the service is defined on a shared promise, and the good ones are designed as a whole.

Design of notifaction services

As the products and services are adaptive the way you need to design those services is different from traditional design process. The focus of this article is not on the detailed description of the design process, that is a good topic for another post. But let me give some cues:

  • Define the proposed change the product service need to achieve, the outcomes is need to have. Design the meaning it need to deliver, not only the presence.
  • Stop making personas as starting points, describe contexts of use, and environments of usage.
  • Stop making static customer journeys. Use that time for the rule design.
  • Create a system that can adapt, that learns by the use. By every use. Under-specify rather, leave the behaviour free to evolve.
  • Start testing early. Almost right away. Iterate. The right process is continuous testing and improving, and in the end this remains the practice during the whole of the life cycle, although the cycles changes in lengths.
  • Do less, offer more.

Some related developments are the use of wearables as interaction means and the rise of the conversational UIs as these cater the open dialog we will have with the things we use. That is also a topic to address in a seperate post.

Timely interaction

The most visible implementation is that of the smartwatch operating systems. Both Android Wear and Pebble Time have a strong timely interaction concept. The now-state is start of the functions and the default actions are based on a look in the future (agenda) and the past (timeline). The apps are only in a second layer reachable and often less important than the functions connected to the moments.

Apple WatchOS 1 has still an app overview screen. But as user you discover soon that this is not a real important starting point. Notifications and glances are much more used entries. In WatchOS 2 a fully timely interaction will be added by giving the possibility to turn the digital crown to look into the future from the watch face.


To conclude; what are the consequences of this new approach in a more broader sense?

  • We will have a New Design approach; We mentioned the new design that is evolving around adaptive rules-based systems.
  • Media change; Media will be less separated from the context of use, but wil also integrated part of the product. Products are media.
  • Manufacturing: new materials. Less variation between products, more variation in the products itself.
  • New products; in the end products are changing and our valuation of products is changing.
  • New interactions. More use of haptics and conversational UIs that create new connections with the products and services we use.

The Notifaction Model aims to create a new approach of thinking about the design of services. For the new products some iterations will be necessary.