Published on | Prototyping: From UX to Front End — Medium Shriman Kalyan

Every user is different. Why offer them all the same experience?

Amigos UX designers!

Imagine a day when we are able to craft app experiences that adapt to each user and their behaviour, on the go… that day is not too far!

Yes, you read it right. One version of the app, one base design, but a design that gets customised to each user based on what we know and learn about their interaction patterns!

Think of this as personalised product experience, a concept very similar to personalised content, but now for design.

App usage patterns are the most direct and valuable form of user intelligence.

Why a data driven and dynamic design concept?

The push for data driven UX and design is going to be driven by the long term usability improvements, cost saving and also the limitations in the current / traditional design process.

Current design process in the industry is usually lean, iterative and user-centered. A combination of ideation techniques, workshops, user surveys, user persona driven journey mapping, interviews, prototyping, usability testing and iterative design practices are used in the app design process to best understand target users and design an experience that is aligned to users’ mental model and expectations.

Let’s design ‘for every user out there’ rather than designing for personas.

In addition to these, A/B tests, post launch analytics, app store feedback are the some of the common methods obtaining quantitative input and direct user feedback post launch aimed at continuous design improvements.

But some of the key limitations of this approach are:

Extremely limited coverage of user personas
Assumptions about the target user segments
Possible research bias as well as design bias
Limited population sample for user research and testing
‘1 design fits all’ solution

Moving forward, designers would be working very very closely with the tech teams.

With data driven design approach, we would still be adopting most of the current design tasks and process but in addition to these, designers would also be working closely with the tech teams to analyse, predict, define and document the design constraints and boundaries to ensure that usability standards are met during the dynamic decision making. This is pretty much the process we adopted at Outware when building the proof of concept demos for the above use cases as part of the recent Hackathon.Data and patterns

Every action that a user performs on the app is potential raw data. Right from the app launch, time spent on screen, navigating to different sections on the app, scroll behaviour, gestures and literally every action performed can indicate so much about the usage behaviour. This along with struggle scores and error analytics add up to a huge data set. This can then be used to create User Intelligence data points.

Now most of this data would be unstructured. As with big data in general, the challenge is to make sense of data to make meaningful inferences. Identifying and analysing patterns help us define the rules and also plan for applicable use cases for the dynamic behaviour. This is also a great opportunity to utilise Machine Learning systems to learn from user behaviour rather than having to define rule sets for specific use cases.

The tech challenge is to make sense of unstructured data to make meaningful inferences.

Example use cases

So what are we talking about? What could be some examples of dynamic design? Couple of generic ones …

Layout Restructuring Algorithm:

Based on the usage of different sections or information sets on a screen, the app could adapt to present the most used and most relevant content in an easy to consume layout. On screen interactions, time spent viewing different scroll states and basic usage analytics can be used to understand usage patterns.

For instance, if a user has regularly scrolled to view particular content then wouldn’t it be great for the app to ‘learn’ this behaviour and present this content on the landing view itself? This directly removes the interaction cost associated with scrolling.

UI Resizing:

Similarly, if we know that a user has regularly had trouble interacting with a UI element then the app could scale up the UI element sizes to allow for an improved usability. This is almost equivalent to catering to accessibility on the go rather than having to turn some accessibility related setting on / off at the device level.

In addition to these use cases, tailored algorithms can also be built to cater specifically to the needs of the product. Also, additional parameters such as location, time of use, type of device, device orientation and other contextual information could be factored in to fine tune the decision making system. As you see, possibilities are limitless and this is potentially the future of digital product design process — not just for apps but also for websites, wearables, voice based UX, auto and what not!

Product and usability considerations

But such a drastic move away from the traditional way of designing will need careful considerations around performance and app bloat. Smart solutions for tracking such as global listeners rather than asset level interaction tracking, managing design assets such as scaling, logic hosting on the cloud rather than on the app so as to keep the app lightweight, etc. would need to be adopted to ensure an enhanced product experience.

On the usability front, defining the logic to allow the system to make design decisions on the go becomes key for this framework to be successful. Also, smarts around user communication when a design adaptation occurs, option for users to opt in/out of this feature and the option to roll back to the previous state of the experience would provide a great sense of control for the users.

Defining the design boundaries and logic to allow the system to make design decisions on the go becomes key for this framework to be successful.

And finally there needs to be a conscious effort to take into account any particular business priorities that might influence the dynamic design decisions. For instance, with the Layout Algorithm, on a particular screen, it might make sense not to apply the dynamic design update for the hero area which serves critical info to the users.

Future

In the world of UX and usability, user is the king.

As UX designers, our primary focus in product design is to have user empathy and design experiences for the users. This is the foundation of user centered design approach.

And a dynamic decision system for design that is based on individual user’s usage patterns has a great potential for continuous design improvements leading to user experiences that are impactful, frictionless, subtle and graceful.

In a nutshell, it’s all about improving the usability and reducing the associated interaction cost.

All in all, a promising step forward in the quest for the ‘optimal design’!

Data Driven Dynamic Design was originally published in prototypr on Medium, where people are continuing the conversation by highlighting and responding to this story.

Data Driven Dynamic Design

Cool stuff, every week

More Articles