Gone are the days when design used to rely mainly on the color palettes and the creativity of the designer. In the rapidly expanding technological world of today, it is essential to work across departments to enhance the screen experiences.
Three years back when I started working in user experience, I started with setting the groundwork: empathize, define, ideate, prototype, and test. I continued working on similar grounds for the next three years until I discovered what data analytics was!
Sometime back, I had a problem statement which required me to concentrate on personas to create the targeted user experience. I was learning data analytics at that time and thought of integrating data for the user experience I was designing.
And I was surprised. A bit more time on the data and its analysis gave me a more accurate, more human-centered user experience. I could infer the best of the strategies and scopes for the user experience I wished from the data.
So basically you use some data visualization tools like Tableau, Power BI, or Adobe Analytics, and make some pretty dashboards with your user data. Then you can make inferences on the target market, user sentiments, grouping data, survey data if you have any that can be well summarized, and so on.
The opportunity to leverage insight from data has never been greater.
Humans tend to generate a lot of data. With new technologies and products surfacing every day, we cannot miss our chance to analyze the data generated effortlessly and provide insights for better-decision making in business.
With the abundance of data being generated today, the value of data and related skills primarily go up, creating better opportunities for optimization in almost every technology domain.
UX + Data
What comes to your mind when you think of data? Accuracy, more insights, numbers, objective judgement, decision making.
For a designer, intuition and experience are the assets.
Data and design fused together gives us a term Data Driven UX, which I would like to define as the opportunity to explore the experience around data itself — how to make data easier to work with, how to get more value out of data, and how data enriches our work and lives.
The primary goal is for many tech companies these days is to create user-centered products. Therefore, having design and data intertwined and part of the same development cycle is a good idea. Companies are challenged to create memorable experiences for their customers across channels and platforms. So, how do we start on using the data in our design life cycle?
I used data primarily in four design components:
- Understanding the user personas
- Designing the task models
- Redesigning the UX
- Conducting heuristic evaluation
Every time you work on Data Driven UX, I would recommend you ask these three questions to the design team:
Why is data essential for this design?
How can you collect data?
What is the right way to use data once you’ve collected it
If you get answers to the above ones accurately, you believe there’s scope to talk on this more with the data team, go ahead! Integrate data into your design process.
1. Understanding User Personas from Data
Personas are character representations of a person, object, situation, problem, challenge, or redesign created based on the research to represent the different user types that will use the service, product or site. Creating personas helps to understand users’ needs, experiences, behaviors, and goals.
I usually end up using data the most while creating personas for background research, qualitative and quantitative research, and the objective market research.
To create user personas for an optimum UX, conducting and condensing user research is the most important part of the design cycle. Data Driven UX for creating and understanding user personas would felicitate:
- A near accurate portrait of your important and target users
- Outline the core user’s motivations, goals, needs, demands, attributes, behaviors
- Unites designers and stakeholders around a common understanding of who the user is
- Help designers in the decision making what product features would be the most important to the user
Data analytics and data visualization techniques and tools can always help for clustering, segmentation, determining relationships, and making insightful conclusions from the data. With them I can get an idea of the age group of my users, geography, their spending capacity, average screen time, their behavior, characteristics, the personas with the competitors, and so on.
For instance, say you are Company A that sells baby strollers, from $30 to $3000 (Yes! There are baby strollers worth even $8000, trust me! Check online if you don’t believe :D) Now, you are redesigning your app/website and revamping your brand experience and you wish to know the age group of the males that shop from your portal. You do not need to ask a student intern to create a custom SQL query as:
FROM Users U
WHERE U.Gender = "Male";
Just open up a data visualization tool as Tableau (Ask your data team what Tableau is, they would love to answer you!). Next up, visualize the age attribute from your data set against the gender of your users, filter the data out on males, and circle out on a scatter plot.
Voilà! We have a relationship trend for age group of the males that shop from Company A! Easy.
2. Designing the task models
I like to define Task Models as a description of each task in a workflow. It often involves documenting the comprehensive business and user information requirements in an accurate and consistent format. Now, as information comes into our picture, we must definitely rely on data for the insights.
Designing and implementing a successful user experience, one that meets the needs and requirements of an organization, requires a logical task model.
For instance, we wish to find the trend for users that spend more than 4 hours on phone; what apps do they use, what percentage of time is on social media, and similar questions to answer. In that case, we go ahead and plot visualization using the tools and observe the trend lines, and use bubble plots to show the segregation. Data team, step in please!
Implementing Data for Task Models:
- Identify the task to be analyzed
- Break this high-level task down into sub tasks and segment the data as categorical or continuous specified in terms of objectives and the area of interest
- Draw a layered task diagram of each sub task ensuring it it in line with the data we segmented
- Produce a data visualization that covers techniques to visualize our task model in hand.
3. Redesigning the UX
A bad UX design can definitely doom your next project, no matter how good the idea or concept behind it really is. For that very reason, you need to make sure that you support your decisions with the numbers. Here’s when data plays an important role.
Say, this time Company A has a problem; Users get to the payment page but drop out mid-transaction. Identifying the problem and redesigning the payment page is the task on hand. So, we now have to visualize how long after reaching the payment page the user drops out. So, from the amount of time spent by the user on that screen, you know what and where it’s going wrong.
Analyze factors that worked or did not work well for your user experience design supported by the data clustering or segmentation. Divide your data into positives and negatives – what works or not should be clear by now. Draw out a correlation between the factors of consideration and determine if the relations are strong or weak.
Working out relations and plus/minus for your data works best when redesigning your user experience.
4. Heuristic Evaluation with Data
Heuristic evaluation is a method for finding the usability problems in a user interface design so that they can be attended to as part of an iterative design process. It involves having a small set of evaluators examining the UI/UX and judge its compliance with recognized usability principles.
- Start your UX design with a qualitative data research to learn the why’s of user behavior using data. Visualize the data into clusters, derive regression models and pull up the qualitative outcomes that are non-numerical — preferences, feelings, inclinations using data segmentation.
- When a product or prototype is available for user testing, employ quantitative data research to learn how people actually interact with the product. Quantitative data is based on measurable evidence (e.g. success rates, task completion times, etc.) and is acquired through methods like analytics, A/B testing, and eye tracking.
I hope this read helps you in a new direction towards extensively implementing data into your user experience design life cycle. Happy reading! Cheers!