Data Driven Design

Data-driven design is quite counterintuitive most of the time. A fancy hypothesis that comes from your most creative insights might gets rejected by the users. 

So, it is essential for design teams to embrace the data-driven approach to design. Let’s get started with data-driven design, what it means and the different nuances involved with it.

When customers interact with the design elements of a website, they leave rich information behind - like clicks, hovers etc. Using this data to inform our design decisions is data-driven design. This research data helps designers to craft the ideal user experience.

Data is the foundation for data-driven design, and it comes in 2 forms: 

• Quantitative:  Data that shows the who, what, when and where. It can be captured via clicks, hover, and heatmaps (which section of the website people stay on). Tools like Google Analytics, Mixpanel and hotjar help capture this data. 
• Qualitative: Data that shows the why and is focused on user motivation and intent. This is captured via user interviews, usability testing, and focus groups.

‍

A lot of times, people dismiss qualitative data as anecdotal and ignore this data completely. But, this data is precious and helps provide clarity to a lot of the design decisions. Now let's understand how to use data in the design process.

🗣️ Understanding the user needs and expectations

Understanding the user needs and expectations: The data you collect must inform you about the users' behaviour and motivation. Any product must help users accomplish a task, and the data must be aligned with that. So you can decide on what data can help you get a better user understanding.

🧐  Right Metrics

Fix the right metrics that make business sense and anchor the design decisions on those metrics. Ideally, these metrics should be 1-2. A good rule of thumb: Will this metric inform me about user behaviour? Will this metric help move a core business metric?

For example, if we consider bounce rate? (Bounce rate is the percentage of users who visit the main homepage but leave without taking any action such as clicks, survey etc.) Understanding the bounce rate helps you understand which section of your users are leaving too soon. Arresting bounce rate can help improve the business.

‍

🧐  Present Data Visually

The right visualization of the metrics helps inform the stakeholders. Present clearly how the data supports or does not support the hypothesis.

‍

Understand that the data-driven design process is circular. You will get a few insights in the first iteration that help you to frame better hypotheses for the next iteration of data collection.

‍

Few tips for data-driven design

‍

1. Segment users while looking at the data. Returning users and new users might have different needs when visiting the website.

2. Maintain the balance between intuition and data. Don’t blindly go by what the data says. If the data is inconclusive, go with your gut and restart the experiment with a different hypothesis.

3. While showing data and metrics, always assume the stakeholders don’t understand anything and speak from the basics.

4. Be open-minded. When data conclusively shows that you have to move away from a given best practice, go ahead and do it.

5. Keep creative juices active. Pure data-driven design may stifle creativity. Always keep a 15% buffer in the design experiment funnel to try wild ideas. Do crazy experiments from time to time that makes sense only in hindsight.

Every great design begins with an even better story. -  Lorinda Maso

‍

Tool of the day

Heurio

‍

This chrome extension helps you with UX check, design & development review, heuristic evaluation, user testing and more. With this extension, you can make comments directly on any website or live application.

‍

‍