The process of creating a user-centered product should be guided by research that shows what works for people who use the product and what does not. It’s not uncommon for a promising concept to not translate well as it goes from the drawing board to physical product. To minimize the risk of failure, more and more product designers are starting to collect and analyze data as they are designing products.
Data-driven design helps product designers get a better understanding of how users interact with their products. In this article, we’ll break down how to connect data to a web design and use the findings to design future improvements.
What is data-driven design?
It’s possible to define data-driven design as the way of designing backed up by findings from data. It’s a process of developing or improving a product based on things you can measure. Designers who ignore data and rely solely on their instincts risk wasting time and money on creating ineffective solutions. Many designers suffer from false-consensus effect—they project their behaviors and reactions onto users and make decisions based on their own thoughts and experiences. While all designers have hypotheses and assumptions on what solutions will work for their users, every hypothesis needs to be validated.
Data can tell designers a lot about user behavior: the preferences of your users, what they like/dislike, how they interact with digital products, and what devices they use.
How to use data in your design process
As a product designer, you have a lot of opportunities to rely on data when planning your design process. Basically, at any stage of product design, data can support designers’ decision making and help them to find solutions to problems. But this only happens when designers know what data they need and how to make a proper design based on it.
Here are a few general things that you need to remember to effectively use data in your process:
Understand user needs and business goals
Data-driven design isn’t just about gathering as much data as possible. It’s about collecting data that can help you find insights about user behavior and using this knowledge to help you make a better product. Thus, it’s essential to understand user needs and business goals first. Invest your time and effort in user research and select a set of metrics that match your business goals.
The Jobs to be Done (JTBD) framework can help you find your target audience’s goals. With JTBD, you can think of data as something that helps people accomplish tasks.
Define your primary data sources
When it comes to using data in your design process, you should always start by evaluating the data sources you have. If we compare a startup that has just released its first product on the market with a company that has a sustainable and profitable online business, it will be evident that both companies will have a different data-gathering process. Generally, a data-gathering process will be harder for a startup because it won’t have a stable user base, and it can be challenging to find a cost-effective way to measure user behavior. Thus, it’s important to understand how much time and effort you are able invest in collecting the data upfront. This information will help you plan your design activities—team members will prioritize activities and create a better research-design-validation process.
Present data visually
Many people consider themselves “visual learners.” Presenting data visually is one of the best ways to draw attention to key messages. Using data visualization as a tool, it becomes much easier to captivate your audience and convey your message. This is especially true when you work with large data sets. Even simple charts and graphs can make it easier to comprehend the information.
Visuals work especially well when you want to view correlations between a few metrics, such as time-on-task, user confidence, and task completion on different platforms.
Data-driven and data-informed design: What is the difference?
Data-driven and data-informed design represent two approaches to working with data. The data-driven design approach uses data as its foundation, and every decision is evaluated using that data. In data-informed design, data is used as a reference when a design decision is made.
Depending on the nature of your project, both models can be beneficial. For example, a data-driven approach may be appropriate when you want to do performance optimization—quantitative metrics such as time-to-load will help you to understand when you will face performance bottlenecks. Data-informed design, on the other hand, is great when you want to know what problems a user faces when they interact with your product, introduce changes in your design that will solve problems, and measure the impact of your changes.
Use quantitative and qualitative data together
Many UX practitioners believe that data only means numbers. This is a common myth. While quantitative data is the foundation of data-driven design, you shouldn’t rely on it solely when making your decision. When it comes to designing with data, it’s recommended to use a combination of quantitative and qualitative methods. Why? Because quantitative data will tell you what actions users take when using your product, while qualitative data will tell you why they do it and, even more important, how they feel about the overall experience.
It’s also important to remember that using just one method of research isn’t going to give you the depth of insight needed to make useful changes. So let’s explore popular quantitative and qualitative methods.
Quantitative data-collection methods
Quantitative data-collection methods provide the answer to the question, “what do people do when using your product?” The methods below will also help you focus on what parts of your product (features) you want to measure first and how you want to do it (what metrics you want to use for that).
A/B testing & multivariate testing
A/B testing (also known as bucket testing) involves running a simultaneous experiment between two or more pages or screens to see which performs the best. A/B tests change one element in a design (such as the color of a call-to-action button) between two versions to see which one performs better. A/B tests are fairly easy to conduct—you can show version A to one half of your audience, and version B to the other. The goal of this testing is to find out what version of your design works better for your users by measuring a level of conversion (e.g., for a landing page, this might be the number of sign-ups).
Multivariate tests, on the other hand, change more than one element (like an entire header of a page). When teams run multivariate testing, they define combinations of variables. The goal of multivariate testing is to determine which combination performs the best out of all defined combinations. Running A/B or multivariate tests regularly will help you garner higher conversion rates because the data will tell you which solutions work best for your target audience.
Web design analytics can tell you who has come to your website, how they got there, how long they stayed, and what they clicked. Tools like Adobe Analytics and Google Analytics can help you collect valuable metrics such as average length of the user session, bounce rate, etc. If you want to optimize a conversion rate on your app or website, it’s recommended to start with highly trafficked areas because they will help you gather valuable data faster.
You can also use eye-tracking tools such as heat maps. If you notice that large groups of users focus their attention on a particular part of your page/screen, you will be able to define interaction patterns based on this behavior.
While surveys can be used for collecting both quantitative and qualitative data, it’s much easier to use them for quantitative data. Why? Because in many cases, it’s easier to prepare questions focused on quantitative data rather than qualitative. A good survey should be well-designed. You need to ensure that your questions are not leading, and the meaning of questions are clear to your test participants.
Qualitative data-collection methods
Quantitative data-collection methods provide the answer to the question, “why do people do something when using your product?” The methods will help you focus on what parts of your product provide maximum value to your target audience and reveal your target audience’s wants and needs.
Interviews can be used as a follow up for a survey. After you’ve uncovered where people do when using your product, you will likely want to know why they do it, and it’s easier to find the answer to this question by talking to the test participants in-person or over the phone.
User feedback on social media
Social media sentiments show what expectations users bring to an experience. By exploring common user complaints, you will be able to prioritize product requirements. This activity will also help you get a more relevant Net Promoter Score (which measures the willingness of customers to recommend a company’s products or services to others).
A contextual inquiry is a type of observational research. A UX researcher watches how a user interacts with a product in their environment (e.g., their workspace) and makes notes. The researchers might want to clarify some interactions and ask questions like “Why did you click that button?” or “What do you think about this message?’
User flows analysis
User flows are the ways users travel through a site, from the page they first land on to the last page they view before exiting the site. Typically, UX specialists have a certain flow that they want the user to follow. However, the actual flows might differ greatly from expected, and it might indicate a problem with the user experience. The easiest way to conduct a user flow analysis is to ask users to complete a specific task, watch how they will do it, and ask specific questions at the end.
Regardless of your goals, data-driven design can help you improve product performance and increase conversions. It’s always helpful to think of data not as numbers, but more like a member of your product team. No matter what you do, data can always support your design decisions.