Data and design may seem like polar opposites, but in the world of User Experience (UX), they go hand in hand to drive substantial business outcomes. In fact, according to McKinsey, companies that employ data-driven design outperform industry-benchmark growth by as much as two to one! Another research by MIT’s Center for Digital Business found that companies in the top third of their industry using data-driven decision-making were, on average, 5% more productive and 6% more profitable than their competitors.
Why data-driven UX design
In addition to growth, productivity and profitability, data-driven UX design yields more great benefits for digital products and organisations. Let’s delve further into them.
Improved user experience – key customer analytics such as customer satisfaction, lifetime value, churn and acquisition numbers – all of this data can help product and design teams understand target users better such as identifying their pain points. In this process, internal teams can refine, optimise and tailor both products and communications.
Improved user engagement – with data, not only can your teams start to understand what motivates users to engage with a product and its features, but they can also begin testing out ideas such as personalisation to create more meaningful engagements.
Increased efficiency – using real-world data to inform design decisions means you and your teams aren’t working off intuition and hunches. This reduces the need for reiteration and rework later down the line. Furthermore, once you have those data collection processes in place, it becomes easier to collect more and more data – it’s compounding!
Better design decisions – making informed design decisions based on real-world data reduces bias in the design process. There’s a phrase in the design world that rings true here – ‘designers aren’t users.’ That said, they do need to understand how the product or platform is performing and whether design updates have had the desired effect. Using data to inform this understanding is key.
The power of data-driven UX design
Slack
Everyone knows Slack. But not everyone knows just how integral data-driven design is to the development of the platform, and additionally how much the team involve their customers in this process.
Back in 2020, Slack underwent a major interface overhaul to tackle increasing platform complexity. But they didn’t just do this on a whim, they brought in customers for feedback on a new interface before they made any design changes. They also ran a study comparing the old Slack experience with the new one, asking people to complete essential tasks. This was no mean feat for the team, as they were having to balance their enterprise user needs with individual user requirements.
Following the redesign, Slack underwent one of its most game-changing periods, and the results speak for themselves. Between March 2020 and April 2021, the company took $902 million in revenue, a 43% increase year-on-year. 2021 was also when the brand was acquired by Salesforce for $27.7 billion, a testament to just how much that interface redesign factored into the Enterprise market!
Airbnb
Airbnb is another brand that uses data to inform platform design changes, but often with very different outcomes in mind! In 2021, the Airbnb team decided to do some data digging. And they discovered that when people search for places to stay, they prioritise finding unique locations over a specific destination.
This surprising insight led the company to implement the ‘flexible destinations’ feature, which adapts to the user’s location and brings up unique properties they may not have considered. Users are greeted with inspiring castles in Spain or intriguing tree houses in England – not the typical city trip or beach break!
While the Slack story was very much focused on the financial results, Airbnb saw a wholly different set of outcomes, the most important of which was in Europe where bookings were diverted away from the most saturated tourist hotspots and peak travel dates. Airbnb continues to work with local tourist offices in Europe in support of more sustainable bookings and decentralised travel trends, as detailed here.
Black Dog Institute
Closer to home, the team here at Adrenalin worked with the Black Dog Institute on a new digital platform that put the customer at the heart of both the research process and the platform design. Black Dog Institute needed a platform that better served their customers – people living with mental health conditions – particularly following the global pandemic. The Adrenalin team used extensive customer research, including moderated focus groups, usability lab testing and multivariate testing, to inform the UX design. The result was a new site and platform that saw a remarkable surge in key metrics within just two months of its launch, all achieved with minimal marketing expenditure:
Over 51,000 unique completions of self-assessments for anxiety, depression and bipolar conditions
Over 800,000 interactions with mental health questionnaires on the website
Significant increases in donations, with over $300K raised
The how of data-driven UX design
It is unmistakably evident that data-driven UX design will bring significant benefits to your digital product. But how do you go about implementing a data-driven UX design approach?
Step 1 – Collect the data
First up, it’s crucial to understand the different forms that data can come in, so that’s where we are starting with the collection process. Data comes in two main forms – qualitative and quantitative. However, this isn’t about a quantitative versus qualitative debate, rather it’s about combining the two approaches to fully understand customer behaviours and usage, and to drive decisions that are as informed as possible.
Quantitative data
Quantitative data is, very simply, numerical data that shows who is taking an action, what they’re doing, when they’re doing it, and where they’re doing it. For example, survey and A/B testing.
Quantitative data is usually obtainable in larger volumes and at scale. On the flip side, however, sample sizes need to be adequate enough to achieve a confidence level and to be ‘statistically significant’ and collecting data in larger volumes can take time. Finally, quantitative data can tell you what, but it can’t tell you why – why customers are behaving in a certain way or what’s driving them to take specific actions.
Qualitative data
With qualitative data, we’re dealing with information and feedback that demonstrates the why and how, so it’s typically expressed in non-statistical means and often called ‘unstructured data’. For example, interviews and focus groups.
Qualitative data provides rich insights, helping your team to develop a deeper understanding of the challenge or pain point at hand. Typically, sample sizes are smaller and so can be quicker to work through. That said, you still need to ensure the data is significant enough – for that, we advise using diminishing returns to find the correct sample size. What’s more, be sure to allow for sufficient time after data collection to analyse it, as qualitative data is much more hands-on!
With both quantitative and qualitative data, there are two main ways in which you can collect it: primary and secondary methods.
Primary sources
This is when you typically run the data collection process yourself, such that the results are yours exclusively. While this approach can be expensive and ‘in the moment’, it is the closest type of data you’ll get to the origin and source of the information. For example, Google Analytics.
Secondary sources
Secondary data is existing research or information that you draw on from outside your organisation, sometimes called desk or secondary research. Often, this method is faster, lower cost, and provides more background and context. On the other hand, that context may be quite different to yours so this should be considered. For example, white papers and journals.
Top tips from the Adrenalin team on collecting your data
Get the rest of your product team involved in the user research – ask them to sit in on sessions you’re running so they can see how real customers are interacting with their product. Believe us, it makes all the difference!
Have two UX team members in as many research sessions as possible so one can focus on running the session and one can focus on taking notes. Then swap the roles to get the best spread of insights.
Step 2 – Analyse the data
From data… to insights. It’s important to turn your raw data into something useful and ‘intelligent’. Here are our top tips to achieve that.
Sort and categorise the data in order to understand it.
Layer in demographic and geographic data.
Look for patterns to form insights.
Be careful not to flatten’ the data or get lost in the details!
Write up your findings in a clear and concise manner.
Make recommendations on what to do next based on insights, areas of focus, additional research that may be required, or design directions to pursue, and features to add or change.
Present and share these results with the rest of your organisation in a compelling way to support buy-in for your recommendations. Displaying data visually is one of the best ways to communicate this and get full stakeholder buy-in to the next steps.
Throughout this process, the most crucial thing is to think about the context and the outcomes you’re trying to achieve with the data: what does it tell you about the product you’ve been designing?
Step 3 – Integrate and implement
Here’s where we action any design, feature and platform recommendations!
For example, this could involve changing an existing feature, adding a new feature, or even changing the positioning of a product. Our best advice at this stage is to go through the normal process your organisation uses to design and execute feature development, ensuring you’re prioritising issues just as you would in your normal design process. Don’t forget to keep actively involving business stakeholders too!
Step 4 – Measure
There’s no point coming this far without measuring what’s been done! Here, you’ll want to assess the performance of the new feature or functionality to understand if it is adding value. Critically, you should ensure you compare KPIs from before and after the feature update, just like Slack did in the example above.
You might want to consider these metrics as a starting point. And don’t forget – you may need to run additional research in order to understand if the change has been effective.
Step 5 – Iterate
Keep repeating steps 1 to 4! Take the findings from your original research and continue to uncover insights on UX improvements to implement and test.
As all best digital teams do, adopt a continuous research approach, allowing users to provide honest feedback on an ongoing basis to inform your UX design features and prioritisation lists.
Data-driven UX design has recently emerged as a revolutionary way in which we approach user experience. And it’s one that companies of all sizes can do. By leveraging data throughout the design process, organisations can gain invaluable insights into user behaviour, preferences, and pain points. Armed with this knowledge, designers can then create experiences that resonate deeply with their target audience, leading to increased user satisfaction, loyalty, and, of course, business success.
Adrenalin is a leading digital product and technology agency for Australia’s top brands and organisations. Stay informed about the latest digital product trends, strategies and tactics by subscribing to the Adrenalin newsletter below.
Learn from us
Join thousands of other Product Design experts who depend on Adrenalin for insights
