Data-driven design provides measurable, essential insights that support designers in their decisions by helping them better understand their audience, objectively. Data shows user patterns, pain points and new opportunities. As UX designers, we have more of an inclination to focus on collaboration and qualitative research, and might struggle to fit within the realm of data-obsessed businesses.
One case study of data-driven ROI took place in 2014, when airline Virgin America carried out A/B testing as part of its process to design a new website. This led to a 14% increase in CR, a 20% drop in customer call centre load and a wider spectrum of devices being used to book flights at almost twice the speed.
“Designing with Data” by King, Churchill, & Tan shows 3 types of methods that are influenced by data. A company can be:
So… what is the difference?
A data driven company takes the narrow path, wherein it tinkers with processes to improve efficiency and performance, its decisions very much based on quantitative data and going into the fine details.
Data informed businesses go beyond quantitative data, and their decision-making also takes into consideration other factors such as experience or instinct. A/B testing different experiences or conducting structured usability tests are both carried out in this sphere.
Finally, those following a data aware process keep the strengths and limitations of data collection in mind while looking at each problem individually to make decisions, therefore finding equal value in multiple testing & data collection solutions.
How to put this into practice
1. Use Visual Aid: It is proven that when presenting data in a visual manner viewers tend to extract more value from it. As we all know, convincing stakeholders is key in a project, and this method works.
2. Establish realistic goals: Being realistic when planning a project or product is essential not only for when presenting it to stakeholders, but also to actually get stuff done, on time and within budget. It is a skill very much honed mostly with experience, and knowing whether that feature should be on the drawing board or backlogged to an iteration phase could be the difference between launching the product or losing it.
3. Choose the most suitable design data approach for your project
4. Create a UX Hypothesis: This statement should define the sample user group you are testing, the variable, the expected effect, the reason for such expectations and what kind of result will be measured. An example: “If buttons with round corners are perceived by users as ‘friendlier’ to click, then changing the sharp corners of the “subscribe” button on our website to round will increase the number of subscriptions through our homepage.” Furthermore, as a best practice tip, aim to add the learnings from failure and/or success and to…
5. …. calculate the amount of work (time and money) is deemed worth investing in this. The right questions to ask here are: How big is the issue we are trying to solve? Will it be an upgrade of the subscription experience that we are currently offering? Or are we aiming for the best subscription experience offered out there? There is a big leap in time and budget between those two targets, yet it is annoyingly easy to overlook, only to realise when it is too late.
6. Ensure you have the defined user sample available and test only 1 variable at a time.
7. Specify whether you are going to collect qualitative or quantitative data – which will in turn define the kind of testing/data collection methodologies you will use.
Being realistic when planning a project or product is essential not only for when presenting it to stakeholders, but also to actually get stuff done, on time and within budget.Tweet
In conjunction with the above, it is essential to mention the importance of always taking a step back and looking at the big picture. Data analysis often narrows down our perspective to a specific point of reality, with the possible consequence of forgetting the other variables at play. Stakeholders could strongly lean in favour to seeing a lot of data before committing: this is because data has that wow factor which provides measurement and a sense of tangibility that the human mind constantly searches for to make sense of what is seeing. As UX/UI designers we should adapt to data-heavy processes, as it helps uncover and provide insight into underlying issues, and evaluate which solution is to them is the most effective. However, this should only complement – and not replace – the intuitive and creative edge that drives our reasoning, leading us to what the problem is and how it should be solved; so let’s keep this indispensable tool close and use it for the greater good, while keeping the users a lot closer.