ARTIKEL

AI is taking over my job!

  • Esther Meijers - UX Designer
  • 29 september, 2023

How AI is helping me, as a UX designer, to reduce research time and gather insights quickly.   

"Who is your target audience, what do you want the target audience to do, what does your target audience see, what does your target audience do, what does your target audience say, what does your target audience hear, and what does your target audience think and feel?" are all questions you want to answer as a ux designer through qualitative and quantitative data/research. After an intensive research phase, you finally have all these answers and insights. Isn't it time for long surveys to make room for the faster AI tooling?

It's Tuesday morning and my colleague asks if I want to write an article on AI research. My first thought here is. What exactly do I think about this? As a ux designer, I prefer to research qualitative data to find out what exactly the user wants. I classify AI as quantitative data. But is this a bad thing? I decide not. After all, we need quantitative data to see how the user currently interacts with a website or app. And AI makes me understand this ten times faster! Pure time savings. I currently see AI as a very valuable colleague. 

But, how do I deploy colleague AI in the most useful way possible? Well that's a question I've been pondering for some time. What I know is that AI helps me in my research to quickly create insights and validate ideas. In this, AI is the supporting colleague. I use smart tooling that uses data from existing eye-tracking results, and then gets smarter through AI. This allows me to test for usability, focal points and accessibility (for example: https://vas.3m.com/, https://expoze.app/) at an early stage. Incidentally, I think it is important that when I use these toolings for heatmapping and improvement points, we should be transparent about this to our customers.  

In addition, I always use at least two different tools because a good research always contains multiple sources this ensures that it is more reliable, even if this is AI. Why? You show connections between the different results and your own assumptions, this gives you a more balanced picture of the situation. This also provides a fuller understanding of the problem. Neatly, they also call this triangulation research "triangulation". 

Striking results during research are for instance: use of colour, placement, use of photography and size. But sometimes, using multiple sources can also produce duplicate results. Like in a case where I tested a small difference with two banners. One banner had two colour fonts and the other one. In the one with two colour fonts, the heading stands out more than the call to action button. Logical, but so in the other banner, the call to action button stands out 15% more. Then I ask the question, "Which has higher priority? The CTA or the heading?" this I cannot decide but is a finding.   

I can't actually write very well myself, but fellow AI can help me with that too. As I am writing this article, I have written down "AI article" as a temporary title. For this, I no longer need to use "Lorem ipsum" or padding text. I asked Chat GPT (chat.openai.com) to write a catchy title. However, you should always be careful to be specific when generating texts and images. So throw "Lorem Ipsum" in the bin because AI doesn't get that (and yes, once someone asked me if this is Spanish). My college AI came up with a wonderful and specific title: "Navigating the frontier of artificial intelligence: Unravelling its promises and challenges". That promises something! 

When developing personas, fellow AI not only helps me quickly identify target groups, but more importantly, it helps me write out the scenario of how the end user will use the website. In doing so, AI takes into account demographic information, motivations, goals and preferences of these potential personas. But I cannot rely on this data alone, I still validate insights and research on human aspect, of course.  

AI is actually a very fine colleague. It helps me carry out tasks faster and more automatically, this gives me more time and space for creativity and innovation. AI helps me during work, but I still have a beer with my real colleagues.   

Side note! No, this article is not AI generated, as it is based on my views around colleague AI and how I work with it. They can't generate this piece of outspoken personal opinion yet ;). 

Deel artikel

Nog niet uitgelezen?

MEER NIEUWS

Let's work together to build powerful brands

LET'S TALK