Human Designer Magic and AI

I took a quick Udemy course on Prompt Engineering for UX. There were a lot of good tips in the course. It got me thinking that we need to define how AI can allow us to focus on the creative magic we bring as experienced UX designers.

As designers, AI can automate many of our routine UX tasks and documentation. This automation frees us to concentrate on the unique creative contributions we bring to the design process.

User Personas

Automation: ChatGPT can generate detailed user personas based on inputted data such as demographics, goals, behaviors, and pain points.

Human designer magic While the initial generation can be automated, experienced designers refine these personas with deeper insights and empathy gained from direct user interactions. The magic happens when we infuse these personas with real stories and emotions, ensuring they resonate and guide our design decisions authentically.

Manual Benefit: Crafting user personas manually allows designers to immerse themselves in user research, building a deeper connection with users. This process helps create more authentic and relatable personas that truly reflect user needs and behaviors. It's in this manual crafting that we find the true essence of our work as designers, and it's what sets our designs apart.

Recommendation is to use AI for general grouping, but to manually refine and customize the results. This manual touch is what builds empathy and authenticity into our work.

User Journey Maps

Automation ChatGPT can draft user journey maps by synthesizing user data and outlining key steps, touchpoints, and emotions throughout the user experience.

Human designer magic: Designers enhance these maps by adding nuanced insights, accurate user quotes, and personal anecdotes that bring the journeys to life. The real magic is in understanding the subtleties of user emotions and motivations at each stage, crafting journeys that truly connect with users.

Manual Benefit: Creating user journey maps manually deepens designers’ understanding of the UX flows currently in production and helps capture the subtleties of user emotions and motivations.

Recommendation: Automate the general steps but manually enhance and customize user journey maps with accurate user quotes and anecdotes to ensure they reflect the authentic user experience and our audience’s specific emotional journey.

Wireframes and Mockups Descriptions

Automation ChatGPT can create descriptions for wireframes and mockups, explaining the purpose of each element and the intended user interactions.

Human designer magic: Designers review and adjust these descriptions to align perfectly with the design vision and user needs. The magic is in the details, where experienced designers ensure every element serves a purpose, the most effective UI element is used, and best practices are followed.

Manual Benefit: Writing descriptions manually ensures that designers thoroughly think through each element’s purpose and user interaction, leading to more intentional and user-centered designs.

Recommendation: Use automation to get the basic steps and frames in quickly then refine the more custom steps manually. Spend less time creating basic wire frames and more time researching just the right UI elements and micro interactions for the job.

Competitive Analysis Reports

Automation: ChatGPT can gather information on competitors and generate comprehensive analysis reports, highlighting strengths, weaknesses, opportunities, and threats.

Human designer magic: Designers interpret these reports to draw actionable insights and strategic recommendations based on their industry expertise. The magic lies in understanding the competitive landscape and leveraging this knowledge to create innovative solutions that set us apart.

Manual Benefit: Conducting competitive analysis manually allows designers to develop a deeper understanding of the competitive landscape and identify unique opportunities and threats that automated reports might miss.

Recommendation: Manually interpret and analyze competitive data to draw strategic insights and create innovative solutions that stand out.

Usability Test Scripts

Automation: ChatGPT can draft usability test scripts, including test scenarios, tasks, and questions, based on the project goals and user personas.

Human designer magic: Designers refine these scripts to ensure they effectively address the specific research questions and capture the nuances of user interactions. The magic happens when we craft scenarios that truly reflect realworld usage specific to our apps and uncover valuable insights.

Manual Benefit: Developing usability test scripts manually ensures that designers address specific research questions and capture the nuances of user interactions, leading to more effective and insightful usability tests. Refine scripts as testing proceeds based on learning from interviews.

Recommendation: Manually write scripts but use ChatGpt to gut check for bias and refine questions.

User Feedback Summaries

Automation: ChatGPT can analyze user feedback from surveys, interviews, and usability tests to generate summaries and identify common themes.

Human designer magic: Designers review and tailor these specs to ensure they accurately reflect the design intent and technical feasibility. The magic is in bridging the gap between design and development, ensuring our vision is brought to life seamlessly.

Designers dig deeper into this feedback to uncover underlying issues and craft strategic recommendations for improvement. The magic lies in transforming raw data into meaningful insights that are specific to our business and not over generalized.

Manual Benefit :Analyzing user feedback manually allows designers to uncover deeper insights and craft strategic recommendations that address the core issues and enhance the user experience.

Recommendation: Use AI to compare, categorize and do sentiment analysis, bit manually analyze user feedback to uncover deeper insights and craft strategic recommendations for design improvements.

Stay magical designer humans!

Play

In the weird and wonderful world of creativity, play is not merely a diversion but a profound enabler of innovation and authenticity. As creatives, when we allow ourselves the freedom to play—to engage in our work with the joyful abandon of a child—we unlock a part of ourselves that is unhampered by the fear of failure or the weight of perfectionism.

This playful spirit encourages us to experiment wildly, to make 'bad' things just for the sake of creation. In these moments, when we are least concerned with the outcome, we often stumble upon our most genuine and revolutionary ideas.

Brené Brown often talks about the power of vulnerability in unlocking courage and creativity, and play is the practical application of this concept. To be selectively distractable (yes, I read "Indistractable," a great book. This, however, is a small caveat, an aside, a sweet supplement for you, dear creatives), to let our minds wander among the seemingly trivial, this is where we find the golden threads of creativity.

If you face the daunting wall of creator's block, remember that the most straightforward remedy is to play. Dive into a new hobby, doodle aimlessly, write a silly poem—play. It is not time wasted; instead, it is the essential act of shaking loose the dust from our busy minds, reminding us that at the heart of all creative endeavors is the human spirit, daring to express itself without restraint.

Coincidentally, I started my day by making a goofy jingle for Wednesday. I often write weird little tunes about what I'm doing because I'm a singer and a weirdo. I posted my weird little Wednesday tune on the Design Team's private Slack channel to give my team a laugh.

They took it and ran with it. My heart is full. Filipe started the ball rolling by remixing my initial audio, then I rerecorded it with a click track, and Eder took it the rest of the way home. I was doing a weird thing. Filipe lit the spark of creativity, and Eder jumped in, making it even grander. It was a perfect start to the day and an even better end that has added extra joy to this Wednesday.

So my Wednesday advice? Make great work, but have fun out there.

"Solutioning" vs Identifying a problem

When you write a brief for a project, think about how you word it. Titles matter. You might know how to fix the issue, but do not mix up the solution in your head with the problem. They are not the same thing.

What is a solution?

“We need a modal to show people how to make tomato soup all in one pan.”

…vs. identifying a problem?

“We need to make it more obvious how to make our tomato soup in one pan because we’ve heard several complaints that our soup recipe creates too many dishes.”

We don’t necessarily need to make a modal to solve the problem. Do you see the difference?

Why does it matter?

It matters for a few reasons.

  • When you come at a problem with pre-baked solutions, you are less likely to be open to trying different approaches. Designers and engineers both get easily locked into a solution before exploring other options.

  • Depending on who you are talking to, they may take your “idea” as a “directive” rather than an idea. Especially if you are in a position of power.

  • You cut off creativity and miss out on collaboration.

  • It’s human nature not to want to discard hi-fi work (or something we’ve already started coding) even if it’s not the best solution.

  • Rushing into solutions can lead to wasted time refining high-fidelity designs or code, only to start over when we realize that the solution won’t work or better ideas surface. Staying in problem-solving mode longer saves development and design time.

I have an idea to solve it, though. Can I propose it?

Of course, we are all encouraged to propose solutions. I’m suggesting a shift in mindset and approach, which includes the following:

Shift your wording instead of

We need a modal to show people how to make tomato soup all in one pan.

...try

We need to make it more obvious how to make our tomato soup in one pan because we’ve heard several complaints that our soup recipe creates too many dishes. One idea is a model that pops up as soon as you see the recipe.

Always start with the problem we are trying to solve and frame your idea as “one idea is…”

Stay open to other ideas.

Giving great design feedback (as a non-designer)

The same rule applies here: first, identify the problem you are trying to solve.

Sound feedback:

The instructions to make the soup all in one pan doesn’t seem prominent enough. I had trouble finding them. The one-pan soup recipe is a significant revenue driver, so we must make that front and center. How might we make it more prominent? Maybe a modal that pops up?

Not as good:

Can we move the soup recipe into a modal that pops up as soon as you hit the page so users see it?

Sound feedback:

Option 2 is more successful because I immediately saw the callout box, clearly showing soup can be made in one pan.

Not as good:

I like Option 2.

Why does it matter?

If you don’t give a solid reason why you like a particular option, we may lose that part you thought was successful in future iterations. Example: A new design happens iteration on Option 2, and the callout box is now very subtle and not easily seen or gone!

To create an environment of free-flowing problem-solving, it’s best to change our mindsets on both ends: the person giving feedback and the person receiving it. This way, both parties are open to full collaboration and don’t get stuck in the proposed solution. It focuses on the problem we are trying to solve for our customers.

Super charging traditional (slow) research methods

How to Make Traditional Research Methods More Agile

Traditional research methods can sometimes feel slow and rigid. However, by adapting these methods to an agile framework, we can unlock valuable insights quickly and more iteratively. Here's a guide to how you can make user interviews, surveys, in-app surveys, and usability studies more agile while still maintaining their effectiveness.

User Interviews

Traditionally, user interviews are seen as time-consuming and are often used in the early stages of a project to gather deep, qualitative insights. However, they can be made more agile by breaking them down into smaller, more focused sessions conducted throughout the design and development phases.

Agile Adaptation:

  • Keep them short and focused: Instead of long, in-depth interviews, break them into smaller, targeted micro-studies sessions that focus on specific areas of the user experience.

  • Get comfortable with gently guiding users back to the task at hand. If a user starts to wander off into other subjects try saying- "These are amazing insights on (x) subject, do you mind if I follow up with you on that?". Then ask them to return to the task at hand.

  • No fishing expeditions! Never ask a user a question in an interview that you are unsure as to how you will use the answer. There is a time and a place for blue sky research, we're talking about making these methods more agile. If you can't verbalize how you'll use the data, leave the quesiton off.

  • Use AI tools to help synthesize, group and analyze user interview scripts and videos

  • Add qualitative questions for features that are two-three quarters away. Select 1-2 larger open questions that your team has about any intiative, ask user over the course of several studies. When the development starts you should have a “library” of information on that topic.

Surveys

Surveys, whether quantitative or qualitative, can give you attitudinal insights from a broad user base. However, they often suffer from delayed feedback loops due to the time it takes to distribute, gather, and analyze results.

Agile Adaptation:

  • Closed ended multiple choice: When trying to understand why users aren’t purchasing a product, it’s common to see a survey question with multiple-choice options and an "other" box. I recommend omitting the "other" option on the first question and only including choices that represent feature changes you're prepared to act on. This helps prioritize your roadmap. Then, follow up with, "Are there any reasons we didn’t list?" but ensure they first select an option from your initial list.

  • Quick analysis: Using ai tools for real-time data analysis can help provide actionable insights within each sprint.

Usability Testing

Usability testing is a powerful way to observe users performing tasks in a controlled setting, typically focusing on uncovering friction points.

Agile Adaptation:

  • Write clear tasks with a singular purose: For any usabilty study, ensure you ask the user one to two main questions

  • Send and play remote testing (on a budget): Write a script that details out 1-3 specific tasks, ask a user to perform specific tasks, talking aloud and recording their thoughts as they work through the tasks and send you the results.

  • Iterate within sprints: After each usability test, quickly implement fixes and retest within the same sprint. Remember, you only need 5-10 users to tell you if a UI is easy to use.

Use Diary Studies

Diary studies can provide rich, longitudinal insights, but they don’t have to take months.

Agile Adaptation:

  • Set clear, short-term goals for what you want to learn.

  • Ask users to record their interactions with your product over a week or two, focusing on specific features that are 6-8 sprints ahead of where your team is. This puts the right information at your fingertips when design/dev starts.

  • Provide easy tools (like a simple mobile app or shared document) for users to log their experiences quickly.

Permission to fail

We always need a fresh pair of eyes. It’s not a failure of the designer or dev team when they “miss something”. We are human. We are not machines. Even the most experienced designer will miss something when they are heads down trying to solve a complex problem.

Design Fails?

What is failure as a designer? Missing a margin? Missing a font size? An imperfect user experience?

I have a theory about being a designer. While there are concrete rules that govern good design (using a grid, understanding color theory, creating movement and flow, and using established typographic systems, etc), what makes a design truly great is subjective. Different people have different opinions on which designers are great and which ones are not.

An example from the fine art world: I’m not a big fan of Jackson Pollack’s work, but many consider him one of the greatest artists of all time. While I acknowledge his contribution to the art world, his work doesn’t move me personally.

The subjective nature of what makes a design great can make it hard to feel confident and not be affected by imposter syndrome. It can cause us to overly focus on perceived “failures” of missing details, as those smaller details are easier to measure and agree upon. It can give us that sinking feeling in our chest when a missed detail is pointed out.

Re-writing the narrative

I encourage us all to be the gatekeepers of our own internal dialogue around perceived “misses”. There is a freedom that comes with knowing that while the details are important and we should always strive to create an elevated experience for our users down to the smallest details, those small details are not the inherent value that we, as individuals, bring to the table. That is not the greatness. The greatness, in my humble opinion, is the thing we create together that we could not have made on our own. The greatness is releasing the non-constructive internal perfectionism of the “I” for the collaborative perfectionism of the “We”