AI Focus Group Feedback
Prompt
Role
Act as a focus group of three distinct members of my target audience (User A, User B, and User C). I will give each user a brief one-sentence backstory. We are looking to understand how well our app prototype meets the meets of our target audience. However, you must follow these strict rules: Each simulated user must provide objective, high-level scrutiny and never give me compliments. A simulated user must always voice when they think my implementation is lacking, and provide suggestions for improvement when they do so. Users must always cite their sources, with links and quotes if possible. Finally, all users must behave with the same level of fairness, objectivity, scrutiny, and support that a member of a thesis review council would have.
Backstories
User A is a 17-year-old name Marcus who experienced one of their favorite content creators get cancelled on the internet. They're shocked, and anxious and want to know what they did and what actually happened. User B is someone named Bartholomew, a 19-year-old who loves rhythm games and developing them, whose witnessing rumors spread about an artists past work they were inspired by as being problematic. They want to get to the bottom of this controversy to be able to continue his project with peace of mind. User C is a 18 year old girl who loves indie animations and come to find out that one of the creators has been getting a huge wave of controversy and she wants to know exactly what's going on.
Questions
Based on our problem statement, how effectively does the current implementation of our prototype address the core issue? Please provide the specific, individualized feedback of User A, B, and C. How well do the specific features of this prototype satisfy the needs and improve the daily life of our users? Do any of the three users disagree on its effectiveness? How well does our prototype actually align with our stated value proposition formula? Have the three users briefly debate the biggest barrier to adopting this prototype based on our implementation. In what specific instances or areas is this prototype lacking, and what actionable suggestions do the panel members have to improve those features?
Feedback
The prototpe seem to fail to satisy user needs, provide actual decision-grade evidence,and reduced cognitive friction or synthesize complex controversies. There's also the fact that it restricts strictly searching by influencer handle. User A says that the biggest is trust and safety as the "Hunting down an influencer's current whereabouts and other info" opton implies that the site is a doxxing tool. User B says the primary barrier is the ultility (aka the search bar). It lacks primary source integration and narrow query scope, like a name-only search, and just spouts out generic "related posts". User C said that usability and routing. For instance, if you were "I want to fact-check something I saw" and get an unfiltered search bar.
User Testing Log
The Tester and Task
Tester: MariusZ, age 18, dude who loves Persona, Sonic, and Ace Attorney.
The Task: Look up any creator that comes to mind and see if they actually show up.
Observation Notes
The user got confused when they tried to look up a different creator that wasn't in the sample of the ones that were there. Turns out there ARE certain creators it was able to look up, however they didn't show a lot of info.
Interview Synthesis
Other than what I just mentioned there weren't anything they were confused on. For the rumor button, he expected a text box asking who exactly they're looking for so that they could type information about them. For the fact checking button, they expected two text boxes, one that asked, again, who you want to fact check and and then another text box asking what you want to fact check about them. For the 'expose' button, Marius expected one text box asking who they wanted to expose and another asking what they want to expose them about. Make sure the profile pictures work and to make the sample sizes bigger. That's what they suggested.
Actionable Fixes
One of the fixes I could possibly implement is option-based pages, I could also try and change the options to fit better with the appearance of the app.
Feedback Synthesis
After looking through the feedback of both the AI Focus Group, I have concluded on adding option-based search inquiries which deals with the issue of making the sample size bigger as well as the unfiltered search bar problem. I'm also gonna change some of the option names.