As soon as I encounter an issue, a moment of confusion, or a potential improvement (positive or negative), I immediately pause and document it. This includes taking a screenshot or video clip whenever possible, and noting the exact time and the specific page/level I am on. I never rely solely on memory. Before I report an issue, I always attempt to reproduce the problem at least one more time, ideally using a slightly different path or sequence of actions. This helps me confirm if it's a genuine bug or a one-time anomaly. If I can reproduce it, I record the exact steps to trigger the issue. I organize my final report clearly, using a consistent format for each item of feedback.

Yes, I have. I recently tested a mobile game called All in Hole, where the goal was to put fruits and vegetables into a hole using physics-based mechanics.

I was testing a search tool for a creative platform (like a stock image library) where users were looking for assets that matched a mood or a vibe, not just a literal object. The problem I noticed was that the search function was too technical. If a user typed "nostalgic warmth," the system only focused on the literal noun ("warmth") and ignored the emotional qualifier ("nostalgic"). It was prioritizing object tags over feeling tags. I conducted extensive searches using only subjective language (like "fierce energy," "hazy memory," "soft-focus dread").I demonstrated that the system was effectively deleting the human emotion from the search request. My detailed examples showed the engineering team exactly where the algorithm needed to be adjusted to prioritize subjective adjectives. This ensured the user found the right feeling—not just the right object—leading to a measurable jump in user satisfaction and interaction.