Design small experiments to discover your best methods and create original results

Early in my career, I tried to learn by doing everything at once: videos, articles, full projects, and random tips from forums. Progress was slow and noisy. Then I started running tiny experiments. One week I replaced re‑reading with five‑minute free recall after each study block. Another week I held coding time constant but switched from long tutorials to building tiny features end‑to‑end. The changes were small, but the signals were clear. My recall score jumped with free recall. My time‑to‑ship dropped when I stopped binging tutorials. The coffee on my desk stayed warm because I wasn’t grinding for hours without finishing anything.

I once spent a week doing design drills in grayscale only. No color, no trendy fonts, just spacing, alignment, and hierarchy. It felt boring on day one. On day four, a friend looked at my mockup and said, “It just feels cleaner.” That was the signal I needed to keep parts of the drill in my routine. I might be wrong, but most breakthroughs are little changes that remove a hidden drag, not a big new tool.

Small experiments also help you find your style. After copying a few photographers’ compositions for a week each, you start to see what feels like you. You keep one person’s negative space, another’s rhythm of lines, and you toss the rest. The experiments make the choice obvious because you’ve felt the method in your hands, not just read about it.

This approach borrows from the scientific method: vary one factor, hold others steady, and measure. It also fights the fixed‑mindset trap by assuming methods are improvable and personal. Over time, the habit of weekly experiments compounds into a toolkit that fits you so well it looks like talent from the outside.

Choose one variable to test this week—maybe you’ll swap re‑reading for recall, or replace color design with grayscale drills—and keep your time, topic, and goals steady so the signal is clean. Track a simple metric like recall score, error count, or time to complete, then make a call at week’s end to keep, tweak, or toss the change. Queue another tiny experiment for next week so improvement never depends on a lucky tip. Put the experiment on your calendar and set a five‑minute Friday review to decide.