I've spent the last decade working as a statistician across tech and healthcare, and I'm convinced that the overuse of p-values in research is causing more harm than good. When a study reports p < 0.05 but fails to account for multiple comparisons or small sample sizes, the result often doesn't replicate—I've seen this happen in clinical trials where a promising drug later showed no effect. Statistical significance without practical significance is a trap that even seasoned analysts fall into. Should we shift the focus toward effect sizes and confidence intervals instead?