Data science job postings keep asking for a PhD plus five years of experience, yet most real-world analytics work is just cleaning messy spreadsheets and explaining p-values to stakeholders. I've hired juniors from bootcamps who outperform PhDs in practical modeling because they focus on business impact, not publication metrics. The credential inflation in data science is pricing out talented people and bloating team budgets. Should we ditch the degree requirements and hire purely on portfolio evidence and problem-solving tests?