For startups, the question isn’t whether to pivot, but how to do it in a way that improves performance. Many founders make changes based on customer feedback, hoping to stay relevant. But without a clear strategy, these shifts often amount to little more than tweaks that fail to unlock real value.

A better approach is to pivot with purpose. Consider the case of X (formerly Twitter). It began as Odeo, a podcasting startup. But when Apple’s iTunes moved into the space, Odeo’s founders didn’t just adjust their model—they made a bold shift to microblogging. The result changed the internet.

When to pivot

New research from the University of Maryland’s Robert H. Smith School of Business explores what makes a pivot successful. Using AI, the researchers analyzed how entrepreneurs decide to change course and why it matters. Their findings point to two common mistakes: acting without a clear reason or overthinking to the point of inaction. The most effective pivots, they found, come from founders who clearly define their goals and test their ideas through experiments.

The study combines AI-driven analysis with business case studies, drawing from 1,600 interviews with 261 London-based entrepreneurs. It highlights two key ingredients of a strong pivot: “theorization” (explaining why the change makes sense) and “experimentation” (testing assumptions).

Entrepreneurial support

Beyond theory, the research has produced a publicly available AI-generated dictionary of business terms and decision-making tools. This could help entrepreneurs, investors, and policymakers refine startup training programs. Institutions like the National Science Foundation and the National Institutes of Health, which fund business research, may also find the insights useful.

The takeaway? A pivot shouldn’t be a desperate reaction or an aimless tweak. The best founders make changes that are deliberate, well-reasoned, and backed by evidence.