A project should ideally move you toward the kind of role you want. If you want software work, build systems that require the technologies and engineering practices used in those roles. If you want digital marketing, choose work that lets you demonstrate research, execution, and measurement.
Difficulty matters too. A project that is slightly beyond your current ability forces growth, while a project that simply repeats what you already know adds little new evidence.
Look for projects that produce visible outputs and meaningful decisions. You want enough complexity to discuss challenges, tradeoffs, and learning.
Career-building projects should form a progression. Each one should make the next opportunity more realistic than it was before.