Are we talking generic campaign-level targeting, or platform-level audience selection?
For the former, you absolutely need a mental model of who you're after. Even if the KPI is raw awareness, the messaging still has to land. Describe the sort of person who'd care about your brand - what do they wear, what music do they listen to, how do they make decisions? A Taylor Swift fan and a Billie Eilish fan might both be Gen Z, but their values, purchase triggers, and attention patterns differ significantly. If you don't segment at least psychographically, your creative bleeds into a generic void.
I usually build a quick audience matrix in a spreadsheet - income bracket, price sensitivity, brand loyalty, platform of choice. Then I map the copy tone to that persona. For example, budget-conscious students vs brand-conscious young professionals need completely different hooks, even if both are 18-24.
If you're asking about platform-level segmentation (e.g., one campaign for everyone vs separate ad sets by age/gender/interests), the short answer is no, you don't have to. But you'll waste budget. I've run tests where a single "everyone 13-24" campaign had a CPM 40% higher than a split into 13-17 and 18-24, purely because the younger cohort has cheaper inventory. That's not a targeting insight - it's just math.
Happy to dig into the long answer if needed - there's a Python script I use for cluster analysis on platform audience overlap that might be relevant.