"Just find some comps" sounds like the easy part of a valuation — pick a few similar companies, average their multiples, apply it to the company you're analyzing. In practice, the peer set you choose can move the resulting valuation as much as, or more than, the multiple itself. Building one you can actually defend takes real discipline, not just a shared industry label.
Start broad — same sector, similar size range — then narrow using the criteria that actually drive value: similar business model, not just similar products, similar end markets and customers, similar growth stage, similar margin structure, and comparable geographic footprint. Two companies can share a sector classification and still be poor comps for each other — a capital-light software company and a capital-intensive semiconductor manufacturer both get labeled "technology," but their margins, growth drivers, and appropriate multiples have almost nothing in common.
Correct moves you up, wrong moves you down — reach 100 to master this lesson.