Would independent AI agent benchmarks influence your buying decision?

If an AI vendor provided third-party benchmark results showing performance on real-world tasks, would that impact your decision? Why or why not?