The pointing consensus problem asks each agent in a multiagent system to agree on their headings toward a common target. This paper proposes a decentralized approach to the pointing consensus problem by simultaneously solving three smaller problems: bearing-only measurement based network localization, target decision, and heading coordination. The proposed solution guarantees that all agents' headings almost globally asymptotically target any weighted centroid of the agents' positions. Furthermore, based on this approach, two decentralized solutions for the Fermat-Weber location problem are proposed and analyzed. Simulation results are also provided to support the analysis.