With inherent support for storing and analyzing highly interconnected data, graph and RDF databases appear as natural solutions for developing linked open data applications. However, current benchmarks for these database technologies do not fully attain the desirable characteristics in industrial-strength benchmarks (e.g. relevance, verifiability, etc.) and typically do not model scenarios characterized by complex queries over skewed and highly correlated data. The Linked Data Benchmark Council (LDBC) is an EU project that aims the development of industrial-strength benchmarks for graph and RDF databases. The Social Network Benchmark (SNB) is an LDBC benchmark intended to test various functionalities of systems used for graph data management. The objective of this talk is to present the status of the SNB and to describe the problems concerning its development.