The life-science domain has been discussing the merits of data integration for several years. Approximately five years ago, knowledge management promised to solve many of the bottlenecks of the drug discovery pipeline. It was envisioned that companies would purchase off-the-shelf software solutions or hire consultants or informaticians to solve their informatics problems. However, knowledge management has not delivered all that it promised. Better investment in integration is required to deliver a system that significantly reduces the time-to-market and R&D overheads, and increase the rate at which novel chemical entities (NCEs) progress through the pipeline. Such investments require strategic decisions that, if implemented successfully, deliver substantial benefits and act as the bedrock for NCE selection.