The use of RFID in libraries has grown rapidly since the late 1990s, and has become a fundamental element in the design of many new major libraries around the world – from Seattle in the USA to Shenzhen in China. In some of these large libraries – Shenzhen has a stock of 2 million items, Seattle has capacity for nearly 1.5 million – this has included not only self-service but also automated stock handling. At Seattle, between 1100 and 1400 items can be dealt with per hour, returning them to the correct part of the library's ‘book spiral’, spread over four storeys.

However, the primary focus of most current implementations remains self-service, with the supply chain applications commonly found in other industries being adopted only in those countries where a centralized process is already in place. While frustrating to some observers, anxious to see the full benefits of RFID being exploited by libraries, it is perhaps understandable that library managers have concentrated on the ‘quick win’ that self-service provides. It is also possible to argue that, to some extent, libraries’ use of only a limited element of this technology's potential is no different from that of software generally.

According to the widely quoted statistics researched by the Standish Group (Johnson, 2002), only 20% of the features in enterprise software are ever used, with 19% rarely used and 45% never used. This may be partly a reflection of the trend towards software ‘bloat’, but also aligns with the familiar findings of Pareto's principle, with 80% of effects attributable to only 20% of causes.

More importantly, there are some aspects of RFID that, once taken for granted, are now being questioned. A statistic often quoted in the early days of library RFID was that tags would be able to achieve at least 100,000 cycles (or 50,000 loan and 50,000 return transactions) and so would easily outlast the items to which they were attached. More recently, however, 3M has suggested that not all tags are created equal, with some becoming inactive within a very short time, due to environmental effects (mainly heat) or simply the use of lower quality materials in construction (3M Library Systems, 2007).