Information counting gathers pace

Page downloads from web sites can be monitored more easily than use of a print journal could. This means that the rise of online information should be enabling librarians to discover exactly which journals and books are the most popular with their researchers and, hence, which publications to subscribe to. However, it is not that simple. Different libraries, publishers and database providers developed their own ways of counting these downloads, so until recently it was difficult to make comparisons.

What was needed was an international effort, involving vendors, librarians and intermediaries, to develop acceptable, global standards for measuring online usage. The result was Project COUNTER (Counting Online Usage of NeTworked Electronic Resources), which is now the leading standards initiative in the field.

The aim of COUNTER is to develop agreed international Codes of Practice governing the recording and exchange of online usage data for different categories of content. Release 1 of the Code of Practice for journals and databases was published in January 2003 with Release 2 following in April 2005.

The state of play today

More than 40 vendors/products are already compliant with Release 1 and several more applications are currently being processed. We estimate that, in terms of journals alone, COUNTER-compliant vendors now represent more than 60 per cent of the annual output of articles covered by the Science Citation Index. The list of COUNTER-compliant vendors is available on the COUNTER website.

But compliance with Release 1 is not the end of the story for these vendors. Release 2 contains a number of additional features (see box) and from January 2006 it will become the valid release with which vendors must comply. Until then, Release 1 remains valid.

The COUNTER initiative does not stop with journals and databases. In January 2005, the project published the draft of a new code of practice covering online books and reference works. This draft will be freely available for comment on the COUNTER website until December 2005 and the final version will be published in 2006.

This marks the first expansion of COUNTER's coverage beyond journals and databases, and is the first attempt to introduce a comprehensive industry standard for recording and reporting online book usage data. The Code of Practice for online books and reference works has been developed with input from a task force of librarians and publishers with expert knowledge of these products.

Its overall format and structure are consistent with the existing COUNTER Code of Practice for journals and databases; only the content of the usage reports has been changed and the set of definitions of terms expanded. The specifications for report delivery, data processing, auditing, and compliance are identical to those already prescribed in the Code of Practice for journals and databases.

The problem of consistency

One of the main challenges we faced in developing this draft Code of Practice was the lack of consistency among publishers in the ways in which they define, structure and distribute online books. In the case of online journals there was a wide consensus that the most important content unit to measure usage of is the full-text article. Even before COUNTER, most journal publishers were measuring downloads of full-text journal articles. COUNTER's main role was to ensure that they all did so using the same standards and protocols.

For books there is no such consensus. Some publishers make online books available only as a single file that can be downloaded in its entirety, with no further vendor monitoring of usage being possible. Other publishers allow the downloading of individual chapters or entries, such as dictionary definitions or chemical structures. We felt it was appropriate to cover both these scenarios in the draft Code of Practice and this is reflected in the usage reports.

The draft COUNTER Code of Practice for Books and Reference Works provides a set of five basic usage reports that cover full-text requests for a whole title, as well as for sections (chapters, encyclopaedia entries) within a title.

The reports specified are:

• Book Report 1: Number of successful title requests by month and title;
• Book Report 2: Number of successful section requests by month and title;
• Book Report 3: Number of turnaways by month and title;
• Book Report 4: Total searches and sessions by month and title; and
• Book Report 5: Total searches and sessions by month and service.

The report formats, data processing guidelines and delivery protocols are exactly the same as those already in use for journals and databases. Likewise, searches, sessions and turnaways have been defined in the same way as for journals and databases and the usage reports relating to these (3, 4 and 5 above) are parallel to those for journals and databases. Book Report 1 and Book Report 2 contain the key new elements on the draft Code of Practice.

The first of these, Book Report 1, is designed to provide usage statistics for those titles that can only be downloaded in their entirety and for which online usage cannot be monitored further by the vendor. The second, Book Report 2, is designed to allow the vendor to measure usage of individual sections within a title. A 'section' is the next level of organisational structure below the complete title, such as 'chapter' or 'entry'. We felt that this single report, based on the more comprehensive definition, was more appropriate than having separate reports for 'chapter' or 'entry' because a title whose structure is based on 'chapters' is unlikely to contain 'entries' and vice versa.

This latest code also defines the protocols to be observed when an aggregator or gateway is involved in the delivery of vendor content to the customer. These protocols are particularly important to avoid duplicate counting of usage by publisher and aggregator in situations where an intermediary aggregator or gateway is involved.

Compliance with the Code

As with journals and databases, compliance with this latest Code of Practice will be encouraged in two ways. Firstly, customers will be urged to include a clause in all relevant licence agreements specifying that vendors provide COUNTER-compliant usage statistics. A standard form of words for this clause is provided in the Code of Practice.

Secondly, to obtain 'COUNTER-compliant' status for their usage reports, vendors will be required to sign a formal declaration of COUNTER compliance. They must also allow COUNTER to review those of their usage reports that they claim are compliant. These reports will then be listed in the register of vendors on the COUNTER website. Only reports listed there may be regarded as being COUNTER compliant.

Report delivery will have to conform to the following standards:

• Reports must be provided either as a CSV file, as a Microsoft Excel file, or as a file that can be easily imported into Microsoft Excel;
• Reports should be made available on a password-controlled website (accompanied by an email alert when data is updated);
• Reports must be provided monthly;
• Data must be updated within four weeks of the end of the reporting period; and
• All of last calendar year's data and that of this calendar year to date must be supplied.

In 2005 more than 200 organisations have applied for and have been accepted into COUNTER membership. As COUNTER's only source of financial support is its members, these new organisations are important - not only for promoting standards in measuring online usage but also in funding further standards development.

Measuring use of journals and databases: the basics

Release 1 of the Code of Practice for Journals and Databases was published in January 2003. It contains seven usage reports, covering journals and databases. These have been divided into Level 1 and Level 2 reports:

• Journal Report 1: number of full-text article requests by month and journal (Level 1)
• Journal Report 2: turnaways by month and journal (this report is only applicable where the user access model is based on a maximum number of concurrent users). (Level 1)
• Journal Report 3: number of item requests by month, journal and page type (Level 2)
• Journal Report 4: total searches by month and collection (Level 2)
• Database Report 1: total searches, sessions and full-text requests by month and database (Level 1)
• Database Report 2: turnaways by month and database (Level 1)
• Database Report 3: referrals by aggregator or gateway (Level 1)

Only Level 1 reports are required for COUNTER compliance, but those vendors who can provide the more detailed Level 2 reports are welcome to do so, and also encouraged to use the COUNTER definitions in any other usage reports they may provide to specific customers. Journal Report 1 is proving to be the most widely implemented of the usage reports.

The latest release, Release 2, builds on the work done in Release 1 but contains the following new features:

• A modified Journal Report 1, which, in addition to the data provided in Release 1, now also reports total usage statistics for HTML and PDF full-text requests separately. This is the only modification to the set of reports already contained in Release 1.
• Examples of the required usage reports in Excel and CSV formats, along with the detailed display rules for each report, in order to facilitate compliance. In addition XML DTDs are also provided for those vendors who wish to provide the usage reports in XML (an optional addition to the already required formats)
• Specifications for the usage reports to be provided at consortium level.
• The protocols to be used for recording and reporting usage when an intermediary aggregator or gateway is involved have been collected together in a new Table. The objective of these protocols is to avoid duplication of counting by the publisher that owns the content and the aggregator/gateway that provides access to it.

An important new feature of Release 2 is that compliant vendors will have to be audited on a regular basis in order to maintain their COUNTER-compliant status. To facilitate this, a set of auditing guidelines and test scripts has been published as Appendix E to Release 2. In developing these standards COUNTER has tried to meet the need of customers for credible usage statistics without placing an undue administrative or financial burden on vendors. For this reason, audits will be conducted online using the test scripts included in the auditing standards and procedures document.

Defining the terms

The original Code of Practice for Journals and Databases contains an extensive list of data elements and other terms used in the usage reports and other parts of the code. Where possible, existing definitions from standards organisations such as NISO, ISO and ARL have been used. Among the terms defined are: 'vendor'; 'aggregator'; 'search'; 'item request'; 'consortium'; and 'consortium member'.

This comprehensive list of definitions is proving to be a useful industry resource and is becoming more and more widely used for purposes that are not directly related to COUNTER.

With the latest Code of Practice the list of terms has been expanded to cover books and reference works. As with journals and databases, where an appropriate existing definition exists, this has been used and the source, such as NISO (the National Information Standards Organization) given. Other definitions have been developed by the books task force, using a number of sources.

New definitions include:

• Book: 'A nonserial printed publication of any length bound in hard or soft covers or in loose-leaf format. Also called monograph' (NISO);
• Chapter: 'A subdivision of a book or of some categories of reference work; usually numbered and titled';
• Entry: 'A record of information in some categories of reference work (eg a dictionary definition)';
• Reference work: 'An authoritative source of information about a subject: used to find quick answers to questions'; and
• Section: 'A subdivision of a book or reference work (eg. chapter, entry)'.

Further information

• www.projectCounter.org
• An account of the early development of COUNTER can be found in Shepherd, PT, Learned Publishing 16 (2003), pp201.