Home | Code of Practice | Best practice guidance | Best Practice on Aggregating Metrics

Best Practice on Aggregating and Compiling COUNTER Metrics

Some consortia and national bodies seek to compile usage metrics from multiple institutions. These guidelines have been put together to help simplify that process and make it more consistent. A minimal template workbook is available to help you with metric aggregation.

The guidelines are generic: they do not cover every eventuality. If your consortium or national body wants to aggregate other metrics, and requires assistance in creating rules for members to follow, we are happy to assist.

This guidance was updated on 24 August 2026 and applies to Release 5.1.1 of the COUNTER Code of Practice (R5.1.1).

PDF Version

Template workbook

Principles

Use Reports, not standard views

We always recommend using the COUNTER Reports rather than the standard views. We built Release 5 from the ground up to be a powerful, flexible and comprehensive usage reporting framework. It replaced the more content-specific reports required in Release 4. To help report consumers (libraries) make the transition, we designed pre-configured filters of the COUNTER Reports that mimicked the old R4 reports, which we call standard views.

We know that some aggregation groups still request that libraries report using the standard views. This requires that libraries collect multiple standard views instead of a single Report. We think this is inefficient, and recommend that aggregation groups transition to using the appropriate Reports.

There is more information available on our website: Why not to rely on standard views.

Use Unique metrics

R5.1 reports usage through Total and Unique metrics. Total metrics reflect every interaction with content during a user session, meaning a user accessing one piece of content four times will generate four Total metrics. Our Unique usage metrics deduplicate user activity to one count per piece of content, per user session. That is, four Total_Item_Requests are deduplicated to one Unique_Item_Request. As such, we recommend using Unique metrics when aggregating usage reporting.

For more information about our metrics, visit Education: Metrics.

Aggregate annually

While COUNTER reporting typically operates on a monthly basis, this can result in a very large amount of data for aggregation groups to manage. We recommend requiring reporting for a twelve-month period.

Do not mix metric types

Different types of metric serve different purposes and possess varying levels of utility. As such, you should not sum up different metrics to reach a proxy for ‘usefulness’.

For example, Searches_Regular and Unique_Item_Investigations represent fundamentally different types of activity and value. They must be treated as distinct categories within aggregated usage reports.

Assessing book usage using the Title Report (TR)

Relevant Host_Types

You can identify these services using the Host_Type field in the COUNTER Registry.

Databases which aggregate full-text content at title level, Host_Types Aggregated_Full_Content and eBook_Collection.
eBook platforms, which are non-database platforms offering access to individual ebooks.

Configuring TR

Harvest the TR including the Attributes Access_Type and YOP, to cover the full reporting period.

Once harvested use Data_Types: Book and Reference_Work to filter the TR.

If you are aggregating metrics for monographs, and wish to exclude metrics for textbooks, encyclopedias, and similar large texts, only Data_Type: Book is relevant.

Usage Metrics

For each title within the TR, report:

Unique_Title_Investigations
Unique_Title_Requests

Reporting Title metrics allows direct comparisons across platforms, regardless of whether books are available by chapter or only as whole titles.

We recommend requiring the Access_Type and Data_Type for each book, in addition to the usage metrics. This will facilitate usage comparisons between Controlled, Open, and Free_To_Read materials, as well as highlighting usage patterns across Books and Reference_Works.

If you wish to know usage by year of publication, you will also need to require YOP in your reporting.

Assessing serial usage using the Title Report (TR)

Relevant Host_Types

While primarily relevant for journals, these recommendations also apply to conferences and newspapers or magazines. You can identify these services using the Host_Type field in the COUNTER Registry.

Databases which aggregate full-text content at title level (Host_Type Aggregated_Full_Content).
eJournal platforms, which are non-database platforms offering access to individual serials.

Configuring TR

Harvest the TR including the Attributes Access_Type, Access_Method, and YOP, to cover the full reporting period.

Once harvested use Data_Types: Journal, Conference, and Newspaper_or_Newsletter to filter the TR.

If you are aggregating metrics only for journals, and wish to exclude metrics for conference proceedings, magazines, and other serials, only Data_Type: Journal is relevant.

Usage Metrics

For each title within the TR, report:

Unique_Item_Investigations
Unique_Item_Requests

We recommend requiring the Access_Type and Data_Type for each title, in addition to the usage metrics. This will facilitate usage comparisons between Controlled, Open, and Free_To_Read materials, as well as highlighting usage patterns across journals and other serialised content.

If you wish to know usage by year of publication, you will also need to require YOP in your reporting.

Assessing database usage using the Database Report (DR)

Relevant Host_Types

You can identify these services using the Host_Type field in the COUNTER Registry.

Databases which aggregate full-text content at title level. By that, we mean that the database gives access to full-text journal or book content, or other content collected into a title. These are Host_Types Aggregated_Full_Content and eBook_Collection.
Databases which offer full-text content as individual items. By that, we mean that the database gives access to full-text or multimedia content that is not collected into a title. These are Host_Types Full_Content_Database and Multimedia_Collection.
Discovery databases, which hold metadata without any full text or multimedia content. These should not have any Request metrics to report and are Host_Types A&I_Database and Discovery_Service.

Configuring DR

Harvest the DR to cover the full reporting period.

Usage Metrics

For each database within the DR, report:

Unique_Item_Investigations
Unique_Item_Requests

Note that discovery databases will only have Unique_Item_Investigations, as there is no full-text in these services.

Search Metrics

Search metrics are less informative than the usage metrics. If you wish to include search metrics in your compilation, then for each database within the DR, report:

Searches_Regular
Searches_Automated
Searches_Federated

It is unusual but not impossible for both Searches_Regular and Searches_Automated to appear on a single database, depending on how the publisher platform is set up.

Assessing other usage with the Platform Report (PR)

Relevant Host_Types

You may wish to aggregate information about usage of other types of content. You can identify these services using the Host_Type field in the COUNTER Registry.

Research data repositories (Host_Type Data_Repository)
Platforms offering individual items of multimedia content (Host_Type Multimedia)
Institutional and subject repositories (Host_Type Repository)
Services used to share information about research (Host_Type Scholarly_Collaboration_Network).

Configuring PR

Harvest the PR to cover the full reporting period.

Usage Metrics

For each Data_Type within the PR, report:

Unique_Item_Investigations
Unique_Item_Requests

Non-Release 5.1 metrics

Principle

An extension of the principle of not mixing metric types: metrics from different types of reports must not be aggregated. That is, if you have Unique_Title_Requests from R5.1 they must not be added to book metrics from non-COUNTER reports.

Release 5 COUNTER metrics

Some publishers still offer reporting under Release 5. Before aggregating metrics from these sources, please bear in mind that they have not been audited since 2024.

The same principles apply: use reports, not standard views. Use unique metrics. Aggregate annually.

Assessing book usage using the Title Report (TR)

For each title within the TR, report:

Unique_Title_Investigations
Unique_Title_Requests

These metrics are comparable with R5.1 Unique_Title_Investigations and Unique_Title_Requests for books and reference works.

Assessing serial usage using the Title Report (TR)

For each title within the TR, report:

Unique_Item_Investigations
Unique_Item_Requests

These metrics are comparable with R5.1 Unique_Item_Investigations and Unique_Item_Requests for journals and other serialised content.

Assessing database usage using the Database Report (DR)

For each database within the DR, report:

Unique_Item_Investigations
Unique_Item_Requests

Note that discovery databases will only have Unique_Item_Investigations, as there is no full-text in these services.

Search metrics are less informative than the usage metrics. If you wish to include search metrics in your compilation, then for each database within the DR, report:

Searches_Regular
Searches_Automated
Searches_Federated

It is unusual but not impossible for both Searches_Regular and Searches_Automated to appear on a single database, depending on how the publisher platform is set up.

Assessing other usage with the Platform Report (PR)

For each Data_Type within the PR, report:

Unique_Item_Investigations
Unique_Item_Requests

Release 4 COUNTER metrics

We are aware that some publishers may still offer reporting under Release 4. We do not have records of these publishers so cannot offer support in understanding their reporting. Before aggregating metrics from these sources, please bear in mind that no R4 report has been audited since 2018. Our original guidelines for aggregation suggest the following:

For books, use Book Report 1 and the Reporting Period Total (Title Requests) metric.
For journals, use Journal Report 1 and the Reporting Period Total (Full-Text Article Requests) metric.
For databases, use Database Report 1, Regular Searches metric.
For multimedia, use Multimedia Report 1 and the Reporting Period Total (Multimedia Full Content Unit Requests) metric.

Non-COUNTER data

We have no insight into what kinds of metrics or reports may be offered by non-COUNTER publishers, and as such cannot provide recommendations.

How this best practice was developed

In 2019 COUNTER created guidelines for aggregating usage metrics, which were utilized by consortia and national bodies to standardize reporting from their member libraries. Since then the reporting landscape has changed. We consulted the community on our draft update to the guidelines in Summer 2026, and published in August that year.