What are ISBN Clusters?
Many LibraryData reports display title-level data in broken out by ISBN clusters. This differs from SalesData, which reports data at a per-ISBN level. This page will help you understand why ISBN Clusters are used, and answer some common questions about ISBN clusters!
Why does LibraryData use ISBN clusters?
Libraries report their collection data to us broken out by their Catalog ID or record. Each record can contain multiple ISBNs, often all of the editions of a single work. When you use the library, you put a hold on a copy of Pride and Prejudice, for example, and you usually don't mind which particular edition or ISBN you receive - just that you get the first available copy! So in the library's ILS (Integrated Library System), the circulation and holdings data is counted against the record.
Each library includes different ISBNs in their record for the same title, often depending on which editions they've had in their collection, their data sources, etc. So the ISBN cluster for a particular title can differ from library to library. When you look at a multi-library aggregate, like the ALL Libraries, we merge clusters together to create the most inclusive cluster based on common ISBNs.
For example, if we're looking at the cluster for Pride and Prejudice, you can see that libraries 1, 2, and 3 each have different ISBNs in their record. Library 1 has one ISBN in common with Library 2, and Library 2 has one ISBN in common with Library 3. Those common ISBNs are used to create the ALL Libraries ISBN cluster for this title, pulling in all of the ISBNs from these three library records. As additional libraries join the ALL Libraries Panel, the ALL Libraries ISBN cluster will grow based on the common ISBNs in the libraries' records.
| Library 1 | Library 2 | Library 3 | ALL Libraries |
|---|---|---|---|
| 9780141199078 | 9780141439518 | 9780141439518 | 9780141439518 |
| 9780755331468 | 9780679783268 | 9780679783268 | |
| 9781593083243 | 9780141199078 | ||
| 9781904633013 | 9780755331468 | ||
| 9781551110288 | 9781593083243 | ||
| 9780679783268 | 9781904633013 | ||
| 9780679601685 | 9781551110288 | ||
| 9780679405429 | 9780679601685 | ||
| 9780307290922 | 9780679405429 | ||
| 9781853268335 | 9780307290922 | ||
| 9781853268335 |
What do ISBN clusters include?
Often ISBN clusters include different editions of the same work. These can include different formats, like a paperback, hardcover, and mass-market edition, or editions released by different publishers. Libraries don't necessarily own copies of each ISBN in the ISBN cluster — old ISBNs may have been weeded from the collection, or other ISBNs may have been added via bibliographic data imports to a library's ILS. This is why you may also see NCR editions' ISBNs listed within a cluster.
Why am I seeing an ISBN cluster containing multiple works?
Some ISBN clusters may contain ISBNs for different but related works. For example, occasionally all of the titles in a series or set may be clustered as a single work. Libraries can catalog ISBNs in a single record for many different reasons, and LibraryData displays the ISBN clusters based on the data delivered to us by libraries. Human error is always a possibility!
If you believe we should review an ISBN cluster and work with a library on adjusting their catalog record, you can submit feedback on a cluster by clicking the "What are ISBN Clusters?" button on the Book Activity results for any cluster.
Individual library clusters and the ALL Libraries cluster
When you're viewing data for a single library (for example, looking at Marigold Library System ALL), the data you see represented in the clusters will be according to their ILS. (The only exception being that if one ISBN is present in multiple records, those records are combined.) If you are viewing data for multiple libraries (for example, in a Collection Gaps: Compare to Library report) the data will be reported using the ALL Libraries cluster, to ensure that the comparisons are apples-to-apples and the circulation data represents the same group of ISBNs.
An information box on the results page of all reports will tell you which aggregate's ISBN cluster you're viewing.
Why am I seeing so many different clusters in the search results?
Queries via the search bar will look for any clusters matching your search term in the title, series, subtitle, and contributor name fields of the bibliographic data. It's also the case that there can be multiple clusters for the same work, if there are no ISBNs in common between multiple libraries' records for the work. The results are ranked by the most active cluster in the most recent week.
How do ISBN clusters work in SalesData?
SalesData reports (the Bestseller, Single ISBN, Multi ISBN, Market Share, Market Comp, and Gap Analysis) do not return ISBN cluster data — just sales per distinct ISBN. In LibraryData reports like the Popular Books report, the All Market ISBN Cluster sales column adds together the sales for each ISBN in the cluster for the selected reporting period — so that, again, you're comparing apples-to-apples.
Do you have more questions about ISBN clusters? Email us at librarydata@booknetcanada.ca for help!