What are the steps involved in an electronic resources audit?
Answer
Table of Contents:
Find Active Collections:
- Search Electronic Collection, Electronic Collection name and enter * in the search field to get the full list of your activated collections. Sort ascending so that it's easier to compare with your database A-Z list on your library webpages.
- If needed, export the list to an Excel file and make a column for notes.
Number and Type of Collections:
- How many collections do you have? 100-200 is reasonable for most places.
- Types of Collections:
- Search: usually has a black CZ icon. You may have collections activated only for CDI searching, meaning that search records from these collections will show up in an expanded search and help users find content to request via ILL
- Active Link Resolver: usually has a blue CZ icon. These collections have portfolios that are used for discovery in Primo VE.
- Active Link in Record: could have either black or blue CZ icon. These collections may or may not have portfolios. Ideally, they only use search records from the CDI that include the links to the content.
- Local: usually has blue local (house) icon. These collections are necessary when the content that you need isn't available in the CZ.
- Database: CZ icon. These won't give you title level content in Primo, but you can activate them in order to put orders on them, or if you want to provide a collection level link in Primo (make sure that your specific link is updated on the collection.)
Collection Activation:
- Are all the collections active? If you current subscribe to the content, you almost always want the collection to be active in Alma.
- It's ok to have a handful of collections just for search, like ERIC or ScienceDirect Journals, but most campuses don’t need many because it's only adding search records to your expanded search. Any titles activated for full text are automatically searchable.
- Check for activation dates: if the date is 11/17/19 or Feb 2020, the collection may have been activated through PCI to CDI migration, and you may need to clean some of these up by deleting them. Ex Libris activated hundreds of OA collections for each campus. Also, many of the collections that were active in the PCI as search records are now Hybrid collection, meaning that you may be getting links to content in Primo that you don't own.
- May have inactive database types just for acquisitions, which is fine
- Link in record collections may be inactive at the collection level but active for search and set to CDI-only full text
- Migration collections: you may find collections that were migrated from your previous electronic resources management system that you no longer get. Check the activation date. If it's sometime in June 2019, then it was probably migrated from EDS or another system. The data may not have been accurate in that system.
- Inactive collections: confirm that the collections are only active for search, acquisitions, or Link in Record linking. Over time, collections may have been deactivated from the CZ and left as inactive local collections in your system. You can delete these types of defunct CZ collections.
Collection Troubleshooting:
- Go through list collection by collection and look for problems:
- Local collections: does this really need to be local or could it be activated from CZ?
- Odd collection names like Bacon or KB or something other than what you’d expect – content might be active in wrong collection
- Collections with no portfolios: you don't need collections for things like Harpweek or other database types that don’t have any content or search options, unless you're using them for acquisitions or to make the database discoverable in Primo (Resource Recommender is a better way to do that).
- Collections with a single title: do you really get this title from this place? Check portfolio link to see if it works.
- Duplicate activations: you may find a collection activated just for an order, and copy of it activated for access in Primo. In this case, move the order to the active collection and delete the other one.
- Wrong or multiple versions of same content: e.g. you get psycarticles through EBSCO but have the Proquest version activated
- Obsolete packages like old Alexander Street packages that changed in Fall 2019 in CZ, will look like inactive local packages, need to activate new versions from CZ
- Open Access packages: ok to have a few but some of the old ones may no longer be fully OA, links may not work, worth testing
Collection Settings (Edit Collection):
- CDI Tab:
- CDI Search:
- eBook packages: Make sure it’s inactive for eBook packages so that you don’t get eBooks in expanded search. That content will not be requestable.
- Aggregator journal or mixed media packages: It’s good to turn search on for aggregator journal or mixed media packages – it only impacts expanded search, anything available in Alma will automatically be searchable in Primo
- We subscribe to only some titles: should be set to yes, even for aggregator packages. Otherwise availability publishes on the collection, not the titles, and will match on content in CDI that may not be part of package. Saying yes publishes availability at portfolio level and will match on CDI search records by identifier instead of by collection name. Will lead to more accurate available content.
- Do not show as Full Text available in CDI even if active in Alma: say yes to this for ebooks/videos if campus doesn’t want title level results from both CDI and Alma, also say yes to this if campus doesn’t want chapter level search results
- Link in Record:
- Just because you can use link in record with a collection, it doesn’t mean you have to. Some campuses prefer to use the portfolios and link resolver linking. The portfolios will publish as available and search records will match the identifiers and the portfolio and its link will show up as a link resolver type link on the search record.
- Link in Record linking: if collection doesn’t have portfolios, or if you choose to go with link in record linking, you need to make sure both CDI search and CDI-only full text activation are turned on. The search piece makes the records show up in an expanded search in Primo. The CDI-only full text activation makes the link in record links show and makes the search record show up in a filtered search (if they’re not already showing up because of some other activation)
- If you're going to use link in record linking, it’s good to test to be sure it works. You can find the link in record records in Primo by grabbing CDI collection id (AAC, etc.) from Collection CDI tab, search for it in Primo, after search go to URL in search bar and change “any” in search string to DBID and hit enter. This should bring up only search records with that collection id (doesn’t work 100% of the time)
- CDI Search:
Service settings (Edit Service):
- Activation Tab:
- Activate new portfolios associated with service automatically?
- No for selective packages
- Yes for aggregators.
- Sometimes you will get whole selective package and will set it to yes – you'll get an error when you save but it’s ok to leave it this way.
- Activate new portfolios associated with service automatically?
- Linking:
- Make sure proxy is active at service level, unless you're using static links with proxies, which isn't recommended. Use CZ linking whenever possible.
- Make sure you don’t have an obsolete parser override
- If you need linking parser parameters, make sure they’re in there. EBSCO now uses customer id and OPID, Gale usually needs id and YES for article, Proquest needs id, etc.
- Portfolios:
- Local portfolios: use CZ linking drop-down option and choose local to see if you have any local portfolios that could be added to the CZ, or might be obsolete and need to be deleted. For selective packages, when a title is removed from the CZ, it won’t automatically be deleted from the collection like it would for an aggregator. It also won’t automatically activate a newer version even if your campus gets that. You get notified about this kind of thing in the CZ Task Updates, or sometimes DARA.
Other things to check:
- Custom linking for Central Index databases table in Config, Discovery, Other, Custom linking for Central Index databases. Make sure your local ids are set for any content that needs one for Link in Record linking
- Stand-alone portfolios that might need to go into a collection. May be left over from migration (check for June 2019 date). Do advanced search in Electronic portfolios for Is Stand-alone = yes
- Local portfolios that could be linked to CZ. Do advanced search in Electronic portfolios for Is Local = yes
- Unsuppressed bib records that have no holdings: will show in Primo if not suppressed. Do an advanced Title search for has holdings equals no. Could also filter by linking to the NZ or not, but it doesn’t really matter, can make set and delete them in batch with delete job if it’s safe to get rid of them.
- Do you have SUSHI set up for their collections? Will let you compare usage across collections and use cost per use analysis if you use it with acquisitions.
- Link resolver: is it set up correctly in places like EBSCO, Gale, Proquest, Worldcat, etc.?
- A-Z list: do you really get what’s on the list? Is it linking correctly? Have you updated to https linking (need to change these in conjunction with proxy updates)?
- Proxy: if you’re self-hosted, are your config file stanzas up to date? Do you have everything they need, including the general DOI one? Is your proxy login screen working and accessible (run WAVE tool on it)?
- LibGuides permalinks: if you migrated to a new proxy, did you catch all the old permalinks in LibGuides, Brightspace, etc. Use LibGuides Tools, Search & Replace to find old proxy entries