The question “ISBN or PID?” begins with a category error. An ISBN is one specific international identifier for monographic book products. A persistent identifier, or PID, is a broader concept: a durable reference used to identify a resource or entity across systems and over time. DOI, Handle, ARK, URN, ORCID iD, and ROR ID are examples of PID systems with different targets and governance.
A book does not need a mysterious second number called “the PID” in every case. It may use an ISBN for each edition and format, a DOI for a book or chapter record, a repository Handle or URN for an archived version, ORCID iDs for contributors, a ROR ID for an affiliated institution, and an ISSN for a continuing series. Each identifier answers a different question.
Amazon’s ASIN and an internal SKU are product identifiers inside particular systems. They can remain stable for operational reasons, but scholarly infrastructure does not treat every platform ID as equivalent to an openly governed, resolvable PID. Define the acronym and the namespace before comparing it with ISBN.
|
Dimension |
ISBN | Persistent identifier category | DOI example | Scope | Resolution |
Main use |
| Meaning | International Standard Book Number | Long-lasting reference to resource/entity | A PID system | Specific vs broad category | Varies by system | Product vs durable identity |
| Target | Book edition/format | Object, person, organization, sample, etc. | Registered object | Product-specific / system-specific | doi.org | Trade/catalog/citation |
| Format | 13 digits | Namespace-specific | 10.prefix/suffix | Standard syntax | Often URI/URL form | Interoperability |
| Issuer | National/regional ISBN agency | PID authority/registry | Registration agency/member | Governed assignment | Registry/resolver | Trust and uniqueness |
| Metadata | Distributed across book systems | Registry/record metadata | Deposited/maintained metadata | Variable | Linked landing page | Discovery and context |
| Persistence duty | Never reuse after publication | Governance + maintenance | Registrant updates URL/metadata | Policy-based | Active resolution | Long-term usability |
| Granularity | Edition and format | Any defined entity/object | Book, chapter, article, data, etc. | Product vs chosen object | One registered target | Avoid level confusion |
| Barcode | EAN-13 possible | Not inherently | Not retail barcode | Physical product option | Not barcode purpose | Retail scanning |
| Required for book? | Channel-dependent | No single universal PID required | Optional/workflow-dependent | Use-case driven | Only if registered | Choose purposefully |
ISBN stands for International Standard Book Number. It is a 13-digit identifier assigned through national or regional ISBN agencies to eligible monographic publications. It distinguishes a specific product by publisher, title, edition, and format. Paperback, hardcover, translation, and substantially revised editions receive different ISBNs when they are separate products.
The ISBN supports book-trade ordering, cataloging, metadata exchange, inventory, listings, and product-level reporting. It can be encoded in an EAN-13 barcode for physical retail. The number is persistent in the important sense that it must not be reused after publication, but ISBN is governed as a product-identification standard rather than a general web-resolution service.
An ISBN does not automatically resolve to one authoritative URL. Metadata can exist in agency, distributor, retailer, library, and publisher systems. The product may move websites or go out of print while the ISBN continues to identify the historical edition.
OpenAIRE defines a PID as a long-lasting reference to a resource. The resource can be a publication, dataset, person, sample, funding body, coordinates, report, or software. A PID can connect to a metadata record rather than directly to the object. The goal is reliable identification, verification, and location over time.
Persistence is not created by a clever string alone. It depends on governance, a maintained registry or resolver, organizational commitments, metadata, policies, and participants who update the current location. A DOI continues to resolve after a website moves only when the record owner updates the URL.
Crossref explains that members register DOIs and metadata, then maintain the record long-term. The DOI stays with the work while its associated metadata and URL can be updated. Crossref supports books, chapters, reference works, conference proceedings, datasets, dissertations, journals, articles, reports, preprints, peer reviews, grants, and other research objects.
|
Identifier |
Expands to | Identifies | Typical namespace | Resolver/open reach | Book relationship |
Do not use for |
| ISBN | International Standard Book Number | Edition and format | 978/979 ISBN | Searchable; not single canonical resolver | Primary book-product ID | Author/institution |
| DOI | Digital Object Identifier | Registered research/content object | 10.xxxx/… | https://doi.org/ | Book/chapter/article/data | Retail format distinction |
| Handle | Handle identifier | Repository/digital object | hdl: / handle URL | Handle resolver | Deposited version | Copyright proof |
| ARK | Archival Resource Key | Stewarded object | ark:/… | Naming-authority dependent | Archive/library object | Retail inventory |
| URN | Uniform Resource Name | Resource in a namespace | urn:… | Namespace-dependent | National/repository object | Universal web URL guarantee |
| ORCID iD | Open Researcher and Contributor ID | Person | 0000-… | https://orcid.org/ | Contributor identity | Book product |
| ROR ID | Research Organization Registry ID | Research organization | https://ror.org/… | Open registry | Affiliation metadata | Publisher product |
| ISSN | International Standard Serial Number | Continuing title/medium | 1234-567X | ISSN Portal/records | Series/journal level | Individual book format |
| ASIN | Amazon Standard Identification Number | Amazon catalog item | Amazon-defined | Amazon only | Coexists with ISBN | Open scholarly identity |
| SKU | Stock Keeping Unit | Seller-defined stock item | Seller-defined | Internal | Operational mapping | Global identification |
| Library item barcode | Local piece designation | One physical copy | Library-defined | Local ILS | Copy under ISBN record | Edition-level ID |
A Digital Object Identifier identifies a registered object and can be resolved through doi.org. Publishers and repositories can assign DOIs at book, chapter, article, dataset, report, review, or other supported levels. DOI does not mean “digital-only,” prove peer review, or guarantee that the content will remain free.
The Handle System provides persistent identifiers and resolution infrastructure. DOI uses Handle technology but adds its own governance, registration agencies, metadata, and policies. Many repositories use Handles for deposited items.
Archival Resource Keys are designed for persistent identification and can be used by archives, libraries, museums, and other stewards. An ARK’s usefulness depends on the naming authority and its persistence commitments.
A Uniform Resource Name identifies a resource within a defined namespace. National libraries and repositories may use URNs for publications. A URN can coexist with ISBN when the systems identify different aspects or provide a resolvable preservation record.
ORCID provides unique persistent identifiers for researchers and contributors. It identifies a person, not a book. Authors can connect works and affiliations to an ORCID record under their privacy and permissions settings.
The Research Organization Registry provides open persistent identifiers for research organizations. ROR helps disambiguate institutions and connect them to researchers and outputs. It is not a replacement for ISBN or ORCID.
ASIN is Amazon’s catalog identifier. A Kindle eBook on KDP can publish without an ISBN and receives an ASIN. A print book on Amazon can have both an ISBN and an ASIN because Amazon manages the catalog item while ISBN identifies the book product across the broader book supply chain.
A SKU is chosen by a seller or inventory system. Different sellers can use different SKUs for the same ISBN, and one SKU can represent a bundle that includes several products. A database may label its primary key “PID” for product ID, publication ID, or process ID; that local field is not automatically a persistent identifier outside the system.
Always document the namespace and owner: ISBN 978…, DOI 10.…, ORCID 0000-…, ROR https://ror.org/…, ASIN, repository Handle, or vendor SKU. An unlabeled value called PID is not interoperable.
A book usually needs an ISBN when it will move through ISBN-based book-trade, library, or bibliographic workflows. It may also benefit from a DOI or repository PID when the publisher wants a persistent web link, scholarly citation metadata, chapter-level identification, research-object relationships, usage services, or repository preservation.
The answer is not “in most cases, yes.” A commercial paperback can operate with ISBN and retailer identifiers only. An open-access scholarly monograph may use ISBN plus book and chapter DOIs. A repository manuscript may receive a Handle or DOI without becoming a formally published book or needing an ISBN. Match identifiers to entities and workflows.
|
Scenario |
Book/product ISBN | Additional PID/ID | Why | Maintenance owner |
Do both? |
| Trade paperback | ISBN | Retailer ASIN/SKU | Trade + platform operations | Publisher/retailer | ISBN plus platform IDs |
| Kindle eBook | No ISBN required by KDP | ASIN | Amazon catalog | Amazon/account holder | DOI usually unnecessary |
| OA academic monograph | ISBNs by product | Book/chapter DOI; ORCID/ROR | Citation and research graph | Publisher/member | Often yes |
| Repository manuscript | Usually no new ISBN | Handle/DOI/URN | Identify deposited version | Repository | PID may suffice |
| Monographic series | Volume ISBN | Series ISSN; optional DOI | Different hierarchy levels | Publisher/centres | Yes by level |
| Research dataset supplement | Book ISBN unchanged | Dataset DOI | Dataset is separate object | Repository/DataCite member | Yes for relationships |
| Library copy | Shared ISBN | Local item barcode/control number | Bibliographic vs copy levels | Library | Both, different purposes |
A novel is released as paperback and Kindle. The paperback has an ISBN for trade and library identification plus an Amazon ASIN in Amazon’s catalog. The Kindle edition has an ASIN and does not require an ISBN on KDP. No DOI is needed unless the publisher has a separate scholarly or persistent-linking reason to register one.
A university press publishes hardcover, paperback, and open-access EPUB/PDF products. It assigns ISBNs to the products under the applicable rules. It registers a DOI for the book landing page and may register chapter DOIs. Contributor metadata includes ORCID iDs, affiliations can use ROR IDs, and the series can have an ISSN. Each identifier targets a level of the scholarly record.
A researcher deposits an accepted manuscript in a university repository. The repository assigns a Handle or DOI and records the version, rights, authors, and related published article. The manuscript may cite the published book’s ISBN or article DOI but does not automatically receive an ISBN merely because it is cataloged and downloadable.
|
Entity/level |
Recommended identifier | Example | Why | Related identifier | Relationship |
Common error |
| Creative work | Work/authority ID if used | Catalog work entity | Cluster expressions | Product ISBN | has manifestation | Treating ISBN as work ID |
| Book product | ISBN | Paperback edition | Ordering/cataloging | Book DOI | is version/format | One ISBN for all formats |
| Book landing page/object | DOI if registered | 10.x/book | Persistent resolution | ISBN | identifies product | DOI per temporary file |
| Chapter | DOI if registered | 10.x/chapter | Component citation | Book ISBN/DOI | is part of | Giving chapter ISBN |
| Contributor | ORCID iD | 0000-… | Name disambiguation | DOI/ISBN metadata | contributed to | Using book ID for person |
| Organization | ROR ID | https://ror.org/… | Affiliation disambiguation | ORCID/DOI | affiliated with | Using imprint string only |
| Series | ISSN | 1234-567X | Continuing-title identity | Volume ISBNs | contains | One ISBN for series |
| Physical copy | Local item barcode | A148… | Circulation/status | ISBN | copy of | Using retail barcode as item ID |
ISBN granularity is product-oriented. Different editions and formats receive different ISBNs when the ISBN rules treat them as distinct products. DOI granularity is registration-policy-oriented: a publisher can register a DOI for the whole book, individual chapters, or both, then declare relationships in metadata.
A DOI should not be minted for every file merely because files exist. PDF and HTML can be representations of the same registered content. A repository may create a separate PID for a deposited version because it is a distinct managed object. Version labels and relationships—version of record, accepted manuscript, translation, correction, supplement—must be explicit.
Identifiers should not be used to hide change. A substantially revised book needs a new ISBN. A corrected DOI landing page normally keeps the DOI and updates metadata. A retracted or withdrawn object should retain its PID and resolve to a status page rather than disappearing or being reassigned.
|
Maintenance task |
ISBN record | DOI/PID record | Frequency | Owner |
Failure signal |
| Assign once | One product only | One defined object/entity | At creation | Publisher/registry | Duplicate/reused ID |
| Validate syntax | Check digit/hyphenation | Namespace and resolver | Before release | Publisher/system | Invalid identifier |
| Complete metadata | Title, format, publisher, etc. | Title, contributors, relationships, license | At deposit + updates | Record owner | Thin/disconnected record |
| Update location | Channel metadata as needed | Resolver destination | Whenever URL moves | Registrant/member | 404/broken resolve |
| Update status | Available/out of print | Correction/retraction/withdrawal | On change | Publisher/repository | Misleading access |
| Version relationships | New ISBN for material product change | Relate versions/objects | On new version | Publisher/member | Unclear record of version |
| Monitor | Listing/catalog checks | Resolution and metadata tests | Scheduled | Named steward | Silent drift |
| Preserve | Not provided by ISBN | Separate archive/repository | Ongoing | Institution/publisher | PID resolves but file lost |
| Never reassign | Published ISBN permanent | PID tied to original entity | Permanent | Governance owner | Identity collision |
For a DOI or Handle, update the registered URL when the landing page moves. Use a durable landing page that contains citation metadata, access status, version, rights, and links to formats—not a temporary PDF URL.
A persistent link to thin or incorrect metadata is not enough. Maintain titles, contributors, ORCID iDs, affiliations, ROR IDs, dates, publisher, abstract, license, funding, references, relationships, and status when supported.
Connect a book to chapters, a chapter to a correction, a repository version to the version of record, a translation to the source, and an output to datasets or grants. Avoid relying on title text to infer these relationships.
Resolution and preservation are separate. A DOI can point to a landing page even if the underlying file is not archived. Long-term access requires repositories, preservation services, backups, format migration, rights, and organizational continuity.
A platform can use PID internally for product ID, but the research and open-science meaning is persistent identifier. State the expansion every time and name the system.
ASIN is a valuable Amazon identifier, but it is proprietary and platform-scoped. Do not equate it with DOI, ORCID, ROR, Handle, or another governed PID merely because the value is stable inside Amazon.
ISBNs are permanent product identifiers and cannot be reused after publication. The useful distinction is that ISBN is not normally a web resolver with a single maintained destination. Avoid a simplistic persistent-versus-nonpersistent binary.
Identifier stacks should be purposeful. Register a DOI or repository PID when the workflow, citation, resolution, or preservation record benefits—not to decorate a trade book with more numbers.
A book ISBN should not be applied as the identifier for a chapter, author, institution, transaction, or individual library copy. Use identifiers designed for those entities.
A DOI that resolves to a broken page or has obsolete metadata fails operationally even though its string persists. Assign responsibility, service credentials, budgets, and maintenance checks.
|
Need |
Use ISBN? | Use DOI/other PID? | Other identifier | Decision rule |
Owner |
| Trade print ordering | Yes | Optional | EAN-13/ASIN/SKU | ISBN-based supply chain | Publisher |
| Kindle-only eBook | No ISBN required by KDP | Rarely | ASIN | Platform rule | KDP/account |
| Library cataloging | Useful/standard | Repository PID may supplement | Control/item IDs | Bibliographic vs copy | Library/publisher |
| Scholarly citation | Book ISBN useful | DOI often valuable | ORCID/ROR/ISSN | Granularity and resolver | Publisher/member |
| Chapter citation | No chapter ISBN | Chapter DOI if registered | Book ISBN | Component-level object | Publisher |
| Repository deposit | Cite source ISBN | Handle/DOI/URN | Repository ID | Deposited version | Repository |
| Author identity | No | ORCID iD | ISNI | Person-level identifier | Researcher |
| Institution identity | No | ROR ID | Other org IDs | Organization-level identifier | Registry |
| Internal inventory | Map ISBN if book product | No open PID needed | SKU/local PID | System-scoped operation | Seller |
| Preservation | ISBN remains useful | PID helps resolution | Archive ID | Needs real preservation service | Repository/partner |
In research and scholarly publishing, PID normally means persistent identifier: a durable reference to a resource or entity. A local system may use PID for product or publication ID, so verify the namespace.
ISBN is a permanent, standardized product identifier and is often discussed within the broader persistent-identifier landscape. Unlike DOI, it is not generally used as a single web-resolution mechanism. State the comparison criteria rather than forcing a yes-or-no label.
No. ISBN identifies a specific book product. DOI persistently identifies and resolves a registered object and can be assigned at book, chapter, article, dataset, or other supported levels.
No. Many trade books use ISBNs without DOIs. Scholarly or repository workflows may add a DOI for persistent linking and richer research metadata.
Yes. The ISBN can identify the edition/format while the DOI identifies and resolves the registered book or chapter record.
ASIN is Amazon’s proprietary catalog identifier. It may be stable within Amazon but is not interchangeable with open scholarly PID systems.
KDP does not require an ISBN and assigns an ASIN. A DOI is optional and normally comes from a publisher or repository workflow, not KDP.
ORCID iD is the widely used persistent identifier for researchers and contributors. ISBN identifies the book, not the person.
ROR ID is an open persistent identifier for research organizations. An affiliation record can use ROR while the author’s ORCID and output’s DOI remain separate.
A continuing series can have an ISSN. Each complete monographic volume can have ISBNs, and components may have DOIs.
Not when the publication needs ISBN-based book-trade identification. A repository PID can identify the deposited object while the book product retains its ISBN.
No. Persistent resolution depends on governance and maintenance; preservation depends on repositories, storage, migration, rights, and stewardship.
The registering member or repository is responsible for keeping its metadata and destination current under the registration agency’s rules.
ISBN and PID are not competing labels for the same thing. ISBN is a specific standard for identifying book products. Persistent identifier is a category that includes systems designed for durable identification and, often, resolution of objects, people, and organizations.
Start with the entity. Give each book product the correct ISBN, each registered scholarly object an appropriate DOI or repository PID, each contributor an ORCID iD, and each research organization a ROR ID where the workflow benefits. Then maintain the metadata and links. The right identifier stack is not the one with the most numbers—it is the one that keeps every entity unambiguous and usable over time.
