This research paper was completed in the Fall 2025 semester as part of the "Foundations of Information" course at Pratt Institute, overseen by professor John Lauermann.
Selection and Implementation Considerations for Digital Asset Management (DAM) Platforms
Sabrina D. Chaney
School of Information, Pratt Institute
INFO 601: Foundations of Information
Professor John Lauermann
December 1, 2025
Abstract
Confronted with the challenge of organizing, cataloguing, preserving, and maintaining a digital collection of files, every digital asset manager is faced with the same initiatory question: "Which DAM software should I use?". Every institution has unique needs that result in different decisions about which DAM to adopt. The vast number of available vendors complicates matters for both institutions that house digital assets and the information professionals that manage them. Making a choice that later proves to be an incorrect fit has the potential to incur great costs of time and resources in order to migrate to a better DAM system.
Knowledge institutions have many problems to consider. As data accumulates over time, it becomes a struggle to accommodate a growing collection of assets. Day-to-day operations may be assisted by automation or replaced outright by artificial intelligence depending on the task at hand. Linking assets from one institution to another is another common roadblock. A multitude of DAM vendors have cropped up in recent years vying for the chance to satisfy these challenges, as well as increase their customer base; however, their basic functions appear redundant upon further inspection, and they quickly come to resemble each other.
Therein lies the question: What are the desirable attributes that persuade information professionals to implement a particular digital asset management platform? Potential considerations include pricing schemes to ensure the DAM is continually funded, ease of use to ensure cross-functional accessibility, and customizability to adapt the DAM according to the assets being collected. Ideally, the leading vendors in the DAM market are taking steps towards future-proofing, reducing the chances of obsolescence that would cause their customers to switch to a different system, risking the loss of their data. The following paper proposes that the selection and successful implementation of a digital asset management platform hinges on the tenets of scalability, automation, and interoperability.
I will draw from existing literature about digital software implementation, both in the historical context of libraries implementing digital systems to make their catalogs accessible in the digital age, and the modern-day context of adapting existing digital collections to newly emerging technology. Doing so will aid in defining key criteria that are essential to selecting and implementing a DAM platform such as scalability, automation, and so on. Turning attention to the present-day DAM market, I will refer to verified customer success reports to analyze the leading software features that users seek out the most. I will also examine research papers and journal articles about evaluation and decision criteria in the realm of digital preservation, which will offer further insight about selecting and implementing a DAM system out of a sea of choices, straight from the professionals that regularly deploy digital preservation techniques.
Literature Review
Henry Stewart Conferences a pre-eminent platform for industry veterans and newcomers alike to network and attend panels about the topic of digital asset management. At DAM New York 2025, Henry Stewart's latest conference offering, the event had over 35 digital asset vendors in attendance, ranging from popular companies such as Orange Logic and Adobe to lesser-known services such as Papifly, Air, Sesimi, and Marq. My research implies that it's necessary to rely on a variety of decision criteria when choosing a DAM system, rather than jumping directly into one particular solution, software package, or vendor. Relying on the vendors themselves leaves information professionals at risk of falling victim to sales pitches and marketing-speak, and fails to incorporate the context of prior technological shifts that preceded the practice of digital asset management. Little work has compared the documented decision criteria of information retrieval systems versus the specific, historically contemporary domain of digital asset management.
Presently, most research that delves into defining and implementing effective information retrieval has been applied to digital libraries and the practice of digital preservation. As Dr. Iris Xie posits, "Digital library practice is institutionally/organizationally based and oriented toward a given community, pragmatic development, and practical operations. As expected, the aims are toward the pragmatic problems at hand." (Xie, 2016). This bears a striking resemblance to the continually evolving practice of digital asset management. My research aims to draw from prior digital preservation research to demystify the current DAM landscape and analyze its most integral attributes. Selected works relevant to the discussion include books and articles relating to digital libraries (Xie, 2016) and information retrieval systems (Bates, 2012).
Marcia Bates, a preeminent scholar in the field of Library and Information Science, asserts that "[information] systems need to be designed specifically for their intended functions in order to provide optimal support for the people who use them" (Bates, 2012). Bates extensively surveyed the relationship between information and the subjective experience of those who interact with patterns of information, providing fundamental definitions that informed information theory and collections disciplines (Bates, 2006). My research will examine that which is agreed upon in the selected literature and in the DAM industry: information retrieval, of which digital asset management is a part, is a discipline oriented around the needs of the user. Successful selection and implementation of a DAM platform, therefore, develops in relation to those needs.
In summation, my research aims to focus on specific decision criteria that apply to selection and implementation of a DAM system. Scholars in the field of digital preservation have analyzed influence factors using controlled experimentation and automated measurements to determine decision criteria such as scalability and automation (Becker & Rauber, 2011). Others have sought to define and identify decision criteria based on user practices, measured with surveys (Joo, Xie, et al.), or by analyzing case studies (Xing, 2021). Looking at the digital asset management industry directly, market research and customer reports reveal the industry leaders that have proven track records of successful implementation with a wide variety of digital collections (GlobeNewswire, 2022) (FeaturedCustomers, 2024). My goal is to synthesize the literature in relation to core attributes that guide implementation of a DAM system.
Secondary Analysis
Scalability
Scalability in the context of information retrieval is defined by the “amount of data used and the number of distributed components cooperating” (Bates, 2011). Christoph Becker, Luis Faria, and Kresimir Duretec identified components of scalability including “efficiency in the data processing system”, the “creation of large content profiles”, and “flexible integration with diverse institutional environments” (2014). Their combined research, presented in the paper titled “Scalable decision support for digital preservation: an assessment” explored “capabilities for semi-automated, scalable decision making and control of preservation functions in repositories” (Becker et al., 2014). Their research confirms that a scalable DAM system can handle very large, heterogeneous collections of assets efficiently; as the asset volume increases over time, the system can process the burgeoning file size and metadata with as few bottlenecks as possible.
To prove the above assertion, the article cited controlled tests intended to analyze scalable content profiling with a tool called C3PO (Becker et al., 2014). The controlled test “...explored the boundaries of scalability by attempting to profile up to 400 million resources (12 terabyte [TB]) in a single profile, enabling a further extrapolation of these results to the entire set of 300 TB in this collection” (Becker et al., 2014). According to the C3PO developers’ GitHub repository, the tool is thus described: “Clever, Crafty, Content Profiling of Objects (c3po) is a software tool, which uses meta data extracted from files of a digital collection as input to generate a profile of the content set. It is designed in a way so that different meta data formats originating from different tools can be easily integrated” (Petrov, 2011/2024). The “meta data extracted from files of a digital collection” (Petrov, 2011/2024) portion is important to underline here, as the content profiling tests could be used to analyze the scalability of contents of a digital asset management library, which are equivalent to a digital collection.
C3PO was deployed by Niels Bjarke Reimer from the Danish State and University Library to formulate content profiles for a 12 TB sample that was “taken from a dataset with 300 TB of the Danish Web archive” (Becker et al., 2014). By beginning with the sample, Niels could then determine how the process would apply to the entire 300 TB collection. First, the 12TB sample was converted into 441 million FITS (File Information Tool Set) files (Becker et al., 2014), which extract technical metadata and produce “XML results following a documented schema that can be analyzed straightforwardly” (Becker & Rauber, 2011). The FITS files were then ingested into a MongoDB server, averaging 0.65 ms per FITS file, and the total import process for all 12TB was completed in less than 80 hours (Becker et al., 2014). The data was then analyzed using MapReduce queries, a technology that could support distributed processing on multiple servers, “and took 15 hours and 18 minutes, which is about 4.63ms per FITS file”, demonstrating linear scalability (Becker et al., 2014). By measuring a large “amount of data processed in a certain timeframe using a defined set of resources” (Becker et al., 2014), the large-scale C3PO test demonstrated the feasibility of profiling hundreds of millions of files with parallel processing. Efficient content profiling is inherently desirable for a DAM system in order to “create and maintain an awareness of the holdings of an organization, including the technical variety and the risk factors that cause difficulties in continued access and successful preservation” (Becker et al., 2014). Thus, conducting a similar test on a DAM system to gauge efficiency would ensure the successful implementation of scalability, satisfying my hypothesis.
Automation
Automation is another vehicle for increased efficiency, as it reduces human involvement and minimizes “the amount of labor needed to build and maintain a shared collection” especially when “decisions are routine and highly structured” (Bates, 2011). The wide adoption of artificial intelligence and large language models has allowed for a multitude of opportunities to automate processes within a DAM, from automatic metadata generation, to face recognition tools that can identify the subject in an image, to AI-generated descriptive summaries, and more. According to a market research report of the digital asset management software market published in October 2025, the desirability of automation is clear: “the integration of artificial intelligence and machine learning technologies into these platforms suggests a shift towards more intelligent asset management systems that can automate processes and enhance user experience…These technologies facilitate smarter asset organization and retrieval, potentially reducing manual effort and improving overall workflow” (Dhapte, 2025).
A recent research paper from the Journal of Theoretical and Applied Information Technology sought to “design a digital asset management architecture using AI TRiSM [AI Trust, Risk, and Security Management] and evaluate it to ensure that it is reliable, risk-reducing, and secure, leading to continued organizational sustainability” (Tasatanattakool et al., 2025). The inclusion of AI TriSM is pertinent in regards to viewing artificial intelligence technologies through a critical lens, because it “is a framework that addresses the issues of trustworthiness, risk management, and security of AI systems” (Tasatanattakool et al., 2025). If a DAM system features tools augmented with machine learning technology, for example, but those tools repeatedly generate inaccurate results, trust in the greater DAM system would wane amongst users, defeating the purpose of integrating automation into the system. The AI TRiSM architecture utilized “Application Programming Interfaces (API) to integrate the fundamental components of data, processes, and algorithms” (Tasatanattakool et al., 2025). The deployment of an API facilitated dynamic data management and linkage between digital assets “by enabling the integration and interoperability of diverse data sources and knowledge domains, thereby supporting data-driven methodologies that are advantageous for asset management” (Tasatanattakool et al., 2025), ensuring seamless integration of diverse data sources without human bottlenecks.
The research paper provided quantitative evidence in the form of ratings provided by “nine experts in ICT or IT working in educational institutions” (Tasatanattakool et al., 2025) after evaluating the TRiSM AI digital asset management system architecture. The selection of working professionals’ review and expert insight is valuable because, again, digital asset management is a discipline that hinges on the needs of the user; if the AI TRiSM framework received low ratings after the experts’ evaluation, it would refute my hypothesis that automation is a key criterion for the successful implementation of a DAM system. Luckily, the final results were reported as “excellent overall (mean = 4.53, S.D. = 0.49). This observation aligns with the findings of numerous scholars who have determined that Artificial Intelligence Trust, Risk, and Security Management (AI TRiSM) can substantially enhance knowledge acquisition and decision-making processes for executives within the educational sector” (Tasatanattakool et al., 2025). The quantitative evidence proves that the AI TRiSM framework is a useful guide for successfully implementing a DAM system with automation as proposed in my hypothesis, aiding in such tasks as “monitoring and tracking of an organization's assets”, “mitigating potential failure risks”, and enhancing “the organization and retrieval of digital information while reducing operational costs” (Tasatanattakool et al., 2025) through the implementation of a fine-tuned API.
Interoperability
Marcia Bates defined the importance of interoperability: "As soon as two [information retrieval] systems need to talk to each other, or information is to be sent over a network, then it becomes necessary to develop common standards that can be used by developers to enable the various systems to inter-operate" (Bates, 2011). In the context of DAM systems, metadata standardization tools facilitate interoperability by reducing complexity, clarifying relationships and hierarchies of assets, and solving “the retrieval problem caused by the use of natural language in the search” (Xing, 2021). Ahmed Alkhard, of King Abdul Aziz University, sought to research what interoperability looks like in practice using both quantitative analysis and literature review to understand how integrated facilities data, digital asset management, and metadata strategies improved asset and facility management.
Alkhard’s qualitative analysis consisted of 2,340 facility-related issues reported at a public school in Saudi Arabia with data extracted from the school's system in .xls format "to identify patterns and trends in problem types, severity, and reported frequency, offering insights into prevalent challenges encountered in asset management and informing the research objectives" (Alkhard, 2024). His analysis found several problems that were the result of poor asset management practices that directly correlated to weakened interoperability, including "instances of duplicated data and records within the repository", "incomplete or inaccurate entries", incomplete status reporting, and "confusion between asset and facilities data" (Alkhard, 2024). In the absence of standardization, simply uploading the reports to a DAM would fail to resolve the issues of redundancy or inaccessibility, as different stakeholders (such as school custodians, teachers, contractors, etc.) could run the risk of tagging the reports with scattershot metadata, making search and retrieval needlessly difficult. Alkhard’s analysis therefore serves as proof of the desirability of interoperability tools in a DAM, satisfying my initial research question.
Alkhard’s literature review consisted of real-world case studies from Coca-Cola Enterprises, The Metropolitan Museum of Art, The Port Authority of New York and New Jersey, and Getty Images that demonstrated "how digital asset management (DAM) and Asset Information Modeling and Management (AIMM) have been implemented in organizations to achieve tangible benefits", including "improved operational efficiency, accelerated content delivery, and enhanced customer experiences through robust metadata management, version control, and rights management" (Alkhard, 2024). Alkhard used his research to propose an integrated digital asset management strategy, defining stages like "Implement Metadata Capture and Storage Processes…", "Establish Metadata Governance…", "Train and Educate Stakeholders…", and more (Alkhard, 2024). Alkhard's emphasis on the proper integration of facilities data, standardization practices to reduce redundancies, and protocols for data storage and training, echo the importance of interoperability as a key decision criterion when choosing a digital asset management platform, satisfying my hypothesis.
In the world of digital preservation, Soohyung Joo, Iris Xie, and Krystyna Matusiak conducted “two rounds of online surveys with 30 digital library scholars and 30 digital library scholars” in order to identify evaluation criteria and “rate the appropriateness of specific measures for digital practice using a seven point scale” (Joo et al., n.d.). Their research found that the “Ability to Migrate” criteria was rated the highest by digital library scholars, which was tied to the measures of “migratable data” and “exporting capability” which draw a direct line to users’ desire for interoperability tools within the information retrieval and management systems that they use (Joo et al., n.d.). Alkhard's integrated asset management strategy serves as a useful guide for assessing different DAM platforms for their ability to inter-operate disparate sets of files and data. While Alkhard's analysis focused on the specifics of facilities data, the same conclusions could apply to many different enterprise-level needs, as evidenced by the multiple case studies in Alkhard’s literature review sampled from a wide range of industries and business sectors. A DAM platform that provides interoperability, therefore, features tools that properly classify and categorize metadata and assets in a way that can be easily reviewed, managed, and taught throughout a business, organization, or knowledge institution.
Conclusion
Digital asset management systems purport themselves to be a single source of truth for documents, photos, and other digital media to be housed, connected by metadata and distributed by tools customized to each organization's needs. According to my research, the functionalities that best support this goal are scalability, automation, and interoperability.
Scalability ensures the longevity of a system by allowing for growth in the collection without sacrificing efficiency. Measuring the time that it takes to generate and ingest metadata content profiles onto a server is an effective test of scalability, as it could reveal potential weaknesses in processing power that can be ameliorated with improved software or an upgraded DAM platform.
Automation reduces the amount of man-hours it takes to accomplish tasks, which is a powerful incentive for stakeholder buy-in, because it opens up financial resources for investment. Implementing an API that abides by the AI TRiSM framework allows organizations to incorporate automated tools with artificial intelligence while prioritizing security and trust, which are important for the governance of digital assets.
Finally, interoperability reduces inaccuracies, redundancies, and errors through metadata standardization and training protocols. Controlled or custom metadata vocabularies, detection of duplicate files or versions, or robust training manuals are all tools that support interoperability in a DAM system. An integrated metadata strategy all but guarantees their adoption and maintenance for long-term information retrieval.
While there exists a wide variety of DAM platforms in the market, those that acknowledge and facilitate scalability, automation, and interoperability are positioned for success because their toolsets, features, interfaces, and programming put the needs of the user first, from the inception of a repository to its growth and maturity in the face of continuously changing technologies.
References
Alkhard, A. (2024). Enhancing asset management through integrated facilities data, digital asset management, and metadata strategies. Construction Economics and Building, 24(3), 76–94. https://doi.org/10.5130/AJCEB.v24i3.8741
Bates, M. J. (2006). Fundamental forms of information. Journal of the American Society for Information Science and Technology, 57(8), 1033–1045. https://doi.org/10.1002/asi.20369
Bates, M. J. (Ed.). (2011). Understanding Information Retrieval Systems: Management, Types, and Standards (First). Auerbach Publications. https://go.oreilly.com/pratt-institute-library/https://learning.oreilly.com/library/view/~/9781439891995/?ar&orpq&email=%C3%BB
Becker, C., Faria, L., & Duretec, K. (2014). Scalable decision support for digital preservation. OCLC Systems & Services: International Digital Library Perspectives, 30(4), 249–284. https://doi.org/10.1108/OCLC-06-2014-0025
Becker, C., & Rauber, A. (2011). Decision criteria in digital preservation: What to measure and how. Journal of the American Society for Information Science and Technology, 62(6), 1009–1028. https://doi.org/10.1002/asi.21527
Dhapte, A. (2025, October). Digital Asset Management Software Market Size, Share Report 2035. Market Research Future. https://www.marketresearchfuture.com/reports/digital-asset-management-software-market-1196
Joo, S., Xie, I., & Matusiak, K. (n.d.). Evaluation Criteria and Measures for Digital Preservation Practice. Society of American Archivists. Retrieved October 6, 2025, from https://www2.archivists.org/sites/all/files/SAA_forum_poster_Joo_Xie_Matusiak.pdf
Petrov, P. (2024). Peshkira/c3po [Java]. https://github.com/peshkira/c3po (Original work published 2011)
Tasatanattakool, P., Wannapiroon, P., & Nilsook, P. (2025). System Architecture of Digital Asset Management with AI TRiSM. Journal of Theoretical and Applied Information Technology, 103(2). https://doi.org/10.1109/ICCI60780.2024.10532376
Xie, I., & Matusiak, K. (2016). Discover Digital Libraries (First). Elsevier. https://learning.oreilly.com/library/view/discover-digital-libraries/9780124201057/
Xing, M. (2021). What Are the Core Elements Necessary for Effective Digital Asset and Media Management Practice? Open Journal of Social Sciences, 9(10), 339–350. https://doi.org/10.4236/jss.2021.910024