Abstract – From the very start of digital preservation, KB National Library of the Netherlands (KBNL) has played a pioneering role in the digital preservation community. From the NEDLIB project in the 90s of the twentieth century, the worldwide first operational e-Depot in accordance with the OAIS standard in 2002, to important contributions to international research in digital preservation in the first years of the twenty-first century. KBNL shared its successes and knowledge with the digital preservation community by means of papers, professional visits and conferences.

But at a certain point in time KBNL became a bit quiet, and remained quiet for a while. Until almost ten years later, when the organization brought renewed focus on its preservation activities, and shared amongst others the purchase of Rosetta and the achievement of CTS.

But what happened in those quiet years, and more importantly, how did it happen, and most importantly, how do we prevent it from happening again?

In this paper, I share my personal reflections as a preservation practitioner, about the impact of organizational context and culture on thirty years of digital preservation within KBNL, to unlock the past, present and future of digital preservation within our institution. Although sharing honest stories in public is, unfortunately, not usually an option, we think it’s important to be open to the community about our challenges and things that have not gone as well as we liked, so we can all learn from those experiences and move forward together.

“Many times what we perceive as an error or failure is actually a gift. And eventually we find that lessons learned from that discouraging experience prove to be of great worth.” [1]

Keywords – Preservation, Organizational context, Reorganization

This paper was submitted for the iPRES2024 conference on March 17, 2024 and reviewed by Koen Dobbelaere, Mikala Narlock, Dr. Lee Pretlove and 1 anonymous reviewer. The paper was accepted with reviewer suggestions on May 6, 2024 by co-chairs Heather Moulaison-Sandy (University of Missouri), Jean-Yves Le Meur (CERN) and Julie M. Birkholz (Ghent University & KBR) on behalf of the iPRES2024 Program Committee.

Introduction

On Monday, 12 December 2022, KB National Library of the Netherlands (KBNL) celebrated the twentieth anniversary of its e-Depot [2]. With a large group of colleagues from the past and the present  we celebrated the fact that KBNL had been able to successfully preserve digital publications in the e-Depot for 20 years.

But... did we really have a good reason to celebrate this twentieth anniversary, and what did we celebrate exactly? And were we really that successful? Or to put it another way: How do you measure success when you have been working on digital preservation for more than twenty years? Is successful preservation about state-of-the-art technology and  applications? Or is successful preservation about the growing amount of the preserved digital collections? Or maybe is successful preservation about sustaining digital preservation within an organization only to the best of our abilities, especially where there is tension between the long term goals of digital preservation and the short-term objectives and measurable results at an institutional level [3]?

In the year of our anniversary we decided to share our honest story in public, both on World Digital Preservation Day [4] and on our anniversary [5], to unlock our history of preservation, with all its ups and downs. Although sharing honest stories in public is, unfortunately, not usually an option, we think it is important to be open to the community about our organizational difficulties and challenges, and about things that have not gone as well as we would have liked. Making these undiscussable things discussible, and confronting any threat or embarrassment that might result from this, instead of avoiding it, are, in our opinion, the first steps to learn from those experiences and move forward [3].

Because we often seem to think we're getting better and better over time, but progress is rarely linear. This type of thinking might be a legacy of the concept of whig history. However, in practice we see that early innovators might suffer from first mover disadvantage after initial success [6]. Yet we seem to think that, when it comes to preservation, it is realistic to expect constant improvement. The models for certification and maturity contribute to this idea, which can be deduced from the fact that certification has to be renewed periodically and institutions are required to show progress over the course of years. And in doing so, we fall into the trap of thinking that preservation can improve in a linear fashion, independently of the strategic direction and,  perhaps more importantly, the organizational culture of the institution.

As the goal of (successful) digital preservation is to maintain the ability to display, retrieve, and use digital material in the face of rapidly changing technological and organizational infrastructures, we can still ask ourselves the question how we best “implement short-term strategies for long-term problems” within our own organizational context and organizational culture.

As Rebecca Frank stated in her dissertation The Social Construction of Risk in Trustworthy Digital Repository Audit and Certification [7], the people that create the standards are mainly concerned about the preservation of the data but employees are also concerned about the preservation of the institution itself because they work there.

“standard developers were not concerned with threats posed to a repository by legal issues, but rather to the digital information being preserved by that repository [...] What we were concerned about is that they were sued out of existence before it could hand over its data, its information [...] For this interviewee, the danger to the organization, that a repository would be shut down before they could enact their succession plan, was a significant threat to the digital content.”

This shows that in practice there are competing values for preservationists, depending on whether they follow the standard or prioritize the concerns of their employer. So organizational culture and context is very important for implementation of digital preservation within an institution, for preservation is more than policies, ingest, storage and technique.

The Golden Age of Digital Preservation at KBNL

The Early Years

In the 1990s KBNL was one of the first (national) libraries to become fully aware of the electronic revolution in the publishing world and to develop a strategy for digital publications. Various research reports had already indicated that digital publications have countless advantages, such as almost immediate world-wide distribution through the internet, but that their longevity was threatened by frequent changes in software and hardware [8]. As providing long term preservation and accessibility of the national production of digital publications was seen as a main task of a national library, KBNL had no choice but to delve into the question of how to keep these publications accessible for users in the future. For this reason KBNL actively participated in various international preservation research projects and started a series of experiments within the institution on processing digital publications.

On KBNL’s initiative, a group of European National Libraries together with the National Archive of the Netherlands, some large publishers and IT companies, initiated the NEDLIB (Networked European Deposit Library) project in 1998 [9]. The project aimed to  construct  the  basic  infrastructure upon which a networked European deposit library could be built. These objectives concurred with the mission of national deposit libraries to ensure that digital publications could be used now and in the future. The project ended in 2000, resulting in a Process Model for an e-Depot and directions for its implementation. Attention was also paid to standards for digital publishing, authenticity, and other aspects of digital preservation. In the same period a Functional Model for Digital Preservation was developed by NASA, which would lead to the well-known OAIS reference model for digital preservation. Because of the great similarities between the NASA  Function Model and the NEDLIB Process Model, the NEDLIB’s terminology was aligned with that of the NASA Functional Model. The NEDLIB project has thus actively participated in the development and dissemination of OAIS (ISO 14721) [10].

In 1995, KBNL decided to start a series of experiments and projects which would lead to a deposit system for Dutch Electronic Publications. The organization made a policy decision to include digital publications into its deposit and reached a general agreement with the Dutch Publishers' Association (in Dutch: Nederlands Uitgeversverbond (NUV)) on the deposit of digital publications. This layed an foundation for expanding the depository collection with all kinds of digital publications (off-line and on-line) [11] and marked the start of the project Dutch Deposit for Electronic Publications (in Dutch: Depot voor Nederlandse Electronische Publicaties, hence DNEP) [12]. DNEP functioned both as an operational service as  part  of  the  production  workflow  inside  the  library  departments and at the same time as a test-bed for research into digital archiving [13]. CD-roms, diskettes and magnetic optical disks, but also online digital publications, were catalogued, stored and made available to the public.

Besides KBNL’s activities on digital preservation, also the national program Metamorfoze [14] started in 1997. The goal of this program was to prevent physical deterioration, first by preserving a copy on microfilm, later by switching to digitization by means of the preservation imaging-method.

Ramping Up

After gaining experience with digital archiving, KBNL sent out a European tender for a long term preservation system in 2000. And on 12 December 2002 KBNL’s e-Depot IBM-DIAS (IBM Digital Information Archiving System [15]) was put into operation, being the first operational long-term preservation library system, based on the OAIS reference model, running worldwide [16].

Although IBM-DIAS was developed primarily to secure long-term preservation of Dutch digital  publications in the Dutch deposit library, KBNL acknowledged that digital publications are simultaneously published all over the world by international publishers and often lack a clear-cut “nationality”. KBNL therefore developed a new type of service, namely digital archiving for international scholarly publications. The international e-Depot was born. In about ten years the international e-Depot processed around 18 million e-journal articles from various Scientific, Technical and Medical (STM) publishers [17].  But besides preserving  these international publications, KBNL also processed  CD-roms, ebooks and scientific publications from the Dutch institutional repositories, from its role as a national library.

Besides preserving born digital publications in its e-Depot, in 2006 KBNL also started a web archiving project [18] and large-scale digitization of newspapers and books via the Metamorfoze project and partnerships with Google & Proquest [19].

With the start of the operational e-Depot system digital preservation became strongly embedded within KBNL organization [20].

For daily operations a dedicated team (the e-Depot department) managed the e-Depot workflows. They maintained contacts with suppliers of the born digital publications, developed and maintained new workflows, and monitored the ingestion processes and error handling.  As key users of the e-Depot system, the team was experienced in designing useful improvements and running tests when updates were implemented.

The e-Depot was also a driving force for innovation. The Digital Preservation Department as part of the R&D division had a firm practical basis and focused on implementable solutions. The Digital Preservation Department was responsible for KBNL’s preservation policies and contributed to national and international research projects about various preservation topics [21] like Universal Virtual Computer (UVC) for images, Preservation Manager, characterization, file formats, preservation strategies like emulation and migration, preservation metadata, safe places network and PLANETS (Preservation and Long-term Access through Networked Services) [22] .

The departments worked closely together and with the IT department, and both departments exchanged their knowledge and experience very actively in the international network.

In my opinion as a preservation practitioner, KBNL had organized digital preservation quite optimally during these early years. The organization could have left all seven “checkboxes”, as emerged from the analysis in “What's wrong with Digital Stewardship”, unchecked, and thus had created a rather ideal culture for preservation practices. There was no fundamental misalignment between the senior organizational leaders and preservation practitioners. Management had a long term vision for preservation, recognized the active and ongoing nature of digital preservation practices, and therefore acknowledged long-term organizational commitment was required.  Management also  expressed full confidence in the work of their staff and there was equal cooperation across all hierarchical levels of the organization, creating authority for practitioners to carry out their  preservation activities [23].

Looking for a second generation e-Depot

In 2009 the IBM-DIAS system had processed and stored over 15 million digital objects, mainly e-journal articles [24]. In the same year, after seven years of operation, KBNL started the process to replace the IBM-DIAS system, the “seven-year-itch” (or the system already “living” longer than most other IT systems), only being one of the reasons. KBNL had published an ambitious new policy plan, with the main goals to give access to all Dutch publications, preserve these publications for the long term and facilitate a national infrastructure for access to all these publications [25]. This ambition and the given challenges (enormous growth of digital collections, different types of digital objects, growing complexity, new preservation functionalities, and digital preservation and permanent access as core of the library process) were translated into requirements for a second generation e-Depot. “Revolutionary” concepts such as the implementation of preservation levels for differentially processing types of collections, the two-layered OAIS model and the use of components or so-called “building blocks” in the implementation of a new e-Depot system, played an important role in the requirements [26].

The idea was to place a separate request for information in the market for  each  component or “building block”. It was expected  that each component could possibly be filled out in a different way, and that the  separate components would either be bought, integrated or developed. But at the end, the request for information did not lead to a suitable supplier for any of the components, and KBNL decided to develop its own second generation e-Depot, called “Digitaal Magazijn” (Digital Warehouse) [27].

Impasse / The real Seven Year Itch

A Library in Transition

Around the same time the decision was made to develop a new homegrown e-Depot system, some unforeseen organizational challenges presented themselves. And that turned out not to be a very happy marriage [28].

Reorganization of KBNL

In the aforementioned Strategic Plan 2010-2013 [29], KBNL focused on the development of the digital library. The transition to a digital library made it necessary to review the organizational structure. The main goals of the reorganization in 2010 were to streamline the primary library processes to improve efficiency in processes and communication, to be better prepared for the digital library [30]. With the motto "an integrated approach for the analog and digital collection" for the Collection Development Programme (in Dutch “Fysiek en digitaal integraal” [31]) the core library processes were redefined. As a result, the associated core responsibilities and roles were redefined and reassigned, and the organization was restructured according to this principle. For the organizational structure this meant that activities common to both digital and print assets (selection, processing, storage, and access) would be handled, regardless of asset type, in matrix fashion by all relevant departments (Collections department, Document Processing and Digitization departments, Collection Care department, and Public Service department) [32]. So for instance, digital preservation and analog conservation were at high-level defined as the same process, and the responsibility for this general preserving process was assigned to one department Collection Care. Although this might sound like a good idea, the big question is how to move this theory into good practice?

Although the responsibility for the general preserving process was assigned to one department, for digital preservation practice within KBNL this organizational change meant that responsibilities were divided over many different departments, with strictly defined responsibilities between them. As is known from organizational literature, restructuring an organization may cause increased hierarchical distance, which has an impact on informal lines of communication and advice [33]. Having separated and clearly defined roles is a feature of the formal organization and a means for achieving efficiency, but it hinders coordination of preservation activities across the departmental borders [3]. This is confirmed in the conclusion in the article “What's wrong with Digital Stewardship”, and especially in the chapter “Hierarchical organizations exacerbate stewards’ lack of authority” [34].

And although the long term preservation and accessibility of all Dutch digital publications was one of the main goals of KBNL’s Strategic Plan 2010-2013 [29], in practice it meant that the organization started focusing on the large-scale digitization and accessibility of the physical collection.

As an other result of the reorganization, the Digital Preservation Department of the R&D division became a more generic research department, and the former e-Depot department was abolished and broadened into a more general department for product support for all "back-end" systems in the organization. As a result staff time had to be prioritized amongst many different products and services. This led to a drastic decrease of staff time dedicated to digital preservation activities. 

In addition, there was much less outward looking, and cooperation with the national and international network shrank.

In practice, all of these organizational changes meant that digital preservation focus and pace was hampered when that merge approached.

Political choices and changes

On top of the 2010 reorganization, KBNL had to deal with other ongoing political choices and changes as well.

In December 2011, it was announced that KBNL and the National Archives of the Netherlands would become one organization, with a target date of 1 July 2013 for integration. But work on the merger was discontinued in 2013 [35].

In 2015, as part of the implementation of a new law, defined as the Wet stelsel openbare bibliotheekvoorzieningen (WSOB – Public Library Provisions System Act), KBNL got a central role in the field of public libraries as well, something it has been doing for years in terms of scholarly information provision. With the WSOB the scope of responsibilities of KBNL broadened considerably. KBNL got the task to devote attention to the subjects that transgress the local aspect of public libraries, such as a joint infrastructure, non-geographically bound target groups and supervising innovation, knowledge-sharing and research in the sector [36].

This major addition to KBNL's tasks and responsibilities meant a growing organization, with different types of tasks, resulting in even less time and attention to the work needed to keep the digital collections sustainable in the long term.

Development of the “Digitaal Magazijn”

One can imagine that all of these organizational changes had a great impact on the development of the “Digitaal Magazijn”. The plan was to build the “Digitaal Magazijn” in four releases, where each release would focus on a type of digital collection, and where each release would have more functionality as needed by the type of collection.

Release 1 of the “Digitaal Magazijn” was delivered in July 2012, and focused on conversion software for the migration of the born digital collection from IBM-DIAS to the “Digitaal Magazijn” [37]. The planned further releases were release 2 for processing new born digital collections (so IBM-DIAS could be phased out), release 3 for processing digitized collections, and release 4 for processing websites. Also in each release more functionality would be added, such as access for customers in release 2, preservation management in release 3 and improved user interfaces in release 4 [38].

But already during the implementation of release 1 the clock was ticking. KBNL’s maintenance contract with IBM would run out and components of IBM-DIAS would no longer be supported. And that would jeopardize the processing and preservation of the publications under the existing agreements for KBNL’s born digital collections. In September 2011 management decided a temporary solution would be needed, to give ourselves more time for the development of release 2 of the “Digitaal Magazijn”. The so-called release 1.5 was built, and delivered in 2013. In this release preprocessing of IBM-DIAS remained unchanged, metadata was transformed into the data model of the “Digitaal Magazijn” and access to the stored publications was taken care of. The major principle of this temporary release was that no or only minor changes were foreseen in the software, and that certainly no new workflows would be included in this temporary solution.

In the meanwhile, during the design and development of release 2 of the “Digitaal Magazijn”, every involved department had its own ideas about what the necessary functionalities for this release should be, and how these functionalities should be ideally  implemented.  One can imagine that all these discussions delayed and frustrated  the development of the “Digitaal Magazijn”. In the end, in July 2014, management stopped the design and development of release 2 of the “Digitaal Magazijn”. The temporary solution is, even in 2024, still up and running, under the name DM1.5.

But the management’s decision to stop further design and development of the “Digitaal Magazijn” did (of course) not stop the discussions. The library still needed a Long Term Preservation (LTP) system to preserve all of its digital collections, and not just the born digital publications that were migrated from IBM-DIAS and the born digital publications that were now ingested using the workflows of DM1.5. Which approach would be the best way forward? Different ideas were suggested, coming from all fragmented departments that were in a way involved with the preservation of the digital collections or the DM1.5 system. Should we further develop DM1.5? Or should we start over with what was minimally needed? Or should we perhaps place another request in the market?

And because of the impact of the reorganization (divided and narrowly defined responsibilities, hierarchical distance, lack of  informal lines of communication and advice, cumbersome   coordination of preservation activities across the departmental borders and less cooperation and knowledge exchange with the national and international network), the organization struggled to remain focused on digital preservation and how to approach this need for innovation. We got into a bit of an impasse regarding how to move on.

As far as I'm concerned as a preservation practitioner, the lack of direction, the lack of long term vision, less confidence in staff and the lack of  informal collaberation [34], is also clearly evident from the management's decision to ingest two new digital collections in DM1.5, just three months after the discontinuation of the design and development of release 2 of the “Digitaal Magazijn”. While the major principle of DM1.5 was that no or only minor changes were foreseen in the software, and that certainly no new workflows would be included in this temporary solution. On top of that, the two collections did not adhere to any of the former guidelines, so a non-dedicated team had to design and develop completely new workflows in DM1.5. For instance, new file formats were supplied in these collections, so the team had to look for and implement new tools for file format identification. Also due to new preservation policies, like instant checksum calculation, registration of information about the process and broader use of PREMIS for event metadata, many new features had to be developed to be able to ingest these two collections. One can imagine that with the aforementioned drastic decrease of staff time dedicated to preservation and DM1.5, the fragmented work of the other employees, and the leakage of collective knowledge about preservation, this work took quite some time and had to compete with day to day work.

Or to put it another way and overall, how much as we know as preservation practitioners about our  area of expertise, apparently it is still difficult to use this knowledge to get our preservation requirements implemented within our own organization, especially when the organizational culture is rather formal [3].

Getting Ourselves Together

Renewed organizational focus on digital preservation

After a number of years of struggling with the lack of focus on digital preservation, all the organizational changes mentioned, the next major organizational change announced itself in 2017.

The former organizational integration, that took place in 2015, came short on preservation activities needed to keep pace with the ever changing publishing world. With increasingly higher volumes and different types of digital publications, the library was putting itself behind. Refocus on preservation was needed to keep KBNL’s digital collections safe.

According to the annual report 2017 of KBNL [39] , the organizational structure had to be  changed to create greater effectiveness. This decisiveness was necessary to achieve the ambitions of KBNL as the national library of the Netherlands. Besides that, management acknowledged that the (temporary) facility DM1.5 to preserve KBNL’s digital collections was in need of replacement and required an increase in scale. Therefore, the organization reinstated a dedicated digital preservation department in 2017 to renew its focus on digital preservation.

The newly founded department focused on defining preservation policies including a vision of how digital publications should be handled and kept safe. Also, the department got the responsibility to manage and improve the processes and systems needed to perform the preservation activities, such as the maintenance of DM1.5. And with the renewed focus on digital preservation, the department again  actively participated in the national and international network, such as the Dutch Digital Heritage Network (in Dutch: Netwerk Digitaal Erfgoed (NDE)) [40], the Open Preservation Foundation (OPF) [41] and the International Internet Preservation Foundation (IIPC) [42].

Also a great deal of attention was paid to preparations for the renewal of the “Digitaal Magazijn”. A business case was defined, and a market survey was conducted, in which reputable vendors were asked to give their vision on the next generation of KBNL’s e-Depot. The required improvements were of such magnitude that  a large-scale renewal program was set up, to achieve all the goals for an improved “Digitaal Magazijn”. This program was called the NDM program (in Dutch Nieuw Digitaal Magazijn, which means New Digital Warehouse).

NDM Program

The NDM program [43] for improving the “Digitaal Magazijn” started in 2018. The main goals of the NDM program were:

• Renewing KBNL’s enterprise architecture on business processes, information and application levels

• Renewing KBNL’s storage solution and migrating all digital collections to the new storage system

• Renewing KBNL’s digital archiving solution for preservation of and access to our digital collections

• Bringing preservation policies, planning and actions to a higher level: from bit preservation to functional preservation and gain certification as trusted archive

• Migrating all born digital collections from DM1.5 to the new archiving solution

• Ingesting KBNL’s digitized collection and web archive to the new archiving solution

• Ingesting new digital collections to the new archiving solution, such as e-newspapers, e-magazines and opticals

The program had a duration of six years where originally four years were foreseen. Moreover, the scope has been adjusted severely, limiting its outcomes. The program ended in November 2023 with the following results:

• a renewed and improved enterprise architecture focused on preservation and registration of KBNL’s (digital) collection,

• a purchased and implemented new storage solution based on an object store architecture,

• a selected and implemented Clarivate’s/Ex Libris’ Rosetta as a new Digital Asset Management and Preservation System after carrying out a European tender process [44],

• preservation policies at strategic, tactical and operational level [45],

• a certified e-Depot being awarded the Core Trust Seal [46],

• migration of the digital collections to the new storage solution has been largely completed

• a developed application for pre-ingest workflows,

• migration of the first born digital collection from DM1.5 to Rosetta has started (but we have a long way to go).

It is fair to say that these results were far less than originally envisioned. During the implementation of the program, the scope has been adjusted severely, amongst others due to budget and delay in delivery of parts of the required infrastructure. Preparing and improving the organization for the continuous work on NDM was put out of scope, just like ingesting KBNL’s digitized collection, web archive and new digital collections to the new archiving solution. The scope of migrating all born digital collections from DM1.5 was narrowed to migrating just one collection MVP, so maintenance of DM1.5 is still required. And even so, what has been achieved and delivered took quite some struggles. In the evaluation report of the NDM program entitled "Proud, but not satisfied" [47], the organization looks back on the one hand with pride on the working applications and solutions that have been delivered during the program, but on the other hand, employees of the library are far from satisfied with where we are and what still needs to be done. In such trajectories, there are always many dilemmas, and sometimes also choices in which a certain perspective “loses out”. The ratio between speed, quality and cost is such a well-known trade-off.

The lessons learned from the NDM program are set up as general recommendations, to be able to break organizational patterns. In terms of giving direction, it is important to have a good overview and keep focus on the (real) issue we are trying to solve. This means that we have to constantly think about the question what do we want to achieve, for whom and why? This reflection takes into account different perspectives and conscious well-founded considerations. This helps to stay on course during the process and also provides clarity to everyone who participates in the process. In terms of collaboration and having good discussion, it is important  that there is equality and that everyone is being seen and heard. This means we need to become more skilled in using all the knowledge and expertise within the organization, but once a direction has been chosen, we must stick to it together. This requires mutual trust and respect for each other's expertise. And finally, we need to get better at learning from evaluations in the past, and apply the lessons learned, so that we don't keep making the same mistakes over and over again.

So... how do we move on? Are we ready for the future?

In general, the challenges to preserve the enormous amount of digital heritage for the long term are bigger than ever before. While the volume and diversity of digital collections have become immense they also are becoming the most prominent reflection of heritage in the modern world. An increasing amount of publications made today does not have a physical source anymore as web sites, ebooks and digital newspapers have become the default. Losing this will result in a digital black hole of our written history.

More specifically, with the outcomes of the NDM program in mind,  KBNL has to migrate the born digital collections from DM1.5 to the new archiving solution and start ingesting its digitized collection, web archive and new digital collections to the new archiving solution. Furthermore, we want to make the switch to functional preservation, so we have to implement our new preservation strategies, like just-in-time, risk analysis and the learning approach instead of prescribing.

But there are also opportunities. Many organizations struggle with the same issues in digital preservation, so we can work together on these issues. Collaboration is inspiring for getting new ideas and for validating current ideas. We can work together to create or improve tools. We can discuss problems and find solutions together. And we can  share our experience and knowledge on digital preservation with each other, because we are all learning all the time. And digital preservation is not a world unto itself anymore.  Focusing on future access and re-use also brings our efforts and intentions  with digital preservation into  a broader perspective.

But what have we learned as an organization from thirty years of digital preservation within our institution? With my personal reflection on the digital preservation highs and lows of KBNL, in my opinion  we can only try to sustain digital preservation within our own (rapidly changing) institutions to the best of our abilities, especially where there is tension between the long term goals of digital preservation and all the other objectives at an institutional level. To be able to do that, we have to understand our organizational culture, in order to create room for diversity, because preservation in organizational context is sociotechnical, and solutions need to be adapted to social norms and values within the organization [3].

On the one hand we still have to stress the importance of preservation within our organization, by giving presentations, networking, advice, guest lectures, etcetera, to make clear that preservation demands long term commitment. But on the other hand, we need to be creative, and find different ways to get our message across and have good discussions, because preservation principles like integrity and authenticity are sometimes difficult to understand for other people in our organization.

Further we try to achieve our preservation goals by attaching them to existing goals within the organization, because digital preservation is not a world unto itself anymore. For instance by including preservation in our general policies [48] and content strategy [49], by explicitly attaching preservation policies to KBNL’s other objectives and publish these policies on our corporate website [45], and by showing the importance of certification to make clear what the organization needs to do in order to sustain a trustworthy digital archive [46].  In this way we try to create ways of being involved without taking over responsibility.

And in terms of the formal versus the learning organization, we try to create awareness about different methods and expected results. The NDM evaluation report is an example of this approach [47].

But in my opinion the most important lesson we all have to learn, is to honestly share  our experiences on how we  all try to sustain digital preservation within our own (rapidly changing) organizations,  and share with the community what worked out well, and what didn’t. After all, we are all learning all the time.

ACKNOWLEDGMENTS

Thanks to Daniel Steinmeier and Jeffrey van der Hoeven for their encouragement to tell this story, for reviewing the first draft of this paper, and all their great suggestions. I would also like to thank Patrick for his enormous patience.