| Year | Cumulative TB preserved |
|---|---|
| 2015 | 0.4 TB |
| 2016 | 13.3 TB |
| 2017 | 15.6 TB |
| 2018 | 55.0 TB |
| 2019 | 81.5 TB |
| 2020 | 99.5 TB |
| 2021 | 136.4 TB |
| 2022 | 175.8 TB |
| 2023 | 233.2 TB |
| 2024 | 338.2 TB |
| 2025 | 507.7 TB |
| 2026 (through April) | 605.9 TB |
View chart data as table
| Year | Cumulative TB preserved |
|---|---|
| 2015 | 0.4 TB |
| 2016 | 13.3 TB |
| 2017 | 15.6 TB |
| 2018 | 55.0 TB |
| 2019 | 81.5 TB |
| 2020 | 99.5 TB |
| 2021 | 136.4 TB |
| 2022 | 175.8 TB |
| 2023 | 233.2 TB |
| 2024 | 338.2 TB |
| 2025 | 507.7 TB |
| 2026 (through April) | 605.9 TB |
Every two years, we ask APTrust members to step back from the daily work of digital preservation and take stock: What are you preserving? How are you doing it? What's working, what's hard, and where do you want to go? The 2026 Member Census is our most complete picture yet of a community that has quietly grown into something remarkable.
We're sharing those results publicly today. Everything below reflects only the 12 sustaining member institutions that permitted worldwide aggregate sharing. The full interactive dashboard is available at aptrust.org/wp-content/uploads/2026/census_2026.html.
605.9 Terabytes. 446,767 Objects. 38.3 Million Files.
Those are the numbers as of April 2026, and they represent more than disk space; they represent decades of audio recordings, manuscript collections, theses, newspapers, photographic archives, and born-digital scholarship that will remain accessible long after the systems that first created them are gone.
The growth trajectory is just as striking. The APTrust community crossed the 500 TB threshold in 2025, and the past two years alone account for more than half of all content ever deposited. In 2025 alone, members added 169.5 TB, more than in any prior year. The community started from essentially nothing in 2015 and has grown steadily through changing technology landscapes, a pandemic, institutional reorgs, and funding pressures. That consistency matters.
2026 is already off to a strong pace with 98.2 TB deposited through April, and deposit projections for 2026 and 2027 are broadly stable, with most institutions expecting to add 1–50 TB of new content in each year.
Content is stored predominantly in the Deep Archive (66%, ~398.8 TB) and High Assurance (34%, ~203.3 TB) storage classes.
What's Being Preserved
The 12 publicly sharing institutions deposit an extraordinarily broad range of content. Audio, video, and photographic materials are the most common types, each reported by 11 of the 12 institutions. Books and volumes, manuscripts, and newspapers were each reported by 10 institutions. Theses and dissertations, web archives, scholarly communications, maps and GIS data, and email collections each appeared in at least 7 institutions' deposits.
The breadth here is worth pausing on. These aren't narrowly scoped technical archives. This is the record of human knowledge and culture at research institutions, and it's being preserved over the long term.
A Community Rating Its Own Value: 4.8 out of 5
We asked members to rate the overall value APTrust provides to their organization on a five-point scale. The average across the 12 publicly sharing institutions was 4.8 out of 5. Nine of the 12 gave a perfect 5.
We also asked separately about APTrust's role in fostering community and shared knowledge, the convenings, the shared learning, and the peer connections that don't appear on a storage receipt. The average was 4.6 out of 5, with 8 of the 12 rating it a 5.
We don't share these numbers to congratulate ourselves. We share them because they reflect something real: a community that has chosen to invest in shared infrastructure and finds that investment worthwhile, year after year.
Workflows: Where Members Are and Where They Want to Go
The census asked about current and desired deposit workflows, and the gap between the two tells an important story about where the community is headed.
Today, among the 12 institutions: 7 use primarily manual workflows, 5 use hybrid (some automation, some manual), and only 3 use fully automated workflows.
Where do members want to be? Only 2 want to stay with manual workflows. Seven want hybrid, and 12 want fully automated deposit workflows.
Workflow development and deposit automation is already the single most common current initiative, cited by 8 of the 12 institutions. Members are actively working on born-digital and AV pipelines, scheduled deposit cycles, and direct APTrust ingest integrations. This is a community that knows where it needs to go and is doing the work to get there.
What's Working: Strengths Across the Community
When we asked members to describe their digital preservation strengths, three themes dominated:
- Administrative support (8 of 12 institutions): Leadership buy-in, dean and AD backing, and active institutional commitment to digital preservation. This matters enormously; programs without it struggle to sustain momentum through staffing changes and budget cycles.
- Dedicated staffing (7 of 12): Dedicated digital preservation positions, committed individuals, or specialist units. Many institutions noted that having even one person whose job is explicitly about digital preservation changes what's possible.
- Technical expertise (5 of 12): In-house knowledge of preservation systems, formats, and workflows.
These strengths aren't evenly distributed. Some institutions have all three; others are building toward them. The community itself is a resource for that capacity-building.
What's Hard: Shared Pressures
Six of the 12 publicly sharing institutions reported at least one significant challenge in the past 12 months affecting their digital preservation program, a figure consistent with the broader landscape of budget pressures and staffing constraints across academic libraries.
The most commonly cited resource needs centered on staffing and funding. Members described lean teams that must reliably characterize, move, and verify content, often with tools built around that constraint. Format identification and transfer tools are the most frequently reported across the community, followed by forensics and ingest tools. Three institutions also rely on custom in-house systems.
Current initiatives beyond workflow development include expanding APTrust coverage to include institutional repositories, ETDs, journals, and born-digital archives (6 institutions); policy planning and gap analysis (5); repository migration or rebuilding (3); and large digitization and preservation standards projects (2).
A Connected Community
APTrust members don't operate in isolation. Among the 12 publicly sharing institutions, participation in other digital preservation organizations was widespread: 10 belong to NDSA, 7 to CLOCKSS, 5 to Portico, 4 to the Global LOCKSS Alliance, 3 to Perma.cc, and 3 to regional Private LOCKSS Networks. Individual institutions also participate in the BitCurator Consortium, the Internet Archive, and the Digital Preservation Coalition.
This breadth of affiliation isn't redundancy; it reflects a community that takes preservation seriously enough to invest in multiple layers of infrastructure, shared knowledge, and professional development.
See the Full Dashboard
The interactive public dashboard, with charts for every section above, is available at: https://aptrust.org/wp-content/uploads/2026/census_2026.html
Results shown there represent only institutions that explicitly permitted public aggregate and analysis sharing. Individual institution data remains private per each member's stated preferences.
Questions about the census or APTrust membership? Reach us at info@aptrust.org.