The Stockfilm Visual Memory Index

Key findings

• The 1960s dominate what survives: 41.9% of the searchable catalog was shot in that one decade. The 1930s and 1940s together account for 11.3%.
• Archival demand is time-and-place specific. 19.9% of searches name an exact year and 25.7% name a place, which is why coverage gaps are felt so sharply: a search for one town in one year either finds surviving film or it does not.
• The clearest coverage gap is not a subject but a craft element: title cards and film leader draw 0.9% of searches against 0.3% of the catalog, a gap index of 2.88.
• 6.1% of searches return no results at all. Searches naming the 1990s meet a catalog with almost nothing to show: the archive effectively ends in the late 1980s, when home video replaced film.
• Licensing does not simply follow supply. School and graduation footage is licensed at 2.4 times the catalog base rate, and 19.7% of the catalog is 1970s film that draws a disproportionate share of decade-specific demand.
• Provenance can be stated precisely: 46.1% of searchable clips trace to one of 1,615 cataloged physical source reels.

1. What survives

The measured population is the searchable Stockfilm catalog: 162,603 published clips, cut from 217,560 restored digital masters, of which 115,631 are available for direct on-site licensing. Definitions for each population are in the methodology.

Survival is not evenly distributed in time. Amateur film follows the consumer history of the medium: 8mm arrives in the 1930s, peaks with Super 8 in the 1960s, and collapses in the 1980s when video takes over.

Geography is even more concentrated. The catalog spans 121 countries and 935 cities, but the United States accounts for 77.2% of located clips. That is a fact about whose home movies were made, kept, and later acquired, and it is the single most important bias to carry into any use of these numbers.

Runtime is reported as summed published clip length, about 299 hours. Clips cut from the same source master can overlap, so this figure is not unique archive time and we do not publish it as such.

2. What people search for

Demand data covers 5,145 on-site searches (2,900 distinct queries) between 2026-02-27 and 2026-07-14. This is one archive's audience over 138 days, published with that scale visible rather than dressed up as market research. Raw query text is never published; queries are mapped to a versioned subject taxonomy, and any theme with fewer than 10 searches is suppressed.

The long tail matters as much as the top themes. Most searches do not map to any broad subject at all: they name a town, a year, or both. People do not come to an archive of home movies looking for a mood. They come looking for a particular past.

3. Where demand outruns the surviving record

The Coverage Gap Index compares a theme's share of searches with its share of the searchable catalog. A value above 1 means people ask for the subject more often than the archive can show it; below 1, the archive runs deeper than demand. It is a ratio of two observed shares, not a forecast, and it inherits the small demand sample described above.

Two gaps deserve plain-language description. First, the demand for editorial craft material: searches for title cards and film leader outrun the supply of surviving examples by nearly three to one. Second, the edge of the medium itself: the catalog holds almost nothing after the late 1980s, while searches keep arriving for 1990s subjects. Where filmmakers need what did not survive here, the footage request and footage brief workflows exist to search beyond the published catalog.

4. What actually gets licensed

3,606 clips, 2.2% of the catalog, have been licensed at least once over the catalog's lifetime on the marketplace. That lifetime basis is the honest one available: it cannot be time-windowed, so it answers "what kinds of footage have ever earned a license" rather than "what sold this quarter." Stockfilm launched direct on-site licensing in 2026. The program has completed fewer transactions than this report's minimum publishable cell size, so direct-license outcomes are not segmented in this edition.

By subject, the strongest lifts over the 2.2% base rate:

Work with the footage behind these findings

Every theme in this report corresponds to searchable, licensable footage. Start from a theme, or bring a brief we can research against the full archive.

• Family life and children82,704 clips
• Christmas and winter holidays6,964 clips
• City streets and urban life23,154 clips
• Cars, driving and road trips24,496 clips
• Nature and the outdoors52,427 clips
• Birthdays and parties9,010 clips
• Military and wartime5,734 clips
• Title cards and film leader515 clips

Data downloads and citation

The aggregate data behind every figure is downloadable. Files are immutable per release; corrections ship as a new version with a change-log entry.

• supply-by-decade.csv
• supply-by-country.csv
• demand-by-theme.csv
• coverage-gap.csv
• licensing-by-decade.csv
• release.json (the full aggregate snapshot)

Cite as: Stockfilm Visual Memory Index 2026, v1.0. Stockfilm Research. https://stockfilm.com/research/visual-memory-index/2026

(c) Stockfilm. Figures and aggregate data may be quoted and republished with attribution and a link to the report. Formal open-data licensing is under review.

Limitations

• Demand reflects 5,145 on-site searches over 138 days from one archive's audience; treat theme shares as directional, not as market research.
• Search click-through data was excluded from this edition: an automated scraping incident in March 2026 contaminated the click log.
• Collection bias: the archive is dominated by North American home movies, so supply shares describe this archive, not the surviving amateur-film record at large.
• Country and city values are display-normalized strings, not ISO codes.
• Summed clip runtime is not unique archive time; clips cut from the same source master can overlap.
• Searches naming the 1990s (9) have almost no matching supply: the searchable catalog holds 13 clips dated in the 1990s.

Full definitions, filtering rules, and known biases: Visual Memory Index methodology. How the archive itself is preserved and described: Archive Integrity and Archive Methodology.

Release history

Editions