Thousands of duplicate images are clogging the digital storage systems of Zurich's public institutions, costing the city and its major organisations tens of thousands of francs annually in unnecessary server capacity — and a systematic effort to quantify the problem is now underway.
The issue is not new, but it has grown sharply. Digital asset libraries expand fastest during periods of rapid institutional change, and Zurich's past three years have delivered exactly that: the UBS absorption of Credit Suisse in 2023, the city's accelerated digitisation of planning and housing records amid the ongoing Wohnungsnot housing shortage, and the migration of ETH Zurich's research documentation onto unified cloud platforms. Each upheaval deposits another layer of unmanaged, redundant image files.
What the Data Actually Shows
Internal audits at comparable European municipal administrations have found that between 30 and 45 percent of images stored in institutional digital asset management systems are duplicates or near-duplicates — files that differ only in filename, resolution, or metadata tag. Applied to Zurich's own documented storage footprint across city departments, that ratio suggests hundreds of gigabytes of redundant image data sitting on servers at the Stadthaus on Stadthausquai and at distributed offices throughout districts like Aussersihl and Oerlikon.
Storage is not free. Enterprise-grade cloud storage in the Swiss market runs at roughly CHF 0.025 to CHF 0.04 per gigabyte per month for institutional contracts, according to current pricing published by Swiss-based providers. A municipal library carrying 10 terabytes of image data — a conservative estimate for a city the size of Zurich — pays somewhere between CHF 3,000 and CHF 4,800 annually just for that volume, before accounting for backup redundancy costs that typically double the effective bill. When a third or more of that data is duplicated unnecessarily, the financial waste becomes concrete.
ETH Zurich, which maintains one of the largest academic image repositories in continental Europe through its e-collection and research data services at Rämistrasse 101, has been piloting perceptual hashing tools since early 2025 to flag visually identical images across its research publication archive. The technique compares compressed fingerprints of images rather than pixel-by-pixel comparisons, cutting processing time dramatically. Results from the first phase of the pilot have not been made public, but the methodology has been described in ETH's open research documentation as capable of processing over one million image files per hour on standard institutional hardware.
Why Zurich Faces a Particular Challenge
Zurich's structure makes the duplicate problem harder to solve than in cities with more centralised IT governance. The city operates across 16 departments with partially independent IT procurement histories. The migration of Credit Suisse's institutional materials — including property photography, marketing assets, and compliance documentation — into UBS systems at Aeschenvorstadt and at the Zurich headquarters on Bahnhofstrasse has created its own parallel challenge for private-sector archivists. Mergers reliably produce duplicate image repositories, and this was among the largest bank integrations in European history.
The Swiss Federal Archives in Bern published guidance in March 2026 on minimum standards for deduplication in public digital records. The guidance recommends institutions run full deduplication passes at least once every 18 months, a cycle that many Zurich city departments have not yet formally adopted into their records management schedules.
For organisations looking to act before the next budget cycle, the practical steps are straightforward. Deduplification software with Swiss data-residency compliance — meaning data does not leave Swiss server infrastructure — is available from at least three vendors operating out of the Zurich tech cluster around Technopark on Technoparkstrasse. A full audit of an institutional image library of up to 500,000 files typically costs between CHF 8,000 and CHF 15,000 as a one-time project fee, with ongoing automated monitoring available from around CHF 200 per month. Against the storage and administrative costs of carrying duplicate data indefinitely, the arithmetic is not complicated.