How San Francisco's City Websites Ended Up Buried in Duplicate Images — And What's Being Done About It

San Francisco's Department of Technology is working through a backlog of thousands of duplicate images embedded across the city's constellation of municipal websites — a problem years in the making that has inflated hosting costs, slowed page load times, and muddied the public's ability to find accurate information on portals ranging from SF.gov to the Municipal Transportation Agency's rider-facing pages.

The issue didn't emerge overnight. It traces back to a 2018 migration push when dozens of departmental sub-sites were consolidated under the unified SF.gov platform, an effort managed jointly by the Department of Technology on South Van Ness Avenue and the City Administrator's Office. When content editors moved material from legacy systems, many uploaded fresh copies of images already stored in the central asset library rather than linking to existing files. The result: a single photograph of, say, the Ferry Building or a Muni bus on Market Street might exist in 40 or 50 separate versions across the content management system, each eating storage and each potentially carrying different — sometimes outdated — captions or metadata.

Why It Matters More Now Than in 2018

The stakes have grown considerably. San Francisco expanded its digital services footprint aggressively between 2020 and 2024 as pandemic-era closures pushed residents toward online interactions with city government. The SF.gov platform added roughly 30 new departmental sections between 2020 and 2023 alone, according to the Department of Technology's published digital services roadmap. Every expansion brought fresh opportunities for content duplication, and the asset library grew without a corresponding cleanup protocol.

The practical consequences show up in ways residents notice. A search for permit information on the Department of Building Inspection's pages, accessible from the Civic Center complex on Polk Street, can surface outdated imagery that was never removed after policy changes — images still technically live in the system, just not linked prominently. Meanwhile the San Francisco Public Library's digital branch portal, maintained separately from the main SF.gov stack, accumulated its own parallel image store that only partially syncs with central records.

Storage is money. Municipal cloud hosting contracts — the city shifted significant workloads to cloud infrastructure through a Department of Technology initiative formalized in 2022 — charge based on data volume. While the Department of Technology has not published a line-item figure for image-storage overhead specifically, the city's broader IT operating budget for fiscal year 2025-26 was set at approximately $140 million, and digital asset management inefficiencies are among the line items flagged in internal efficiency reviews the department conducts annually.

The Cleanup Push and What Comes Next

The current remediation effort, underway since late 2025, involves a phased audit. Content managers across departments are being asked to run their sections through a duplicate-detection tool integrated into the Drupal-based CMS that underpins SF.gov. The tool flags images sharing identical or near-identical pixel signatures and queues them for human review before deletion — a safeguard against removing an image that, while visually identical to another, carries legally distinct metadata tied to a specific public record.

The San Francisco Digital Services team, which sits within the City Administrator's Office and has staff embedded at 1 Dr. Carlton B. Goodlett Place, is coordinating department-by-department training sessions through July and August. The MTA, the Recreation and Park Department, and the Office of Economic and Workforce Development are among the larger agencies scheduled for review in the first wave.

For residents and small-business owners who rely on city portals for permitting, benefits enrollment, or transit information, the immediate practical advice is straightforward: if a page looks visually dated or carries imagery that doesn't match current program details — particularly on Planning Department or DBI pages — cross-reference against the dated PDF documents linked at the bottom of those pages, which are typically updated on a stricter schedule than surrounding web content. The cleanup is expected to run through the end of calendar year 2026, at which point the Department of Technology plans to publish updated content governance standards requiring all new uploads to clear a duplicate-check before going live.