SF City Websites Are Drowning in Duplicate Images — Here's What the Numbers Actually Show

San Francisco's city government is sitting on at least 47,000 duplicate or near-duplicate image files spread across more than a dozen municipal websites, according to a digital infrastructure review completed last month by the Department of Technology. The redundancy is not merely an aesthetic problem. It is eating server storage, slowing page load times for residents trying to access permit portals and benefit applications, and generating recurring cloud hosting costs that city analysts estimate have exceeded $280,000 annually across affected systems.

The timing matters. San Francisco is in the middle of a broader push to modernize its digital services under the SF Digital Services program, which has been operating out of City Hall since 2019 with a mandate to consolidate the patchwork of agency websites that evolved independently over two decades. Duplicate image proliferation is a direct symptom of that fragmented history — departments uploaded their own photography, logos, and graphics without any shared asset library or standardized content management protocol. Now, as the city eyes a unified content platform, the image problem has become a concrete metric for measuring just how sprawling the cleanup job actually is.

What the Data Shows

The Department of Technology audit, which covered the period from January 2024 through April 2026, found that the SF.gov portal alone contained roughly 12,400 image assets, of which approximately 31 percent — about 3,800 files — were exact or visually near-identical duplicates. The San Francisco Municipal Transportation Agency's public-facing site added another estimated 6,200 image assets, with a duplication rate the audit placed at 28 percent. Muni's site has historically been updated by multiple teams working on Caltrain integration announcements, BART connection schedules, and internal communications, each uploading their own versions of maps and route graphics.

Storage costs are only part of the equation. Web performance benchmarking conducted by SF Digital Services found that pages on sfgov.org carrying unoptimized duplicate imagery loaded an average of 2.3 seconds slower than pages where assets had been deduplicated and compressed. For residents filing building permits through the SF Planning Department's online portal on Seventh Street — a system that logged more than 89,000 user sessions between October 2025 and March 2026 — those seconds add up. Research on government digital services consistently links slower load times to higher abandonment rates, particularly among users on mobile devices or lower-bandwidth connections in neighborhoods like the Tenderloin and Bayview-Hunters Point.

The problem is not unique to San Francisco. Cities that have undergone similar audits — including Chicago, which published a comparable content management review in 2023 — found duplication rates on government sites routinely running between 25 and 40 percent of total image assets. What makes San Francisco's situation distinctive is scale: the city operates more than 50 separate web properties, from the Recreation and Park Department's booking system at McLaren Park to the Office of Economic and Workforce Development's small business portal.

What Comes Next

SF Digital Services has proposed a phased deduplication and migration project, with Phase One targeting the ten highest-traffic city websites before the end of fiscal year 2026–27. The estimated cost for Phase One, according to budget documents presented to the Board of Supervisors in May, is $1.4 million — a figure officials argue would be recouped through reduced cloud hosting fees and avoided web development rework within three years.

For residents and small businesses that interact with city services daily, the practical upshot is faster, more reliable web pages — something that matters on Cesar Chavez Street just as much as it does in the Financial District. The Department of Technology plans to publish a public dashboard tracking deduplication progress by the end of August 2026, giving advocates and supervisors a running count of how many duplicate files have been resolved and what storage savings have materialized. That kind of transparency will be the real test of whether the audit numbers translate into accountability — or just become another set of statistics filed away in a municipal report.