311 case data is one of the most used datasets on the open data portal. While the fields are fairly straightforward, we get a lot of other questions about the data itself. Below are some commonly asked questions to help you in your exploration of the 311 data.
What is 311 and what's in the case data?
The City and County of San Francisco established SF311 to provide an easy-to-remember telephone number that connects residents, businesses, and visitors to Customer Service Representatives ready to help with general government information and services in 2007. SF311 started taking web requests in 2008, social media in 2009, and mobile requests in 2010 from apps that connect to the open311 API. Departments can also use the 311 system to report inter-agency issues.
This dataset includes cases generally associated with a place or thing (for example parks, streets, or buildings) and created July 1, 2008 or later. Cases generally logged by a user regarding their own needs (for example, property or business tax questions, parking permit requests) are not included. Cases prior to July 1, 2008 are not included due to significant data structure changes in the spring of 2008. General Requests are multipurpose requests that can be routed to the majority of City Agencies. SF311 case data is provided as is and without warranty but works with its departmental partners to continuously improve business processes and data quality.
The following is a list of the included case types:
311 External Request
Blocked Street or Sidewalk
Catch Basin Maintenance
Construction Zone Permits
DPW Volunteer Programs
General Request (Multiple Agencies)
Rec and Park Requests
Residential Building Request
Sidewalk or Curb
Street and Sidewalk Cleaning
Temporary Sign Request
How often is the data refreshed?
Data is refreshed every morning at approximately 10am. In the event of a publishing failure, the system is configured to catch up on the subsequent day.
How do I download the data?
All datasets on the open data portal can be downloaded via the export button on the top right of the dataset landing page.
Why don’t all the cases have a Latitude and Longitude for mapping?
Sometimes cases occur at a location without a valid street address. These cases are forwarded to the departments with a description of the location instead of a verified street address.
How can I identify duplicate requests?
311 captures requests for service from the public, and the public is unaware if the issue they see has already been reported. Often times a single incident generates multiple request. Some incidents do not generate any requests. The subjectivity of public (crowdsourced) data makes trending and identification of duplicate requests difficult.
Users of the 311 data should understand each report may not be unique, such as an overflowing trashcan, but instead treat it as a data for unique complaints around an incident. For example, four 311 cases for litter could be any of the following:
Person A submitted 4 separate requests for the same piece of litter
Person A,B,C,D submitted individual requests for the same piece of litter
Person A & B submitted individual requests for a piece of litter at 10am. Litter gets picked up. Person C & D submitted individual requests for new piece of litter in same location again at 11am
Person A submitted individual requests for 4 different pieces of litter they saw on their walk to work
In all the above cases it is accurate to say there were Four 311 service requests. It is not accurate in all the above cases to say there were four pieces of litter.
Why is the text or photo I submitted through the app not available in the dataset?
SF311 is committed to protecting individual privacy rights, and may choose to retract pictures related to some classes of cases where publishing personally identifying photos would infringe on these rights. As such, SF311 tries to anonymize all the data published to the open data portal by publishing the Category, Request Type, and Request Details of all requests, which provides transparency on the volume of cases and what is being reported. We're unable to automatically verify that personal information is not included in the description, so the description is not published.
For example residents using the 311 app have the options of providing contact information. In general, we advise users their contact information is for the purpose of departments interacting with them to fulfil the request. Users can either use the mention (public) process or direct message (private) process to contact us. While we store the contact information in the database, we don’t associate it to the request on the Open Data portal.
What do the different columns represent?
Each column has properties which are available by clicking the (i) in the column header. The properties include the data type as well as a definition of the columns. For example:
Status Notes - Explanation of why status was changed to current state or more details on current status than conveyed with status alone
Responsible Agency - The agency responsible for fulfilling or otherwise addressing the service request.
How do I use the Source Column?
The column shows the Mechanism or path by which the service request was received; typically ‘Phone’, ‘Text/SMS’, ‘Website’, ‘Mobile App’, ‘Twitter’, etc but terms may vary by system.
If the source shows Integrated Agency it means that a city department created an ‘internal’ request to another city department. For example, if Public Works were tasked to clean sidewalk of loose garbage and came across a refrigerator which is considered a ‘bulky item’ that needs special equipment, they’ll create a request to ‘Recology.’
Why is the category of my request different than I see on the web or mobile?
The categories on the web and mobile application are intended to be clear options to the public, but they do not necessarily map to a form (Category in the open dataset) by the same name. Some categories, such as ‘Neglected Building (Blight)’ are recorded on the General Request form. Graffiti and Litter requests map to multiple forms.
Why are there strange spikes in the data?
SF311 service requests are entered by customer service representatives on behalf of customers, by customers themselves, or by departments as a result of work they are performing. For this reason, the data is subject to spiking based on interest by an individual or organization related to a process, or by a department performing or publishing additional work in an area.
For certain columns there have been coding changes. You can find documentation of each of these changes in the 311 dataset changelog
Can I see this data in my neighborhood?
Yes, check out the 311 explorer to view cases in your neighborhood.