How we source events
Nothing on chi·feed is typed in by hand, and nothing is invented. Every listing traces back to a named upstream source, and this page says which ones, what each contributes, and exactly what chi·feed adds — and doesn't — on top.
The short version: we pull from six public event feeds every three hours, merge duplicates, place each venue in a Chicago community area, and link every listing back to its source for tickets and the full description. We don't rewrite descriptions, we don't sell tickets, and we don't take money for placement.
The sources
- Ticketmaster (Discovery API)
- The backbone: arena and theater shows, sports, and most ticketed concerts. Read through Ticketmaster's official developer API. Timed-entry attractions that list every half-hour slot as a separate event are collapsed to one listing per day.
- Do312
- DoStuff Media's Chicago site — nightlife, smaller music rooms, comedy, and free events with a younger skew that the big ticketing feeds miss. Do312 itself re-lists Ticketmaster and Eventbrite inventory, so this is where cross-source de-duplication does the most work.
- Choose Chicago
- The city's official tourism calendar. Strong on exhibitions, festivals, and long-running attractions. Multi-day exhibitions that appear once per day are collapsed into a single date-range listing.
- Eventbrite
- The long tail: classes, meetups, pop-ups, crawls, and independent organizers. Read lightly from Eventbrite's public browse pages — a handful of date-sliced requests per refresh, throttled — with ticket prices filled in from their open pricing endpoint.
- DCASE — City of Chicago Department of Cultural Affairs & Special Events
- Millennium Park, the Chicago Cultural Center, and citywide DCASE programming. Almost everything here is free. Recurring series ("Fridays, Dec. 4, 11 & 18") are kept as one listing with the schedule shown, rather than expanded into separate rows.
- Chicago Park District permits (Chicago Data Portal)
- The Park District's outdoor event permit ledger, published as open data by the City. Filtered to the public-facing permit classes (festivals and 10,000+ attendance events) — the raw ledger is mostly athletic-field reservations. A festival that appears as hundreds of permit rows becomes one listing.
- Reader submissions
- Anyone can submit an event. Submissions land in a review queue and are checked by a person before they appear.
The news tab is separate from events: it shows headlines and summaries from local outlets' public RSS feeds (Block Club Chicago, WTTW, the Sun-Times, the Reader, NBC Chicago, CWB Chicago, Loop North News) and links straight to the article. We don't republish articles, and that tab is marked not-to-index for search engines.
What chi·feed adds
- Cross-source de-duplication. Two listings are treated as the same event when their cleaned titles are at least 90% similar, their start times are within ninety minutes (or the same calendar day for all-day listings), and the venue matches by name or sits within about 150 meters. The richer record is kept and the other's ticket link is attached to it, so a merged listing shows every place to buy.
- Neighborhood resolution. Street addresses are geocoded once (via OpenStreetMap's Nominatim, cached so no address is ever looked up twice) and placed in one of the City's 77 official community areas. Suburban venues stay neighborhood-less rather than being forced into the map.
- Normalization. Each source describes events its own way; chi·feed maps them to one set of categories, one price range, a free flag, and a ticket availability state (sold out, few left, postponed…).
- Ledger collapse. Timed-entry slots, per-day exhibition instances, and permit rows are folded into one listing each, as described above.
- Images. When a source ships no image, we fetch the listing's own social-preview image so the card isn't blank.
- Human corrections. Readers can flag a wrong time, venue, or description on any listing. An accepted fix locks that field, so the next refresh from the source can't overwrite it. Listings an editor removes stay removed even if the source keeps publishing them.
- The weekly roundup. The editorial posts are the one place we write, not aggregate: which events turned up on more than one feed, what got cheaper or sold out since last week, the free picks, and which venues are busiest — synthesized from the whole feed, then read and approved by a person before publishing.
What chi·feed doesn't do
- No rewriting of event descriptions. The blurb on a listing is the source's own text, credited and linked. For the same reason, individual listing pages are marked not-to-index for search engines by default — they aren't our writing.
- No ticket sales, no affiliate links, no paid placement.
- No aggressive crawling. Each source is read at most once every three hours, with rate limits well under what the source allows, through official APIs wherever one exists.
Refresh cadence and freshness
All sources are re-read every three hours. Events are archived (not deleted) after they end, so old share links still open. A listing's page shows the source it came from and every merged source's link, and each connector's last run is monitored so a source that goes dark is noticed rather than silently missing.
Corrections and removals
If you run a venue or an event and something here is wrong — or you'd rather your event weren't listed at all — use the "Report an issue" link on the listing, or email us. Removals are honored, and the removal is remembered so an automated refresh can't bring the listing back.