A data journalist treats a spreadsheet, a government database, or a leaked dataset the way an investigative reporter treats a document cache or a network of sources: raw material that has to be interrogated until it gives up a pattern nobody had previously assembled. The raw material is different in kind from the rest of the field, though — structured or semi-structured numbers (health records, procurement contracts, campaign finance filings, crime statistics, climate sensor readings) rather than human testimony or a paper trail, and the skill set follows that difference. SQL is the baseline for extracting from structured government or commercial datasets; Python and R do the cleaning, statistical analysis, and automation, with Python more common for scraping and pipeline work and R more common for the statistical and visualization side [current_fact, consistent with industry hiring patterns 2026]. The output is frequently not just a story but a public tool — a searchable database, an interactive map, a "look up your own address" feature that lets a reader find themselves inside the story rather than just read about it.
Much of the daily texture looks less like the popular image of journalism and more like a data analyst's day: writing queries, cleaning messy government exports, cross-referencing two datasets nobody else thought to combine, chasing a discrepancy that turns out to be the whole story. The Discovery gradient dominates because the essential move is finding a pattern that was sitting in public or semi-public data the entire time — nobody had looked, or nobody had looked with the right method. The Explanation and Expression layers arrive after the finding: turning a dataframe into a chart, a map, or a sentence that a reader with no statistical training can actually use.
The role usually sits embedded inside an investigative team or a dedicated data desk — translating a finding for reporters who add sourcing and narrative context — or publishes data-led stories directly, with the data itself carrying most of the evidentiary weight a beat reporter would otherwise build through sources.
Kitsune can talk through anything on this page — whether it might suit you, what to do next, questions this page doesn't answer. Everything here is yours to read either way.
The pay premium for coding literacy is real and specific to this specialization rather than journalism generally: 2026 US figures put average data journalist pay in the roughly $97,000–$108,000 range, well above a typical beat reporter's baseline [survey_aggregator, ZipRecruiter / Glassdoor 2026] — though the figures blend genuinely distinct roles and should be read as directional rather than precise.
The entry pattern runs backwards from the rest of the field. A meaningful share of data journalists never attended journalism school at all — they arrive from computer science, statistics, or GIS backgrounds and pick up sourcing ethics and verification standards on the job, the reverse of the traditional beat-reporter pipeline. NICAR (the data journalism conference run alongside Investigative Reporters and Editors, held most years in early spring) and the IRE conference itself function as the field's actual professional home and training ground more than any single degree [official_body, IRE/NICAR conference programming, 2026]. AI tools are reshaping the routine end of the work — first-pass cleaning and basic data-pulling are increasingly assisted — but formulating the right question and judging whether a discrepancy is a real story or a data artifact remain durably human skills, the same pattern this file's own Supply Chain Analyst archetype describes for a structurally similar role in a different field.
There is no single credential gate. Common routes run in both directions: journalism training plus self-taught SQL and Python, often via NICAR's own bootcamps and online courses; or a computer science, statistics, or GIS degree plus a newsroom internship that teaches sourcing and verification on the job. A public portfolio of data-driven stories (commonly hosted on GitHub) tends to matter more to hiring editors than a journalism credential does. Entry points include dedicated data desks at outlets like ProPublica, The Washington Post, the Associated Press, and Reuters, as well as smaller nonprofit newsrooms building out their first data capacity.
Entry pipeline runs backward from the rest of the field (CS/stats/GIS backgrounds more common than journalism school); AI lowers the technical floor for generalists while raising the ceiling for specialists, and which effect dominates is unsettled
Growing; scheduled and leaked datasets are increasing in volume and the technical-plus-judgment combination remains scarce, reflected in a real pay premium over generalist beat reporting
People drawn to Data Journalist / Computational Journalistare often drawn to these — in the order they're closest. The ones marked sit in a different field entirely.