Data-Led Digital PR for Games, Sports and Esports Desks

Journalists in this sector link to one thing: a number they cannot get anywhere else. When the story is built on a dataset with a defensible method, roughly one pitch in three converts to coverage. When it is built on a commissioned poll of 2,000 people, it is closer to one in twenty, and half the coverage that does land carries no link at all. That difference is the entire service. The dataset comes first: Before any outreach exists, there is a number. In games that has meant concurrency and retention patterns pulled from the Steam Web API and Twitch concurrents, prize pool distribution set against player earnings, or store rank history read across a release window. In sports it has meant training load, participation by district, or ticketing and viewership series that nobody has joined up before. The strongest stories use your own telemetry, aggregated and legally cleared, because it is the one dataset a rival newsroom cannot reproduce. I will tell you within the first week if your data does not support a story. That conversation happens before you have paid for outreach, not after. What lands on your domain: A methodology page on your own site, with the sample, the collection window, the known limits stated plainly, a downloadable CSV, and charts licensed for reuse with attribution. This is what journalists cite when they are doing the job properly, and it is why the links point at a page that keeps earning rather than at a press release. A named analyst from your team is quoted in it and reachable for follow-up questions on the day. Who gets the pitch: Named reporters, not publications. Between 40 and 60 people who have written about this specific kind of number in the last eighteen months, found through Muck Rack and Prowly and then verified by reading what they actually filed. Each pitch names the angle that reporter would take, which differs between a trade desk covering the industry and a national desk covering the culture. Embargoes go out mid-week, mornings in the target time zone, never against a patch day, a tournament final or a results announcement. Refused: Wire distribution. Syndication across a newswire network produces a coverage count with no editorial decision behind any of it, and I will not count it. Stunt statistics, reverse-engineered so that the headline exists before the data. Paying for coverage, which is different work and is disclosed when it happens. Pitching a story I do not think stands up, even when the client is certain it does. Who this is not for: Teams needing coverage inside two weeks for a launch that is already dated. A data story takes six to ten weeks from brief to embargo, most of it in the analysis, and compressing it produces the poll-shaped thing described above. Also not for iGaming, and not for brands unwilling to publish a method that a hostile statistician could read line by line. Counting coverage honestly: The log records every piece: outlet, reporter, whether the mention is linked, the rel attribute, and whether it is original reporting or a syndicated copy of another piece. Syndicated duplicates count once. Unlinked mentions are recorded separately and passed into reclamation, since a masthead naming you without a link is a follow-up email rather than a failure. Alongside that, referral sessions to the methodology page and, from month three, the queries that page begins to rank for on its own, which is usually where the durable value turns up.