Data-Led Digital PR for Games, Sport and Media

Six data-led campaigns run last year earned between 14 and 61 referring domains each, median 29, at a cost between 310 and 900 EUR per earned domain. I quote against that band and against nothing else, and the campaign that returned 14 is in the reference sheet alongside the one that returned 61. The asset is the product: A campaign is one piece of information a journalist cannot obtain elsewhere, presented so it can be written up in twenty minutes. Everything else, the pitch, the page, the follow-up, is distribution. Practices that invert this end up with a good pitch attached to nothing, which is why so much of this market reports open rates instead of coverage. Where the numbers come from: Your own systems. Match telemetry, matchmaking logs, ticketing, refund and chargeback rates, CMS publishing data, community moderation volumes, churn by cohort. In games and sport these datasets describe behaviour nobody else can measure, and a journalist who covers the sector knows immediately that they are real. Extraction is done with your data team under a scope note that names every field leaving your environment, and aggregation thresholds are set so no individual is identifiable. What I refuse to build: No commissioned survey of a thousand respondents dressed up as a study. Sports and games desks discard those on sight now, and rightly. No index that ranks cities or countries on a metric assembled to make your brand the natural quote. No press release distribution. No HARO, Featured or Response Source replies, which produce links from sites that exist to consume them. No embargo offered to eleven outlets at once and called an exclusive. How it is seeded: A first-look exclusive to one correspondent who covers this beat, chosen because of what they published in the last quarter rather than because of a title's traffic. Twenty-four to forty-eight hours later the full dataset goes to a list of between forty and ninety named journalists, built in Prowly and Muck Rack and then hand-verified, because roughly a fifth of any bought media list is at a different employer. Follow-up is once, with something added, or not at all. What failure looks like: About one campaign in four underperforms its band. The usual cause is that the finding was interesting internally and unremarkable outside, and the honest time to catch that is at the data review before the asset gets built. I will kill a campaign at that gate and rebuild the angle from the same dataset rather than proceed and blame the news cycle. That gate is written into the schedule as a stop point with your sign-off. Who this is not for: Companies that will not release operational numbers. This work has no substitute input, and without it what is left is a survey, which I do not do. Also not for anyone who needs coverage in a fixed month, because a strong campaign will be held for a news window and a weak one cannot be forced into one. And not for brands in a live regulatory dispute, where publishing internal data is a decision for your counsel and not for a search consultant. Reporting: Earned URLs with the journalist, the outlet and the link attribute for each. Pickups that mention you without linking, tracked separately, because those still move branded search and I will not pretend otherwise. Referring domains that survive at 90 days, since a share of syndicated pickups will not. Cost per surviving domain against the band quoted at the start.