Quarterly data study built from your own operating data

A study that works earns fifteen to forty referring domains from publications with named editors. Anything advertising three hundred is counting scrapers and aggregator reposts of the same wire item. One study a quarter is the correct cadence for a single brand, because the second one lands on journalists who already know the first was accurate, and that compounding is the entire mechanism. The data comes from you first: Booking records, rate and occupancy history, the listings feed with its price changes and withdrawal dates, delivery and lead time logs, returns by product category. Aggregated, anonymised past the point where any single property, client or customer can be identified, with the thresholds agreed with your side before analysis starts. Public sources fill the frame rather than carry the piece: Eurostat tourism nights, national statistics offices, land registry transaction records, published rental indices. The distinction matters because a study built only from public data is a study any competitor can replicate the week after and no journalist needs you to explain. What I refuse to build: The panel survey. A thousand adults asked whether they would give up something they were never going to give up, written up as a finding. It manufactures a statistic rather than measuring one, property and travel desks learned to discount it years ago, and it dates within a season. I also will not run the analysis toward a headline agreed in advance. If the data says the opposite of what marketing hoped, that is the study, and in my experience the uncomfortable finding is the one that gets cited. How it is built and published: One question, stated in a sentence. A method section with the sample, the period, the definitions and what the data cannot tell you, published on the page rather than available on request. The raw aggregated dataset as a downloadable CSV, which is what makes a data desk trust you and what makes the piece re-citable for years. Charts as static images with the credit line burnt in, plus a copyable embed block. A permanent URL that is not filed under a dated blog path, because those get pruned in the next migration and take the links with them. Distribution: An embargoed brief to eight to twelve named journalists whose last six months of published work show they cover this specific thing, each one written individually, with the dataset attached and an offer of the analyst on a call. Not a release, not a list buy. Coverage from four of twelve is a good outcome and I would rather tell you that now than describe a hit rate nobody achieves. Not included: Illustration and design beyond the study's own charts, wire distribution, ongoing media relations after the launch window, translation into more than one additional language. Coverage guarantees, which nobody credible offers and which would require paying for placement. Who should not buy this: A company that cannot get internal data released for publication within four weeks. That is usually a legal or ownership question rather than a marketing one and it kills the timeline quietly, so I ask about it before contracting. Also not for anyone who needs the asset to convert directly. A study earns citations and authority for the domain that hosts it; it does not sell rooms, listings or sofas, and judging it on assisted revenue in quarter one will end the programme just before it starts working.