Original Data Studies That Earn Editorial Coverage

A thousand-respondent consumer panel survey costs about three thousand dollars to field and earns almost nothing, because any competitor can buy the same instrument next month. The studies that get cited hold a number nobody else can produce. For a direct-to-consumer brand that number is almost always already sitting in your own order data. The question comes before the data: We start from a question a desk editor would commission, then test whether your data can answer it honestly. How fast a colour moves from runway to reorder is answerable if you hold four seasons of SKU-level sell-through. What Gen Z wants from skincare is not a question, it is a headline looking for a survey. If your data cannot carry the question, we say so in the first week and either change the question or end the engagement there, and you pay for the week. What we work from: First-party sources: order lines, returns with reason codes, on-site search logs, waitlist and restock signups, size and shade curves by region. Public sources we combine with them: BLS price series for apparel and personal care, US import records, USPTO trademark filings, Google Trends, and FDA cosmetic facility registrations under MoCRA, which since July 2024 have put a large part of the manufacturing map on the public record for the first time. Joining a private series to a public one is what makes a study hard to copy. The methodology is published, not summarised: Every study states the sample, the exact date window, the exclusions and their effect, and a paragraph on what the data cannot show. We do not round a 3.1 percent difference into nearly a third more, we do not report a subgroup under n of 100 as a finding, and we will not run a survey whose conclusion you have already written. Reporters at the desks worth pitching check this, and the ones who do not get corrected by someone who does. The asset: A page on your own domain. Charts rendered as SVG with the underlying CSV linked beside them, the methodology on the same URL, and the dataset downloadable without an email form. Gated assets do not get cited. We supply chart files sized for print and a plain-language summary a reporter can quote without calling you. Privacy and legal: Aggregation thresholds are agreed with your counsel before the analysis, not after the finding. Nothing is reported at a granularity that identifies a customer or a wholesale account, and legal sees the methodology section before the embargo goes out rather than the morning of. Seeding: An embargo to a short named list, sized to the story rather than to the outlet count, with the dataset attached. Then a second wave carrying a regional cut. Trade press before national in most categories, because trade coverage is what national reporters read to decide a story is real. What we will not build: Infographic link bait. Roundups of other people's statistics. Study pages assembled from secondary sources with a survey stapled on. And we do not write the study after the pitch has been sold, which is the sequence that produces the numbers everybody quietly distrusts. Who this is not for: Brands under two years old rarely have the depth for a trading study, and a survey is the wrong substitute. If your data lives across three systems that disagree on what counts as a return, budget for that reconciliation before this work, not during it. And if you need the study live in three weeks, the answer is no. The median build here is nine to twelve weeks including legal review.