Original Data Studies Built From Your Operating Records

A study earns citations when it holds one number a reporter cannot get anywhere else. Not ten numbers. One. Everything else in the document exists to make that number checkable. Where the number comes from: Your own systems, most of the time. The exports that produce a defensible statistic in these sectors are dull: TMS lane and rate history, WMS pick and dwell times, CMMS work orders and mean time between failures, historian availability by asset class, quoting and RFQ win-rate history out of the ERP. A carrier's own dwell data across four hundred receiving docks is a figure no analyst house holds. Where the internal data will not carry the finding, we field a survey to a named population — an association membership, a licence registry, a buyer list — and publish who was asked, how many answered, and the response rate. A panel survey of a general audience dressed as industry research is not something we will put a methodology page on. Disclosure control: Before anything is calculated, we agree the aggregation level at which no individual customer, site or contract can be identified. Any cell built from fewer than five contributing facilities is suppressed. Rate data is published as an index against a base period rather than in dollars. This step decides whether your legal and commercial teams will allow the study to exist at all, so it happens first rather than as a review at the end. What gets published: A findings page written so each claim sits beside the figure supporting it. A methodology page carrying the sample definition, collection window, exclusions and the questions asked verbatim. An aggregated CSV under CC BY 4.0 — that licence line is what turns a reuse into a link. A chart pack under the same licence, because a chart lifted into a trade article without a data file gets redrawn and unattributed. Schema.org Dataset markup so the release surfaces in Google Dataset Search, and a Zenodo deposit with a DOI where academic citation is realistic, which in energy and materials it usually is. The launch: Three to five reporters are briefed under embargo a week ahead, individually, with the file rather than with a release. Whoever is briefed gets the number before it is public, and that is the only reason they clear space for it. Association newsletters and standards committee circulars go out on publication day; in these industries a committee mailing moves more qualified readers than any technology outlet will. What we refuse: A study whose conclusion is agreed before the data is pulled. If the number can only come out one way it is advertising, and putting a methodology page on advertising is how a company loses a trade journalist permanently. Rounding a finding until it flatters. Infographics as the primary asset — they get shared and not cited. Re-releasing last year's figures behind a new cover. The risk you are carrying: The finding may be dull, or unflattering to your own operation. You hear which as soon as the data is clean and before any publication work starts, and you can stop there with the analysis in hand. Around one study in five that we scope does not proceed for exactly this reason. That is the cost of the ones that do get cited. Repeatability: Value compounds when the instrument is run again. An index in its second wave gets covered as a trend, which is an easier story to place than a first-time finding. We build the first wave so your team can run the second without us, and we say so in the run book rather than keeping the method in our heads.