How the Industry Review is made

What the review is

The Industry Review is an annual, chapter by chapter account of the semiconductor industry, written by The Semiconductor Hub. Each chapter sets out the Hub's own analysis in plain words, the facts it rests on with their sources, and the Hub data behind it, then says what the chapter means for the people who use the Hub and where to go next on the platform.

It is built to be checked. Every statement is one of three kinds, and each is shown differently, so a reader can always tell what a sentence rests on.

Three kinds of evidence

Analysis. The body text is the Hub's own reasoning, written in our own words. It carries no numbers of its own: wherever a size, a share, a rate or a date matters, it comes from a cited fact or from a figure built on Hub data.

Cited facts. A fact from a named publication is restated briefly in our own words and numbered. Its note gives the publisher, the title, the date and the page or section, and, where the publication names the primary source of a figure, that source too. A forecast is labelled with who made it and the year it was published. We never copy a publication's passages, structure, charts or tables, and we prefer the primary source wherever one is named.

Hub data. Figures are drawn live from the Hub's own records when a page is read or a file is made: datasets in the Data Catalogue, published indicators, the company directory, the fab explorer, the jobs board and the events calendar. Each figure is dated with the newest value it shows and carries its source note. Where the Hub holds too few records to say something, or holds only illustrative sample values, the figure says so and invites contributions. The Hub never fills a gap with an estimate of its own.

Sources

Third-party facts come from credible publishers: research firms and consultancies, industry associations, statutes and regulators, and standards bodies. Each is named where its fact is used. The full list for a chapter is its numbered reference list, and the publications cited across the edition are listed on the landing page.

Hub data comes from the Data Catalogue, where each dataset states what it contains, its coverage, its update cadence, how its values are sourced and what it leaves out, and from the Hub's live registers. Follow the link under any figure to see the data and its notes.

Audit and reviewers

Every chapter is audited before it is marked as reviewed. An editor assigns practitioners or analysts who volunteered through the call for reviewers. Each reviewer reads the chapter, checks its cited facts against their sources, records findings section by section and signs off when the text is accurate, or says what must change.

An editor makes the final approval, which needs at least one reviewer's sign-off and no open finding marked as wrong or misleading. Nobody approves their own contribution. Until a chapter's exact text is approved it shows as not yet audited, a draft for review. When the text changes after an audit, the chapter shows that it has changed since its audit until it is reviewed again. Reviewers who choose to be credited are named on the chapters they audited.

Corrections

Anyone can point out an error. Use the discussion thread of the chapter, or the contact page, and say which section and which fact you mean, with a source where you have one. A correction changes the chapter's text, which raises its version and its revised date and sends it back for review, so the audit status always describes the text you are reading.

Where a cited publication is itself corrected or superseded, the chapter is updated to the newer text and the note says so. Hub data corrects itself: figures are read live, and the Data Catalogue keeps a changelog of restatements.

Who can read what

Visitors can read the contents, every chapter summary, this methodology and one sample chapter in full. Registered members can read every chapter. Registration is free. The PDF of the whole edition and of each chapter is part of paid plans and is watermarked with the member's name.

Independence and limits

The review names no company as better or worse than another and ranks nobody. Where a chapter needs an example, it describes the kind of company or product rather than naming one. Publishers are named only as the sources of the facts they published.

The review describes the industry in general. It cannot know a reader's situation, and a forecast is only as good as its assumptions, which is why each forecast is labelled with who made it and when.