The question is registered before the data is pulled, and every row is published.
These are standing studies rather than articles. Each one names its question and what would make it false before any retrieval, states its method, publishes the rows behind every figure, carries its limits on its face, and is re-run on a schedule. When a measurement overturns an earlier one, the correction is published and dated rather than quietly applied.
CPG
Margin, pass-through, and demand read from the filings.
Film
Formats, windows, and what an audience score measures.
Games
Dated launches, public conversation, published numbers.
Cybersecurity
Buyer-intent questions where a wrong shortlist is expensive.
Marketing
What companies actually spend, from their own filings.
Semiconductors
Where a chip crosses a border, and where the money lands.
Streaming
What happens to attention after the opening week.
Predictions
Every call we registered, including the ones we got wrong.
Most recent
Nobody stopped caring. They stopped buying.
The question: Did the Halloween occasion shrink, or did the basket shrink inside an intact occasion?
What it found: The occasion is at a six-year high and the basket is not. September search for the category ran 15.5% above its own 2019–2024 average while sales fell 6.7%, and decorations fell 58.6%. Across 409 comments about trick-or-treating, “skipping” appeared zero times.
A 94% audience score, and they were talking about Batman
The question: When a film posts the best audience score its studio has had, what is that score attached to?
What it found: Not the film. Across 973 comments with ten terms registered in advance, Batman ran 10.28% and a sequel 0.41%. The craft conversation cannot be told apart from a horror film that cost a seventh as much.
Every call we registered, including the ones we got wrong
The question: Does registering a prediction from a measurement actually beat guessing?
What it found: Yes, by about nineteen points, and not by more. Measured priors held 14 of 19; priors from belief held 18 of 33. This page replaces a published record of 17 for 17 that counted an unresolved prediction as a win.
One opened at twice the other. Three months later it was nine times smaller.
The question: Does a strong opening month predict whether a streaming title still has an audience a quarter later?
What it found: It predicts rank and not retention, and at the top of the chart it inverts. Median retention at three months is 7.3% — and the range runs 2.2% to 44.9%, so the median describes almost nobody.
An entire industry vacated twelve days
The question: Does a generational launch take demand from the other games released beside it?
What it found: It cannot, because there are none. Two published calendars contain exactly one dated title in the twelve days around November 19 — and the twelve days before are the busiest of the autumn.
Every study lives under its vertical above. Nothing here is a summary — each page carries its own rows.
What makes it a standing study
- The question and its falsifier are written down before any data is pulled. A prediction recorded afterwards is worthless, and a question that cannot come out wrong is not a question.
- Every row is published, not summarized. A citation list reported as a percentage cannot be checked by the reader, and an unverifiable figure is an assertion.
- Every figure carries its source, its method and its date. Including the model version, the location a search was run from, and what the pull cost.
- The limits are on the page, not in a footnote. Including what the instrument can never see.
- It is re-run on a stated schedule and corrected in place, dated. A measurement that can be taken again is taken again.
Everything else on this site descends from these pages. A finding on the blog names the study it came from; a Decision Brief applies one of these instruments to a named company's own decision. The research is free to read and to cite, with no form.
Every study here is re-run, and the next reading will differ from the last.
Sound Decisions carries each re-run and each correction, with the registered question written before the data is pulled.
Subscribe to Sound Decisions