Editorial policy

Who is responsible

Brandtree is published and edited by Alykhan Alidina, who is accountable for every published fact. Corrections, disputes and takedown requests go to the same person. Contact details are on the about page.

What we publish and what we do not

We publish ownership relationships, the dates they began and ended, the deals that caused them, the regulatory decisions that shaped them, and the documents that establish them. We publish counts and comparisons computed from that data.

We do not publish speculation about deals that have not been announced, valuations that no party disclosed, or claims about a company's conduct that we cannot source to a filing or a regulator. We do not reproduce news article text; where we link to a headline, we link to the publisher and store nothing but the title, the link, the date and a short original summary.

We do not use brand logos or wordmarks anywhere on the site. Brand names appear as text, referring to the brands as the subjects of factual reporting.

Corrections

Every entity page carries a correction link that opens a pre-filled issue naming the record and the file it lives in. We ask for the primary source that supports the correction, because a correction has to clear the same bar as the original claim.

When a correction changes a published fact, we change the data, note it in the public corrections log with what was wrong and what it is now, and update the verification date on the affected page only. We do not silently edit. When a correction reveals that a source was weaker than we treated it, we downgrade the confidence label rather than deleting the record.

Ownership genuinely changes. A page that says something different from what it said last year is usually not an error: check the timeline, which shows every owner with dates, and the changelog, which records what changed and when.

Where language models are used

This matters enough to be specific about.

Entity answer sentences are not written by a model. The first sentence of every brand and company page is generated from the database by a deterministic template. It says what the data says, and if the data changes the sentence changes. No model is in that path.

Editorial explainers may be drafted with model assistance. Long-form articles about industry consolidation may start as a model-assisted draft. Every such draft is reviewed line by line against primary sources by a human before publication, every factual claim in a published explainer carries a citation, and nothing is published while still marked as a draft. The tooling enforces this: draft files are excluded from the build entirely.

Records proposed by automated ingestion are never published automatically. Scripts that scan regulatory filings and regulator websites write proposals into a review queue. A human approves, edits or rejects each one. Nothing in that queue is built into the site.

Independence

Brandtree takes no payment from any company it covers, accepts no sponsored placement, and runs no advertising. There are no affiliate links. If that changes, it will be disclosed here first and on the affected pages.

Privacy

The site sets no cookies and runs no third-party scripts beyond privacy-preserving analytics that store no personal data. There is no cookie banner because there is nothing to consent to. See the privacy page.