Writing
Field notes on analytics, technical SEO and AI discovery for government agencies, universities and B2B teams. What breaks on large, complicated sites, why it breaks, and what we do about it.
Why collecting data just in case creates audit, breach and legal risk, and how to shrink what you store so analytics supports decisions without liability.
A case study of a complex B2B site that lost buyer trust through an off-domain form, outdated PDFs and slow pages, and how fixing them rebuilt it.
Why AI discovery is not SEO with new jargon: citations over clicks, blocking and PDF traps, schema, first-party data and new metrics.
Client questions on the June 15 Consent Mode v2 change answered: why a CMP is not enough, what shifts for Google Signals, and Basic versus Advanced mode.
A five-step audit for finding PDFs that AI cannot read, optimizing the ones you must keep, and moving key higher ed content to structured HTML.
Why agencies reporting on their own ad spend hide waste, and how an independent forensic audit finds ghost scripts and self-referral traffic.
Seven AI discovery strategy mistakes, from blocking AI crawlers and PDF-only content to weak schema, and how to become a source LLMs cite.
How to organize a large site around entities: map your core entities, add nested JSON-LD, and give each one a single canonical page.
How Consent Mode v2 works, the difference between Basic and Advanced modes, and where government and higher ed sites leak data when tags ignore consent.
Which metrics to add to SEO reports as AI answers replace rankings: citation frequency, primary-source rate and freshness.
Quick setup for the GA4 to BigQuery export, why to start before you need it, and how SQL views keep Looker Studio dashboards fast and free of quota errors.
Why enrollment marketing agencies that report on their own spend create a conflict of interest, and what higher ed leaders should ask for instead.
Tell us what you’re seeing. We’ll tell you plainly whether it’s something we can fix.
