Writing
Field notes from the edge of search and measurement in pharma — how brands show up (or fail to) in AI answers, how to keep analytics honest, and the prompts and tools I actually use. Most pieces take about five minutes.
Start here
I built a no-code patient-insight tool in 20 minutes
No code, on-label, MLR-aware, and shareable — a specialist AI agent for pharma patient-question mining. I pointed it at sickle cell; here's what it found.
You can rank #1 on Google and still be invisible in AI search
Pharma brands that won traditional SEO can still be absent from AI-generated answers. Closing the gap is structural reassembly of pre-approved language — not new content.
The enigma of direct traffic
The one number in an analytics report that admits what it doesn't know — and, as the unattributable share grows, what happens to a number nobody owns.
All writing
What pharma can do with ChatGPT ads under OpenAI's policy: no Rx promotion, a narrow unbranded lane, no ads in personal health chats, and how to measure it.
No code, on-label, MLR-aware, and shareable — a specialist AI agent for pharma patient-question mining. I pointed it at sickle cell; here's what it found.
Fierce Pharma Week named the right problem: accuracy in AI answers. Naming isn't measuring — here are four questions that actually score it for a brand team.
Paste any AI or martech vendor's pitch in and get the questions that make the demo fall apart. It won't decide whether to buy — it just makes it harder to fake.
Anthropic's NPI connector is reference data for cleaning HCP inputs, not a targeting tool — and the NPI join is a privacy call, not analytics'.
The one number in an analytics report that admits what it doesn't know — and, as the unattributable share grows, what happens to a number nobody owns.
Snitcher's docs assume Google Tag Manager. Here's the Matomo Tag Manager build that works — the _mtm data layer, visit-scope dimensions, and the traps.
Pharma measurement problems aren't missing dashboards — they're duplicate events, names nobody agreed on, and PII in query strings. This prompt grades them.
An open-source workflow that turns a site spec into a complete analytics implementation — discovery, ARD, GTM container, QA plan, and GA4 checklist — from a single source file.
A Now page is a dated, machine-readable answer to 'what is this person working on right now?' — a question bios never answer and AI assistants get asked.
Answer engines read your site's structure, not your design. This prompt audits your pages the way a machine does and returns a prioritized fix list.
Patients and clinicians are using ChatGPT and Perplexity to research treatments. This is a prompt I use to check what those tools are actually saying.
A patient site's paid search reported a 76% conversion rate. The number was an artifact of how conversions were counted — and the bidding algorithms were optimizing against it.
Pharma brands that won traditional SEO can still be absent from AI-generated answers. Closing the gap is structural reassembly of pre-approved language — not new content.