Posts
Weight Loss Clinic Marketing and the GLP-1 Ad Rules for 2026
What Meta, Google, FDA and the FTC actually allow in GLP-1 ads, the words that get weight loss clinics in trouble, and the content rules we use with a clinic.
Facebook Ads to Amazon: Attribution and the Brand Referral Bonus
How to run Facebook ads to an Amazon listing: why Meta can't see the sale, how Amazon Attribution tracks it, and what the Brand Referral Bonus credits back.
Meta CompleteRegistration Event: Which Signup Step to Track
Meta's CompleteRegistration event counts finished signups. Fire it on the wrong step and campaigns starve. Which step to track, and how to set it up.
How to Track ChatGPT Traffic in GA4 and Tell When It Sent a Lead
A ChatGPT visitor asked us for an audit, and a day later GA4 still had the visit in the wrong channel. How to track AI traffic in GA4 and prove a lead's source.
Social Media Handle Check: Is Your Brand Name Taken?
Is your brand's handle taken? The check we run on Instagram, TikTok and YouTube before launch, what each platform says, and how to set up Meta ownership.
Skool Levels and Points: Why Locking Courses Backfires
How Skool levels and points work, why locking courses behind levels backfires in a young community, and the four settings that fix it, from a real member audit.
Your Instagram Caption Link Goes Nowhere
Links pasted into most Instagram captions aren't clickable, and the comment a keyword, follow first, get the link flow runs on a paid tier. Here's where links actually work and how to start an email list without guessing.
Bakersfield's Two Top-Rated Roofers Have 467 Reviews. AI Named Neither.
When a customer asks ChatGPT or Perplexity for a local recommendation they get one to six names, not ten blue links. We ran the test across three Bakersfield industries and found that ranking at the top of Google does not mean AI knows you exist.
The Ads Filled the Cart. Nobody Paid.
A jewelry brand's Meta ads delivered 972 landed visits, 7 add to carts and 8 checkout starts in twelve days, and nobody bought. Here's how two checkpoints we wrote in advance pointed us past the ads to the offer and the checkout, and what we found there.
Google Ads Reported 444 Conversions. None of Them Were Signups.
A software client's Google Ads program closed at $470.83, 2,726 clicks and 444 conversions. Every conversion was a YouTube channel subscription. Here's how the conversion column decides what a campaign chases, and how to report a real result that isn't the result you needed.
Ten Creator Slots Produced One Post
Matching with creators is the easy part. We audited a live campaign and found the failures are all downstream: shipping, approval, payment, disclosure, and a status label that means something other than what it says.
The Dashboard Said One Lead. There Was No Lead.
A counter read one. The record surface read zero. The download produced no file, three times, on three different days. Here is what we have learned about marketing numbers that restate themselves, and the six rules we now report by.