Hey Reader!
I’ve been doing more work with helping clients talk to their legal counsel on what their default consent states should be, and as a part of this, I review their privacy policies against what actually happens on their site. Most of the time, the gap between what the policy says and what’s happening is larger than anyone would like.
For example, one policy I recently reviewed said plainly that it doesn't collect any information about you using cookies but then in the very next paragraph, it described exactly what their cookies collect and where it goes. Another policy leaned heavily on "using our site means you agree to this" language, even though several of the regions that client serves require an actual opt-in click, not silent agreement.
This happens because the privacy policy is written by legal and the tracking setup is built by a completely different team. It’s very rare that anyone is assigned to check that the two still agree, which is exactly why they usually don't. If you’re agency-side, this is a good audit to offer clients whether or not they've asked for it, before a regulator or a customer finds the gap between what the policy says and what actually happens.
In-house folks, if you haven’t done this lately, put it in your to do list. You might not find anything catastrophic, but you’ll probably find something.
AI Prioritization Framework for SEOs
If your LinkedIn feed has you convinced you're hopelessly behind on AI, Liz wrote this piece just for you. It's a prioritization framework for learning AI for SEO without burning out, built around a self-audit of your niche, your clients, and where you actually feel friction in your week. This way you can learn things that actually make you feel like you’re making progress instead of flailing around in a sea of way too much information.
Articles Worth Your Time
———•
|
Search Console can finally see your social and video content
Google added a new Search Console property type called platform properties, and it’s a big thing for Google: for the first time you can verify a property you don't own the domain for. Connect an Instagram, TikTok, X, or YouTube account and you'll see the clicks, impressions, and actual search queries pulling people to that content in Search and Discover. This provides fantastic insights for a channel that we used to have to track manually (if it was tracked at all), and the query data is the real gold here. I’m excited to dig in, but this is in a staggered rollout so you might not be seeing it yet.
How ChatGPT actually picks its sources
Most "how to show up in ChatGPT" advice is a bunch of guesses, or SEO in a new set of clothes. Suganthan Mohanadasan did something better: he read the raw network traffic that’s sent when you do a search. He found that every web result carries a hidden label sorting it into one of four pipelines, and the licensed-publisher tier (Reuters, WSJ, Wikipedia) is effectively closed unless you own a national newspaper, so the rest of us compete in the scraped tier, which means the whole game is being cleanly scrapable. One big finding is that if your pricing or specs load through JavaScript, ChatGPT often can't read them and will cite a third party instead.
More ways to keep your BigQuery bill from creeping up
This is a paid post, so you'll need a subscription for the full queries (but it’s worth it, I’m a subscriber!). I think it’s a valuable piece because what people don’t realize is that BigQuery is cheap until some big data shows up (e.g. heavy GA4 traffic, Merchant Center benchmarking) and then it gets expensive, fast. If you only implement one thing from this piece, follow their tip on capping your query spend in BigQuery Studio Settings, so you won't cause a spike or get blamed for someone else's. The rest of the piece covers finding your costliest query patterns and who's running them, using metadata to trim cost, avoiding querying the same thing twice, being deliberate about data types, and copying tables without paying twice for storage. Take yourself out for a nice dinner (or buy a restaurant) with the money you’ll save! Google doesn’t need it.
Where You Can Find Me
———•
|
BrightonSEO Workshop: Brand New!
Have you heard of this AI thing? It’s very popular right now. I decided to reorient my reporting workshop around AI — renaming it to Question-Driven Reporting with Data Studio and AI. I’ll be covering how to build reports that answer the actual question someone asked, faster and with far less manual work. Plus you’ll get some fun Claude skills to take home! The training day is the Monday before the conference (September 15 to 16), and your workshop ticket gets you into both days of talks too.
Use this link to get a 10% discount. See you there?
That's it for this edition of The Huddle. As always, if you have questions or want to share what you're working on, just hit reply!