|
SEO updates you need to know
Sponsor: SEO IRL

October 6 & 7, 2026 in Toronto, Canada
Canada’s largest SEO conference returns to MaRS Discovery District for two jam-packed days.
Join Lily Ray, Nathan Gotch, Joy Hawkins, Michal Suski and more for cutting-edge SEO and AEO playbooks with real-world results. Lunch and swag provided both days, plus the highest concentration of SEOs in Toronto.
SEO IRL sold out last year. Get your ticket before it sells out again.
|
Search with Candour podcast

SearchNorwichXL preview
Season 4: Episode 86
Jack Chambers-Ward continues the hype for SearchNorwichXL by highlighting talks from Arnout Hellemans and Brendan Bennett, who will be speaking at SearchNorwichXL.
Arnout’s talk from SearchNorwich 16 is titled ‘How to work on your SEO without paying for tools’, and Brendan’s talk from SearchNorwich 19 is titled ‘It’s not me, it’s you: Why your SEO or agency (probably) hates you’.
|
|
|
Last week's SEO tips

For whom the spam tolls
I agree with Lily Ray that ChatGPT seems to be using lots of site: queries as a quasi-spam filter (OpenAI also bought a ton of link graph data from vendors). Peec AI has shown a massive drop in the exact type of pages we have been highlighting in SEO audits, where companies have decided to use 'pSEO' to scale them and Google has also not been a fan. Things like:
- reviews pages
- 'best of' pages
- 'top' pages
- comparison pages
- 'vs' pages
I was going to write something about this, but Lily Ray beat me to it, and I don't think I can really add much to what she has written.
|

Ride the grounding
I cannot express how important it is that you understand if web searches are happening for your topic in AI platforms like ChatGPT. Great example from Ina Ilieva today ⤵️ Ina was using our free QueryFan tool to explore query fan-outs that were happening when users were asking ChatGPT for recommendations for things like self-help books. I was contacted because Ina suspected QueryFan was broken as it was not listing any fan-outs for any of these queries, which seemed really surprising to me! When I loaded up QueryFan and tested it on my regular "running shoes" topic search, I could see that it was indeed working correctly and returning query fan-outs. I decided to verify locally using Suganthan's "Fanout Fox" tool, and confirmed; all of the "self help book" recommendations are 100% coming from training data. No web searches are happening. I found this surprising. It means there is nothing you can do to change your visibility in the short term. No mentions elsewhere, no rewrites, chunking, prayers or cats.txt are going to help you. Sometimes knowing what _not_ to do is important 💡
|
|

My friend of Delphi
Micro/small businesses, this is for you! AMA (sort of). I get a lot of questions via DM that I have previously answered in my SEO tips, or in webinars, podcasts, or writing, so I am doing an experiment ⤵️ I've trained Delphi on hundreds of my LinkedIn posts, my blogging, Substack writing, videos of talks I've done, podcast audio, webinars, and you can talk to it/(pseudo-me), and it seems to give fairly decent answers to a lot of the more regular questions I get. I'm acutely aware that a lot of professional SEOs / agencies like Candour are not particularly accessible to micro/small businesses due to cost, and they are particularly vulnerable to grifters looking to make a buck, which is one of the reasons I share so much information publicly; to help them make better decisions. No, it's not the same as hiring a professional consultant, but I do believe it will be genuinely helpful to a lot of people starting out, or that want to double-check information they are being given by someone trying to sell them something!
Chat with Robo Mark here.
|

Disposable domains
While there is news about Google 'not enforcing' Site Reputation Abuse penalties in the EEA, it is worth talking about the concept of a site vs domain. 💡 I commonly see SEOs conflate the terms "domain" and "site", and they are very different concepts. I believe this has been the centre of confusion when Google has made various historic claims about not having some "domain wide" scores, when SEOs were really asking about site wide scores, which Google absolutely has. This was made abundantly clear to me when we found the "site_quality_score" metrics in the Google exploit, and we could clearly see that Google was giving individual (and very different) scores of a www. version of a site vs the blog. and forum. subdomains for instance. They were treated as their own, separate units, in terms of re-ranking.
|
Refer subscribers and earn rewards!
|
|
|