|
SEO updates you need to know
💸 |
|
Google confirms ads coming to AI Mode. A leaked document explains that ads will be based on the wider context of the conversation, not just the current query. So far, ad formats eligible for AI Mode include Text & Shopping ads. |
🔁 |
|
AI Overviews start appearing in AI Overviews. We've seen AIOs in People Also Ask but this is the first time we've seen AIOs within AIOs. This could be a bug or maybe Google heard you like AIOs so they put AIOs in your AIOs... |
Sponsor: fatjoe
The hottest trend in link building right now is digital PR campaigns.
Our State of Link Building survey showed it, and you know it.
But how can you offer quality PR campaigns for your clients at scale?
Easy - with fatjoe’s done-for-you digital PR campaigns.
- Clear coverage guarantees.
- Transparent pricing.
- High-impact in-content journalist links.
|
Search with Candour podcast
The future of schema & structured data
Season 4: Episode 31
Join host Jack Chambers Ward as he sits down with returning guest Alex Wright, a managing partner for SEO at Dentsu, to uncover the untapped potential of organic shopping feeds in the ecommerce landscape
Learn about the importance of optimising your feeds, the pitfalls to avoid, and the bright future of agentic search in 2025.
|
|
|
This week's solicited tips:
All we have to decide is what to do with the IP that is given to us
If you're setting up automated crawling of your site as Googlebot, it is beneficial to make the origin of these crawls the US.
Googlebot is normally crawling from US-located IP addresses, so also crawling from this location can catch edge cases, such as automatic geo-IP redirects (which you hopefully don't have).
|
It's a dangerous business, relying on LLMs
Who is liable when Perplexity says your company has "communication issues", but they are incorrectly pulling reviews from a similarly named company? 🤔
This is a whole new area of SEO opening up: Making sure that information being fetched about you is correct. In the not too distant future, people are going to be outsourcing their decision making to these agents. If this were the case, how much business would Candour be losing because LLMs are spreading processing information incorrectly?
|
Even the smallest leak can change the course of the future
Leaky data alert! 🚨Understanding how the LLM layer is processing and filtering the information it gets is an increasingly important part of SEO. Now, we know a bunch about how Perplexity works ⚙️
This is down to some ✨fantastic✨ work by Metehan Yeşilyurt, definitely worth following. Metehan has been dropping some of the best SEO stuff I've seen recently.
Metehan discovered 59 ranking patterns through browser interactions with Perplexity, and has been kind enough to share his discoveries with the SEO community.
There is some pretty basic (and brittle) stuff here, like manually specified domains and various threshold settings, but all very valuable to learn!
Here's Metehan's blog post.
|
This task was appointed to Comet and if it doesn't find a way, no one will.
I have no doubt that users will be outsourcing a chunk of decision making and research work to agents in the future. You can test how your site performs with agentic browsers today. ⤵️
I've been testing Perplexity's Comet browser to see how various sites perform.
The video shows the Comet assistant looking at the Candour website to try and determine if we offer digital PR, if we look any good at it - and if so, to send us an enquiry.
✅ The task was performed perfectly, and much to my surprise, the agent managed to complete Cloudflare's "Check this box if human" test!
I've also been testing some e-commerce websites, asking if they stock specific products and if so, to buy the cheapest variant it can find.
✅ The task was performed successfully, with the assistant finding the right colour, type, and quantity of paint I wanted, and adding it to my basket, leaving at this stage, just for me to do the checkout [I'm not trusting it further at this stage!]
In terms of "building SEO friendly websites", I can see this becoming a standardised block of tests that we run, to ensure there is nothing on the site that is causing agents to become confused.
Very interesting space to see develop, as new standards emerge.
|
What can men do against such reckless indexation?
If you need to quickly remove search results like ChatGPT recently did with their shared conversations, you can use the Removals Tool in GSC ⚒️
However! It is worth noting the removals tool is temporarily hiding those URLs (or pattern of URLs) from search results, it is not “removing” them in the way a noindex does.
So why would you use this tool?
1) If you noindex pages, they need to be crawled again before it can be honoured. This can take hours, days, or weeks depending on your website.
2) Dev stuff doesn’t always happen instantly, does it? While you’re waiting for that noindex dev ticket to be prioritised, the Removals tool gives an instant fix.
While URLs are “removed” from search, they will still be crawled and such as normal, so if you did use noindex directives on URLs, they would start to be honoured during this period.
The removals tool will hide URLs for six months.
|
Refer subscribers and earn rewards!
|
|
|