Google has updated its search help documentation around canonicalization this morning. The Google Search Relations team split in three distinct sections and updated a lot of the content to provide clearer details around how Google Search and canonicalization works.

The three sections include:

All of this use to be on a single help page, which you can review on the Wayback Machine over here to compare.

With this, Gary Illyes from Google dropped another LinkedIn tip on the topic of canonicalization, he wrote:

Friday ramble: you can stack canonicalization signals to strengthen that hint.

You have a rel=canonical pointing from A to B, but A is HTTPS, it’s in your hreflang clusters, all your links are pointing to A, and A is included in your sitemaps instead of B. Which one should search engines pick as canonical, A or B?

If you just change the URLs from A to B in your sitemaps and hreflang clusters, combined with that rel=canonical it might already be enough to tip over canonicalization to B. Change the links also, and you have an even greater chance to convince search engines about your canonical preference.

Recently, Gary also mentioned to use absoluate URLs for rel-canonical.

So check out these new docs and learn a bit more on canonicalization and Google Search.

Forum discussion at LinkedIn.

As a reminder, since April, Bing has been slowly testing a new BingBot user agent, slowly rolling it out to more percentages of crawls over the year. It should have rolled out to 100% of all crawls last month. But now Fabrice Canel said this week that it is near 100%, and an announcement is coming sometime today from Microsoft.

Fabrice Canel from Microsoft Bing wrote on Twitter, “We are near 100% and are proactively monitoring and rolling back any website having issues. Stay tuned for more communication Friday.”

Initially, Bing said it would be rolled out by Fall 2022, then January 2023. Maybe today is the day?

Here has been the rollout so far:

• April 2022: Less than 5% of crawls
• July 2022: 5% of all crawls
• September 2022: 20% of all crawls
• October 2022: 50% of all crawls
• February 2023: New 100% of all crawls

Let’s see what Fabrice announces today.

Forum discussion at Twitter.

Update: Confirmed, it is fully live according to Fabrice on LinkedIn.

Yes @rustybrick 💯% today (we fallback to previous user agents where appropriate). More information on Bing Crawler User Agents: https://t.co/n0ILIEUz0I https://t.co/49ILtJDuL0

— Fabrice Canel (@facan) February 3, 2023

Last night I reported on Search Engine Land, based on the Google earnings report and earnings call, that Google’s CEO Sundar Pichai has confirmed the search company will release a chat based search feature based on its own AI, LaMDA, in the coming weeks and months. Plus, there is a big new search event this Wednesday – so maybe we will hear about it there?

Yes, the other day we reported about Apprentice Bard, the reports of Google’s ChatGPT answer to OpenAI and Microsoft Bing.

Here is what Sundar said on the call:

Sundar Pichai said, “In the coming weeks and months, we’ll make these language models available, starting with LaMDA, so that people can engage directly with them. This will help us continue to get feedback, test, and safely improve them. These models are particularly amazing for composing, constructing, and summarizing. They will become even more useful for people as they provide up-to-date more factual information.”

Sundar Pichai said that he “first spoke about Google being an AI-first company” more than “six years ago.” “We have been preparing for this moment since early last year, and you’re going to see a lot from us in the coming few months across three big areas of opportunity; first, large models. We published extensively about LaMDA and PoN, the industry’s largest, most sophisticated model plus extensive work at DeepMind,” he continued to say.

During the question and answer period, Sundar added “We’ll be launching — we’ll — more as labs products in certain cases, beta features in certain cases and just slowly scaling up from there. Obviously, we need to make sure we’re iterating in public, these models will keep getting better, so the field is fast changing. The serving costs will need to be improved.”

“So I view it as very, very early days, but we are committed to putting our experiences, both in terms of new products and experiences, actually bringing direct LLM experiences in Search, making APIs available for developers and enterprises and learn from there and iterate like we’ve always done. So I’m looking forward to it,” he added.

Sundar goes on to add, “In terms of Search too, now that we can integrate more direct LLM type experiences in Search, I think it will help us expand and serve new types of use cases, generative use cases. And so I think I see this as a chance to rethink and re-imagine and drive Search to solve more use cases for our users as well. So again, early days, you will see us be bold, put things out, get feedback and iterate and make things better.”

Here is the event live stream on Wednesday:

Everyone is so ChatGPT crazed!

Forum discussion at Twitter.