🐹 Hamsterdam Part 77: Weekly SEO & AI News Recap (11/24 to 11/30, 2024)

By Ethan Lazuk

A weekly look-back at SEO & AI news, tips, and other content shared on social media & beyond.

Opening notes, thoughts, and musings:

• Welcome to another new week of Hamsterdam! 🐹
• I appreciate you being here, and look forward to sharing the week’s news! 🙌
• It’s Black Friday week, so thoughts go out to those for whom this is like the Super Bowl.
• The holiday lights are up in Lake Eola Park (downtown Orlando). We dropped by the other night:

If you’re in a rush, hop down to the news portion. 📰

Or continue reading for two vocabulary lessons, this month in SEO history, plus an introduction.

And we’re off …

---

📊 Marketing word of the week: “Metrics”

Metrics are quantifiable measures used to demonstrate the effectiveness of your marketing activities. They help you identify what’s working or not and plan for future areas of investment.

Metrics vary depending on the marketing channel and your goals within that channel. Examples of high-level metrics could be for website traffic, including overall traffic, unique visitors, or average session duration to measure engagement.

SEO metrics have traditionally focused on keyword rankings, organic traffic, or even backlinks obtained, but the concept of applicable metrics is being rethought in the age of AI, where users can get their informational needs satisfied without clicking through to a website.

Shifts in SEO metrics include a heightened focus on business value, such as CTR and conversions, brand visibility in AI-generated responses, as well as SERP real estate visibility, like for knowledge panels.

In terms of where to measure marketing metrics, GA4 and Google Search Console are go-to platforms for SEOs. Bing Webmaster Tools has also taken on renewed importance considering that ChatGPT Search is partly powered by Bing’s results.

---

📚 AI word of the week: “Mixture of experts”

For “M” words this week, we could go with “machine learning,” which is a program or system that trains a model from input data to make predictions on new or unseen data. That said, let’s get a bit more in the weeds and discuss “mixture of experts” within ML.

Mixture of experts (MoE) in machine learning involves using a team of specialized expert networks instead of one large model, where each expert focuses on a specific part of the problem space.

An analogy would be like going to medical specialists for different areas of treatment as opposed to seeing a general practitioner.

Here’s a basic idea of how MoE looks:

The way MoE works is the problem space first gets divided into smaller sub-problems, each with its own assigned expert neural network that specializes in that area.

A gating network then analyzes the input data and decides which expert(s) is best suited to handle it. The outputs from the selected experts are then combined into the final result.

For example, say we asked an ML model a question about history. The gating network would identify the topic of the problem and then route the question to the history expert network.

The benefits of MoE include improved accuracy, increased efficiency, and scalability.

An example of an ML model that uses MoE is Google’s Gemini 1.5 Pro.

---

This month in SEO history: HTTP v0.9

In November of 1991, the very first version of the Hypertext Transfer Protocol was born, known as HTTP v0.9.

HTTP is a set of rules that allows a web browser (the client) to communicate with web servers hosting websites.

The original HTTP features included a one-line request, containing the GET method and the path to the requested resource. It only supported GET (no POST, PUT, or DELETE).

The first HTTP also didn’t use headers, so no information about the client, server, or requested content was exchanged beyond the basic request line.

Today’s HTTPS uses strong encryption algorithms to deter hackers. It also uses TLS 1.2 and TLS 1.3, moving beyond less secure SSL versions for improved security and performance.

HTTPS is a requirement for modern websites, as indicated by the warning browsers like Chrome will show for non-HTTPS sites.

We also learned a bit more about how Googlebot crawls this week:

Google picks the best protocol that provides the best crawling performance for Googlebot; HTTP 1 or HTTP 2 and sometimes switches between sessions https://t.co/wHw79cDbFH hat tip @gaganghotra_ pic.twitter.com/pxsz2cgDkP

— Barry Schwartz (@rustybrick) November 27, 2024

There’s also HTTP/3, as shown in the chart above. Unlike its immediate predecessors that rely on TCP, HTTP/3 uses a new transport protocol called QUIC, which reduces the time it takes to establish a connection between the browser and server, leading to quicker page loads.

---

⛑️ Introduction to week 77: Lost content recovery

I recently published a short blog post about how to reclaim lost content after a website migration using WayBack Machine.

This stems from an audit I recently did where a site’s YoY traffic was just a fraction of what it was previously:

While we’ve seen HUC-impacted sites lose significant traffic, this site didn’t align with that timeline.

Upon further investigation, I learned the website underwent a redesign last year in which a sizable amount of its pages were removed.

The kicker is the client wasn’t aware the pages were removed. 👀

I’ve encountered this situation a few times, where sites’ pages get removed to streamline the redesign process — likely a cost consideration — without an SEO’s input.

Upon looking at the removed pages, I realized there were some popular blog posts that had business value because they were on the buyer’s journey.

It’s situations like this where Wayback Machine is your friend.

Use Wayback Machine to recover the lost content, put it into a Google Doc for a fresh review to ensure the information is current and the format satisfying by today’s standards, then republish the content at the same URL or using a redirect.

While not ideal, if you come across this situation, there is value in reclaiming the lost content. Whether the original signals fully transfer is an open question, but as long as there’s a business case for the content to exist, there’s no downside.

Thank you for supporting Hamsterdam and the cause of SEO & AI learning. 🙏

Enjoy the vibes:

Missed last week? Don’t worry, I got you! Read Part 76 to catch up.

🌟 Other great sources of weekly SEO news:

• The SEO Weekly – Garret Sussman, iPullRank
• SEOFOMO – Aleyda Solis
• Weekly Video Recaps – Barry Schwartz, SER
• Weekly SEO News YouTube channel – Olga Zarr, Seosly
• Niche Surfer – Yoyao Hsueh

---

Time for our weekly review of SEO social posts, articles, & more …

Now, let’s step inside the white flags of Hamsterdam …

⏩ Jump to a section:

• News, Google updates, & SERP tests
• SEO tips & tidbits
• Fundamentals & resources
• Articles, videos & case studies
• Local SEO
• Technical SEO
• Content marketing
• Local SEO
• Data analysis & reporting
• AI, LLMS, & machine learning
• Miscellaneous & general posts
• Older stuff that’s good!

Or keep scrolling to see it all. ⏬

I’m an independent SEO consultant focusing on custom audits and holistic strategies for brands. Don’t hesitate to reach out, or visit my about page for more information.

Let’s connect!

Hit me up anytime via text or call at 813-557-9745 or on social or email:

• X
• LinkedIn
• YouTube
• TikTok
• Mastodon
• Reddit
• Instagram
• Threads
• Facebook
• Mail

Cheers! ✌️

---

Past recaps