Hamsterdam Part 62: Weekly SEO & AI News Recap (6/10 to 6/16, 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 week of Hamsterdam, and happy Father’s Day to the dads out there! Keep on keepin’ on. 😉
• I used to knock out these recaps by noon on Sundays, but they’ll likely be evening posts (US Eastern time) for a while. (Getting busy!)
• I appreciate you being here, and look forward to sharing the week’s news!
• In related Hamsterdam news, I wrote about a 2008 interview Vanessa Fox gave to Eric Enge about holistic SEO and marketing. Timeless lessons in that.
• Also, if you’re a fan of Google Search Central’s content, I made a summary of all Search Off the Record podcast episode transcripts using Gemini. It’s a fun article.

Feel free to jump ahead to the news recap if you’re in a hurry. Or stick around for vocabulary, this week in SEO history, and an introduction first!

If you’d like Hamsterdam delivered every week, subscribe to the free newsletter.

And we’re off …

---

Marketing word of the week: “Ecommerce”

First off, does anybody know how to spell this word? Ecommerce, eCommerce, e-Commerce, E-commerce, well you get the point.

“Ecommerce” with no hyphen has the higher U.S. search volume per Ahrefs, likely because it’s simpler to type.

As for the definition:

Ecommerce is an abbreviation of “electronic commerce.” It refers to the buying and selling of goods (or services) over the internet.

This can be online retail shopping, auctions, and even digital downloads or B2B sales (recall Part 59’s vocab). Some brands are online only, while others have brick-and-mortar retail locations.

Popular ecommerce website platforms are Shopify, WooCommerce (WordPress), and NopCommerce, but most CMSs offer a store functionality.

Ecommerce SEO can focus on site structure, internal linking, and page experience, and increasingly, Google uses its shopping graph to show products in the SERPs. As we’ll see below, Merchant Center can also use a website’s product structured data for free listings.

Another key aspect of ecommerce marketing is working with other channels, like email, social media and UGC, and CRO.

Some of the most fun projects I’ve been apart of were for ecommerce brands. Of course, it can be a struggle bus ride if all teams aren’t rowing in one direction. 😉

---

AI word of the week: “Embedding spaces”

I made this section a bit longer to account for how important the concept is. If it’s totally new to you, check out some of the linked resources, as well.

Embedding spaces are multi-dimensional mathematical representations where words, phrases, or entire documents (and even websites, potentially) are mapped as points or vectors.

The position of these points (vectors) relative to each other (based on mathematical ways of measuring, like Euclidean distance, cosine similarity, dot product, etc.) reflects their semantic similarity (relationships based on deeper meaning and context).

The close proximity of vectors for a query and a document in an embedding space, for example, may suggest that document’s relevance (search intent alignment) for information retrieval (search engines).

Embedding spaces are fundamental to natural language processing (NLP), which involves enabling machines to understand and process human language. NLP is also foundational to today’s search engines like Google (semantic meaning) and LLMs.

By representing words as vectors (numbers or coordinates to signify their position in a continuous and high-dimensional embedding space), NLP models can capture nuanced relationships of meaning.

Here’s an example of different embeddings of the word “mole”:

Popular word embedding models for NLP tasks include Word2Vec and GloVe. However, the attention in transformers has enabled deep neural networks to encode more dense embeddings with richer context from surrounding words (think Google’s BERT or T5). This shift to transformer-based models has improved the quality of embeddings.

Search engines can leverage embedding spaces to better understand the search intent of queries and the relevance of documents or web pages (and likely much more, like authors or domains). This is the evolution to semantic search, beyond lexical (keyword-based) matches.

“So what’s the difference between traditional keyword-based search and vector similarity search? For many years, relational databases and full-text search engines have been the foundation of information retrieval in modern IT systems. For example, you would add tags or category keywords such as “movie”, “music”, or “actor” to each piece of content (image or text) or each entity (a product, user, IoT device, or anything really). You’d then add those records to a database, so you could perform searches with those tags or keywords.

In contrast, vector search uses vectors (where each vector is a list of numbers) for representing and searching content. The combination of the numbers defines similarity to specific topics. For example, if an image (or any content) includes 10% of “movie”, 2% of “music”, and 30% of “actor”-related content, then you could define a vector [0.1, 0.02, 0.3] to represent it. (Note: this is an overly simplified explanation of the concept; the actual vectors have much more complex vector spaces). You can find similar content by comparing the distances and similarities between vectors. This is how Google services find valuable content for a wide variety of users worldwide in milliseconds.”

– Find anything blazingly fast with Google’s vector search technology, Kaz Sato & Tomoyuki Chikanaga (Google Cloud)

LLMs also use embedding spaces to generate coherent and contextually relevant responses based on user queries (prompts), or perhaps even personalize responses from user interaction data.

Here’s an example of a 2D embedding space:

But keep in mind, embedding spaces can have hundreds or even thousands of dimensions, each representing a different aspect of meaning.

For context, that image above is from 2016, pre-transformer architectures.

The high dimensionality of embedding spaces today allows them to capture complex semantic relationships.

In the realm of search, this can be resource intensive, though, involving tradeoffs. (Think multiple phases of ranking results.)

An example work flow for embedding spaces could be:

1. Raw text data (unstructured, typically) is encoded (one-hot encoding).
2. That input is then compressed into an embedding layer, either lower-dimensional (like Word2Vec or GloVe) or higher-dimensional (like BERT or MUM), as embedding vectors (learned through training).
3. These embedding vectors then exist in an embedding space, which captures their semantic similarity.
4. Search engines or LLMs then use that embedding space for NLP tasks.

Understanding the embedding space can be difficult (a challenge of “interpretability”). The GIF above shows the Embedding Projector from Google (2016). More recent techniques include ELM (embedding language model), where LLMs are used to transform abstract vectors into understandable narratives.

---

This upcoming week in history: “Juneteenth”

This section usually covers “this week in SEO history,” but part of why I created Hamsterdam was to use the platform to discuss other topics of importance, as well.

That’s why I’d like to explain the meaning of the upcoming Juneteenth holiday on June 19th (this Wednesday).

Juneteenth National Independence Day is a federal holiday in the U.S. commemorating the ending of slavery.

On June 19th, 1865, the Union Army, under the charge of General Gordon Granger, made its way to Galveston, Texas, where the general announced that all enslaved African Americans were free.

The Emancipation Proclamation was already issued by President Abraham Lincoln on January 1st, 1863. This proclaimed “all persons held as slaves” in the rebelling states “henceforward shall be free.”

However, the U.S. Civil War continued for more than two years after until April of 1865.

It then took several months for freedom to reach the most western rebelling states.

It wasn’t until the ratification of the 13th Amendment on December 6th, 1865, though, that slavery was abolished throughout the U.S.

Here’s a perspective on Juneteenth from Angela Tate, Curator of African American Women’s History at NMAAHC (National Museum of African American History and Culture):

“When I think about Juneteenth, I think less about it being a specifically American, but how it connects African American thoughts about freedom and emancipation to the same notions across the African diaspora. There is this impulse towards commemorating, celebrating, and remembering freedom. African Americans have always used these moments of memory to think about where the community has come from and what we’re pursuing and striving towards, as well as taking the time to pass down history and culture.

Juneteenth is a time to reflect. What does it mean to really celebrate our freedom? What does it mean to be free in moments where freedom is conditional, and freedom is always a challenge? Juneteenth is a moment to think about freedom being conditional freedom and it is something that we must continuously strive and fight for.”

– Angela Tate (NMAAHC)

You can learn more about Juneteenth from NMAAHC.

In my local area of Orlando, we also have the Wells’Built Museum of African American History and Culture. (I wrote about my visit there in a past recap).

>> You can donate to NMAAHC here.

>> You can donate to the Wells’Built Museum here.

---

Introduction to week 62: “Stick with it”

Yesterday, I wrote a blog post about SEO lessons from 75 Google Search Central podcast episodes, based on responses Gemini gave me from inputting the PDF transcripts into prompts.

That article had three sections, essentially.

One where I only used transcripts.

Another where I used transcripts and content from Bing’s blog.

Then a third section where I used transcripts and SEO-related blog posts from the last 3 weeks of Hamsterdam recaps.

That third section proved the most challenging to edit because the combining of multiple perspectives made it harder to decipher the context behind Gemini’s responses. (You can see more of what I mean in the article.)

However, Gemini did have a takeaway that I found spot on:

“The overall takeaway is that SEO is more than ever a long-term game that requires a holistic approach.” — Gemini

The idea of a “holistic approach” is interesting to think about.

In some ways, it’s a challenging concept to grasp if you’re not in a position to see how different aspects of offline and online marketing, business practices, and socio-economic trends interplay overall.

For someone working on individual elements in audits or editing single pieces of content, for example, it can be difficult to see how that one effort contributes to a holistic approach.

It can also be hard to think about such efforts from a different context.

Roger Montti had an interesting SEJ article this week about replies to comments Danny Sullivan made as Google Search Liaison on X (Twitter) regarding diversifying traffic sources:

Google On Traffic Diversity As A Ranking Factor. via @martinibuster: https://t.co/KkFvYV4S4w#seonews #Google #SEO

— SearchEngineJournal® (@sejournal) June 14, 2024

As another example, I have a blog post that talks about instances of social media content appearing in Google’s search results.

The argument is to treat eligible social content like it’s “SEO content,” because it can appear in the same user journeys in Search.

However, if you search on Google for [social media content ranking in Google Search], the dominant search intent is often “Will social media help my rankings?”:

I think Vanessa Fox and Eric Enge did a nice job discussing holistic marketing in an SEO context during a 2008 interview that I recently learned about.

Here’s an excerpt from their conversation:

In my mind, the simplest way to think about a holistic approach comes from a sentiment espoused by none other than Lester Freamon in The Wire: “…and all the pieces matter.”

But why did I call this introduction “Stick with it”?

Well, that can be part of it, too.

I started putting effort into my website’s blog in mid-2023, but to say I felt some skepticism is an understatement.

Yet, I was confident that if I kept writing about what interested me, I’d grow myself personally and eventually figure out a way to convey my thoughts better to readers.

More and more, people are reaching out because they’ve read something here.

Sometimes they found it in Search, but other times, it’s from social media or other referral sources. (I started a newsletter a few months back, for instance, because someone who reads Hamsterdam requested it.)

I’ll always continue blogging, whether page views come from my wife, peers, or potential clients, simply because it’s fun and helpful.

Yet to Gemini’s point earlier (or more accurately, Gemini’s summarization of the views of several wise-minded people), “SEO is more than ever a long-term game that requires a holistic approach.”

All the pieces matter, even when it feels like they don’t.

Stick with it. 😉

Buckle up for a full week’s recap, and enjoy the vibes (new (old?) music from Puscifer):

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

Missed last week? Don’t worry, I got you! Read Part 61 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

---