Share
Embed
Slide 1
Slide 1 text
@frankvndijk
Slide 2
Slide 2 text
@frankvndijk
Friday 11:10am, where two worlds collide
Slide 3
Slide 3 text
Perfect, because what I'm about
to tell you is technical...
Slide 4
Slide 4 text
@frankvndijk
Slide 5
Slide 5 text
@frankvndijk
What are embeddings?
“Embeddings are numerical representations of
data (like words, images, or audio) in a multi-
dimensional space”
Images Audio
Text
Embedding model
0.9 0.7 0.2 0.6
Slide 6
Slide 6 text
@frankvndijk
cat
Embedding model
0.9 0.7 -0.3 0.6
Slide 7
Slide 7 text
@frankvndijk
cat dog
0.9 0.7 -0.3 0.6 0.9 0.6 -0.2 0.8
Slide 8
Slide 8 text
@frankvndijk
cat
dog
lion
Slide 9
Slide 9 text
@frankvndijk
It’s not that difficult!
Slide 10
Slide 10 text
@frankvndijk
Text-embedding-3-large
By OpenAI, highest performance
but more expensive
Text-embedding-3-small
By OpenAI, excellent performance
and lower cost
Gemini-embedding-001
By Google, flexible in use with
dimensions
Comparison of different models
Slide 11
Slide 11 text
@frankvndijk
Introducing EmbeddingGemma
Open source
Privacy
Small but very
powerful
Slide 12
Slide 12 text
@frankvndijk
Big websites ask for creativity
Comparing 1k+ urls in a
content gap… No way 🫠
Slide 13
Slide 13 text
@frankvndijk
I mean... It shows you
interesting keywords, but
not the whole picture
Content gap 🫤
Slide 14
Slide 14 text
@frankvndijk
Try Plotly for your visualisations
Free Python package
Interactive data showcase
Web app integrated
Slide 15
Slide 15 text
@frankvndijk
Slide 16
Slide 16 text
@frankvndijk
Organic traffic data
Vector embeddings
Competitor & own urls
Slide 17
Slide 17 text
@frankvndijk
API Access >> Ahrefs +
Gemini
API connection
Slide 18
Slide 18 text
@frankvndijk
Ahrefs >> Backlinks,
RefDomains, URL Rating off
and Traffic on
Gemini >> Extract
embeddings from page
content
Right settings
Slide 19
Slide 19 text
@frankvndijk
Extraction >> Store rendered
HTML
Rendering >> JavaScript
Rendering on
Slide 20
Slide 20 text
@frankvndijk
Export to graph with Python
Slide 21
Slide 21 text
@frankvndijk
Slide 22
Slide 22 text
@frankvndijk
Perfect showcase of the right data
Show your strengths
Show optimization options
Makes it visual for non SEO
Slide 23
Slide 23 text
@frankvndijk
Slide 24
Slide 24 text
@frankvndijk
Enable embedding
functionality, semantic
similarity and low relevance
Connect to your embedding
template
With embeddings
Slide 25
Slide 25 text
@frankvndijk
0 1
Identical
No similarity
Cosine similarity
Slide 26
Slide 26 text
@frankvndijk
Try a threshold of 0.8
Crawl to find new internal links
Slide 27
Slide 27 text
@frankvndijk
Similar pages will show up
Source url Target url
Slide 28
Slide 28 text
@frankvndijk
Workflow to find non existing links
Inlinks
Semantically
similar
Slide 29
Slide 29 text
@frankvndijk
Try it yourself, there is a
QR code at the end
Slide 30
Slide 30 text
@frankvndijk
Lorem ipsum dolor sit amet, consectetur adipiscing elit,
sed do eiusmod tempor incididunt ut labore et dolore
magna aliqua.
Audit anchors with cosine similarity
0.9 0.7 -0.3 0.6
0.9 0.6 -0.2 0.8
Slide 31
Slide 31 text
@frankvndijk
Finding irrelevant anchors
Inlinks
Embeddings
Slide 32
Slide 32 text
@frankvndijk
Vector embeddings are so interesting
because… Google has been using them for
years
Slide 33
Slide 33 text
@frankvndijk
Retrieval Augmented Generation
Identify Retrieve information Generate
AI
Slide 34
Slide 34 text
@frankvndijk
We need to check what content is
used to generate the AI overview
+more sources
Slide 35
Slide 35 text
@frankvndijk
How good is their content based
on a semantic level
+more sources
0.81
0.77
0.83
Slide 36
Slide 36 text
@frankvndijk
Should we optimize for a
cosine similarity of 1.0?
Slide 37
Slide 37 text
@frankvndijk
Don’t forget your SEO
It’s the foundation of a successful
GEO strategy. Without a strong SEO
base, your GEO strategy will fail
Slide 38
Slide 38 text
@frankvndijk
Success is no longer just about
matching a query, it’s about
deeply understanding the intent
and the questions behind it, and
answering them
Slide 39
Slide 39 text
@frankvndijk
PHASE 1 PHASE 2 PHASE 3 PHASE 4 PHASE 5
Find relevant
queries
Extracting
content
Generate
embeddings
Calculating
(cosine) similarity
Optimizing
content
Slide 40
Slide 40 text
@frankvndijk
Start by finding the full search intent
Get insights from:
Slide 41
Slide 41 text
@frankvndijk
We can take a look behind the
scenes at the competition with Ahrefs
Slide 42
Slide 42 text
@frankvndijk
Use regex to filter out long tail keywords
that might be relevant to your query
Slide 43
Slide 43 text
@frankvndijk
Or check what ChatGPT is using to find
information to generate an answer
Great tip from Mark Williams-Cook
Slide 44
Slide 44 text
@frankvndijk
All these insights give us a better insight
of what our target audience is looking for
SERP features like people also
ask, people also search for
Insights from Gemini
Trends from Google
Trends
Slide 45
Slide 45 text
@frankvndijk
Extract the right content to compare
Get the content by using:
Slide 46
Slide 46 text
@frankvndijk
We need to scrape the content that
is used in the AI overviews
Slide 47
Slide 47 text
@frankvndijk
Chunk your own content so we can
use it check if it’s good (enough)
Slide 48
Slide 48 text
@frankvndijk
Start my finding the full search intent
0.9 0.7 -0.3 0.6
0.9 0.6 -0.2 0.8
Slide 49
Slide 49 text
@frankvndijk
Sorry…
It was an workflow all
along… Sorry not sorry…
Slide 50
Slide 50 text
@frankvndijk
Please don’t start keyword
stuffing now…
Slide 51
Slide 51 text
@frankvndijk
Start adding value instead
Slide 52
Slide 52 text
@frankvndijk
Script that checks relevancy of
backlinks based on Ahrefs export
DR + relevancy
Slide 53
Slide 53 text
@frankvndijk
What’s next Start calculating cosine similarity
Switch from search query to intent
Start automating with the Google Sheets
01.
02.
03.
Slide 54
Slide 54 text
@frankvndijk
Your [SEO/GEO/AEO] strategy should
still draw the map and set the
destination, let vector embeddings be
the compass that guides you
throughout the journey
Slide 55
Slide 55 text
55
DEPT®/SEO Brein | @frankvndijk
Thank you for listening :)