Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Mining the SERP for SEO, Content & Customer Insights

September 20, 2019

Mining the SERP for SEO, Content & Customer Insights


September 20, 2019


  1. Mining the SERPs for SEO, Content & Customer Insights Rory

    Truesdale // Conductor http://cndr.co/brighton @RoryT11
  2. •SERPs are a great resource to learn what Google ‘thinks’

    our customers want •Workflows that will help you understand the intent of the people you want to reach •How to use these insights to improve the quality of your on-page optimisation What To Expect @RoryT1
  3. We can deconstruct & analyse the language in SERP displayed

    content to learn what Google thinks our customers are interested in @RoryT1
  4. Google isn’t ranking a page based on how it uses

    a keyword. Google provides accurate results based on intent, query context & word relationships. On-page Optimisation @RoryT1
  5. • User intent • Query context •Topical relevance • Word

    relationships Target the keyword, but optimise for this. On-page Optimisation @RoryT1
  6. Structure landing pages to help Google understand context & how

    it meets the needs of the searcher @RoryT1
  7. There are four ways you can get this. You need

    Jupyter Notebook What is that?
  8. The Jupyter Notebook is an open-source web application that allows

    you to create and share documents that contain live code, equations, visualizations and narrative text. Uses include: data cleaning and transformation, numerical simulation, statistical modeling, data visualization, machine learning, and much more. Jupyter.org @RoryT1
  9. Jupyter Notebook is an environment on my laptop where I

    can learn Python by copying scripts created by people significantly smarter than me and breaking them or making them do something slightly different. Rory Truesdale Python Charlatan @RoryT1
  10. Resources to get started… Jupyter Notebook – Getting Started Guide

    Robin Lord Find scripts Paul Shapiro JR Oakes Hamlet Batista Find scripts Find scripts
  11. I’ll share a link to a Dropbox with everything you

    need to get you started @RoryT11
  12. Lower case avoids duplication & punctuation adds no value to

    this analysis Lower Case & Remove Punctuation @RoryT1
  13. The process of chopping up a sentence into individual pieces,

    called ‘tokens’ Tokenization @RoryT1
  14. The process of converting a word to its root (i.e.

    “playing” becomes “play”) Lemmatization (optional) @RoryT1
  15. How many times does a word or combination of words

    appear in your SERP content? Co-occurrence @RoryT1
  16. Co- occurrence Understand the types of phrases that Google sees

    as semantically relevant to a target keyword set
  17. •Additional source of data for keyword research •Identify topical content

    gaps on landing pages •Optimise landing page content by incorporating semantically relevant phrases HOW CAN WE APPLY THIS? @RoryT1
  18. Cost: Range: Time to Charge: Battery Size/Capacity: All Wheel Drive:

    Towing Capacity: Semi-Conductor SERP XLT: Product Page £ 44,360 MSV 9,620 MSV 7,470 MSV 380 MSV 3,040 MSV 180 MSV
  19. What are the most frequently occurring nouns, verbs & adjectives

    in a SERP? Part of Speech Tagging @RoryT1
  20. PoS Tagging Uncover the phrases or topics you should include

    in your landing pages to rank for a term Nouns (people, place, thing) @RoryT1
  21. PoS Tagging Get clues around how Google is interpreting the

    context and intent of a search Verbs (action or state) @RoryT1
  22. PoS Tagging Understand the language and tone that might resonate

    with a searcher Adjectives (descriptive word) @RoryT1
  23. PoS Tagging Credit Card Example – P1 Verbs Intent Clues:

    What is the specific motivation our searcher has?
  24. PoS Tagging Credit Card Example - P1 Nouns Context Clues:

    Words that clarify meaning & help us understand what a searcher wants @RoryT1
  25. PoS Tagging Credit Card Example - P1 Adjectives Context Clues:

    Words that clarify meaning & help us understand what a searcher wants @RoryT1
  26. •Create landing pages that are aligned with the intent of

    a searcher •Help copywriters understand the language and desires of a target audience •Tactically incorporate more semantically relevant phrases into landing pages HOW CAN WE APPLY THIS? @RoryT1
  27. Can we use NLP to uncover topical trends in the

    SERPs to help us with content ideation? Topic Modelling @RoryT1
  28. Topic Modelling Topic modelling is an NLP method that assumes

    a corpus contains a mixture of topics. It looks at how words and phrases co-occur in a corpus and attempts to group them in coherent themes or topics. @RoryT1
  29. Topic Modelling OK, computer. Here’s some words. Group them. @RoryT1

    Rory Truesdale Cheapening machine learning since 2019
  30. Topic Modelling The output is an interactive visual on topical

    trends that can be easily shared with other teams @RoryT1
  31. Topic Modelling Uncover topical trends hidden in the language of

    the SERPs that can inform content ideation @RoryT1
  32. •Valuable data point to reference for content ideation •Inform internal

    linking and content recommendations across a website •Incorporate topically relevant phrases into existing pages to improve semantic relevance HOW CAN WE APPLY THIS? @RoryT1
  33. How can we make our scripts work across other data

    sources to understand our customers? Other Useful Applications @RoryT1
  34. With some minor tweaks we can make our scripts work

    across a huge corpus of user- centric content Pretty cool, right? @RoryT1
  35. Potential to ramp up and apply sentiment analysis to these

    sources for useful visualisations @RoryT11
  36. Deconstruct product reviews to find out what really matters to

    customers •Simple •Easy to use •Intuitive •Buggy •Slow @RoryT1
  37. Scripts we create allow us to get these insights from

    lots of other user-centric sources beyond the SERPs @RoryT1
  38. • https://www.searchenginejournal.com/scrape-google-serp-custom-extractions/267211/ • https://www.searchenginejournal.com/mine-serps-seo-content-customer-insights/311137/ • https://www.seerinteractive.com/blog/user-testing-serps-an-audience-first-approach-to-seo/ • https://www.dropbox.com/sh/vl5miyt6sgbvmkl/AAC5365YcWTun_EzkQLtixe1a?dl=0 (Jupyter Notebook

    tutorial) • http://www.blindfiveyearold.com/algorithm-analysis-in-the-age-of-embeddings • https://www.searchenginejournal.com/semantic-search-seo/264037/#close • https://www.slideshare.net/DawnFitton/natural-language-processing-and-search-intent- understanding-c3-conductor-2019-dawn-anderson • https://moz.com/blog/what-is-semantic-search • https://www.slideshare.net/paulshapiro/redefining-technical-seo-mozcon-2019-by-paul-shapiro Useful Resources @RoryT1