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

Stop guessing what works on a page. Start extra...

Stop guessing what works on a page. Start extracting it.

Avatar for Joe Worgan

Joe Worgan

September 23, 2026

Other Decks in Marketing & SEO

Transcript

  1. A VERY, VERY QUICK INTRO • 5 years of experience

    in SEO • Working at Snaptrip Group - a travel accommodation provider
  2. AS SEOS, WE USUALLY OPTIMISE BASED ON INSTINCT, BEST PRACTICE

    OR WHAT WORKED LAST YEAR BUT…. WHAT IF THERE WAS SOMETHING WE COULD DO INSTEAD?
  3. WHAT IS IT? Custom extraction lets you pull any specific

    piece of information off a webpage, across thousands of URLs at once
  4. HOW TO DO IT Option 1: Option 2: Learn: •

    Regex • Xpath • CSS • Use an LLM to give you a hand 🤝
  5. NOW WHAT? • You’ve found the pages…. now what? •

    Connect up your GA4 or GSC account to tie this in with your performance metrics
  6. THE DATA Pages with FAQs grew: 65%+ Pages without FAQs

    grew: 48%+ • Pages with FAQs consistently outperformed those without • This isn't a best practice guess. It's from our own data Knowing that FAQs work isn't enough. We can work out which FAQs work better than others…
  7. LET’S GO EVEN DEEPER • Number of FAQs • ‘Show

    more’ button • Word count of FAQ answers • Number of internal links • FAQ Schema • FAQ position on page (above the fold v below) • FAQ heading text
  8. THE HOTEL GURU Does content freshness work on an editorial

    site? Let’s use custom extraction to find out!
  9. We’ve now worked out the last modified date of over

    2000+ pages Easily pick out pages with outdated content
  10. ORGANIC • Pages updated 6-12 months ago attracted most traffic

    • Freshness does matter, but so does giving time Google to react to the changes
  11. AI DATA • The AI pattern is almost identical to

    organic - win in search and win in AI too
  12. FIVE EXTRACTIONS IN ONE CRAWL 1) Datemodified Schema 2) Author

    name 3) Intro word count 4) Number of H2 sections 5) Number of internal links in the copy
  13. A FEW TAKEAWAYS • You don’t need to be a

    regex expert - use LLMs to your advantage!
  14. A FEW TAKEAWAYS • You don’t need to be a

    regex expert - use LLMs to your advantage! • Stop auditing pages one at a time - save yourself time!
  15. A FEW TAKEAWAYS • You don’t need to be a

    regex expert - use LLMs to your advantage! • Stop auditing pages one at a time - save yourself time! • Don’t just extract one thing - combine elements to give you the bigger picture
  16. A FEW TAKEAWAYS • You don’t need to be a

    regex expert - use LLMs to your advantage! • Stop auditing pages one at a time - save yourself time! • Don’t just extract one thing - combine elements to give you the bigger picture • Do this on your competitors’ site too
  17. A FEW TAKEAWAYS • You don’t need to be a

    regex expert - use LLMs to your advantage! • Stop auditing pages one at a time - save yourself time! • Don’t just extract one thing - combine elements to give you the bigger picture • Do this on your competitors’ site too • Get creative with what you extract - the granular the better!
  18. A FEW TAKEAWAYS • You don’t need to be a

    regex expert - use LLMs to your advantage! • Stop auditing pages one at a time - save yourself time! • Don’t just extract one thing - combine elements to give you the bigger picture • Do this on your competitors’ site too • Get creative with what you extract - the granular the better!