Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Find out how DataScience has revolutionized SEO for OVH

481 views

Published on

Conference Data-SEO Next level ( OVH Summit )
> FR:https://youtu.be/5iDBW3F_cYM
> EN:https://youtu.be/51TI_pkBgLY

Published in: Data & Analytics
  • Be the first to comment

  • Be the first to like this

Find out how DataScience has revolutionized SEO for OVH

  1. 1. DATA-SEO NEXT LEVEL VINCENT TERRASI / REMI BACHA HEAD OF DATA / HEAD OF SEO @vincentterrasi / @remibacha
  2. 2. Any sufficiently advanced technology is indistinguishable from magic ARTHUR CLARKE
  3. 3. BIG DATA ARTIFICIAL INTELLIGENCE
  4. 4. DATA SCIENCE PROJECT SEO PROJECTOVH
  5. 5. SEO IS A BIG DATA JOB
  6. 6. UNDERSTAND DATA MANIPULATE & ANALYSE BRING VALUE TO DATA DATA SCIENCE
  7. 7. EMPIRICISM 01. MAKE OBSERVATIONS 05BIS. REFINE, ALTER, EXPAND, OR REJECT HYPOTHESES 04. DEVELOP TESTABLE PREDICTIONS 02. THINK OF INTERESTING QUESTIONS 06. DEVELOP GENERAL THEORIES 05. GATHER DATA TO TEST PREDICTIONS 03. FORMULATE HYPOTHESES
  8. 8. DATA CENTRIC EMPIRICISM
  9. 9. RANK BRAIN
  10. 10. 01 02 03 04 CHANGING SEO FACTORS NEW FACTORS RANKING MISTAKES ULTRA- PERSONNALISATION
  11. 11. IT’S TIME TO UPGRADE SEO MACHINE LEARNING IA BIG DATA DATA SCIENCE DEEP LEARNING RANKBRAIN
  12. 12. WELCOME TO THE DATA SEO ERA
  13. 13. NEW JOB DATA SCIENTIST SEO
  14. 14. LEARNING DATA SCIENCE – Data Scientist Toolbox – Getting & cleaning Data – R / Python Programming – Explorary data – Machine Learning – Big Data
  15. 15. SEO DATAMART COMPETITORS OTHER TRAFFIC SOURCES DATA SOCIAL NETWORK SEARCH CONSOLE CRAWLS STOCK, PRICES, SALES DATA CUSTOMERS DATA EVENTS WEB ANALYTICS NETLINKING SEMANTICAL WEBPERFS SEARCH TRENDS SERVER LOGS
  16. 16. SEO DATAMART COMPETITORS CRAWLS NETLINKING SEMANTICAL WEBPERFS
  17. 17. XGBOOST 33TREES 10MAX DEPTH 100WAS GRID OF SIZE ROC AUC : 0.915 ? ? ? ? ? ? ? MOST IMPORTANT VARIABLES
  18. 18. Screamingfrog_in_csv Semrush_out_csv Screamingfrog_in_csv_pre pared Semrush_ screamingfrog_out _postgres Majestic_out_ postgres Visiblis_out_ postgres Semrush_ screamingfrog_ majestic_visiblis_ Prediction (XGBOOST_ CLASSIFICATION) on
  19. 19. DATAIKU DSS The most complete Data Science platform Data Preparation Machine Learning Deployment Collaboration
  20. 20. WHY PREDICT GOOGLE RANKINGS?
  21. 21. HOW TO PREDICT GOOGLE RANKINGS?
  22. 22. GETTING SERP DATA FROM SEMRUSH
  23. 23. CLEAN DATA REMOVE INVALID URLS Slow Crawl Rate Non-HTML Content Network Problems Slow Web Servers WAIT TIMES Errors from Web Servers URL Moved Permanently Redirect (301) URL Moved Temporarily Redirect (302) Authentication Required (401) or Document Not Found (404) Cyclic Redirects
  24. 24. CREATE PREDICTION MODEL
  25. 25. XGBOOST Adaptive boosting Gradient boosting Bagging Random forest BIAS RELATED ERRORS VARIANCE RELATED ERRORS
  26. 26. ? ? ? ? ? ? ? XGBOOST 33TREES 10MAX DEPTH 100WAS GRID OF SIZE ROC AUC : 0.915 MOST IMPORTANT VARIABLES ExtBackLinks RefDomains TrustFlow External Outlinks Response Time CitationFlow
  27. 27. TAKE AWAY … AUTOMATED MACHINE LEARNING WITH DATAIKU AUTOMATED KPI REPORTING SEO DATALAKE TEXT GENERATION OPPORTUNITIES DETECTION PREDICTIVE ANALYSIS PROCESS MINING AUTOMATED MACHINE LEARNING WITH DATAIKU SEO DATAMART
  28. 28. NOW, MACHINES CAN LEARN AND ADAPT, IT IS TIME TO TAKE ADVANTAGE OF THE OPPORTUNITY TO CREATE NEW JOBS. Data-SEO, Data-Doctor, Data-Journalist …
  29. 29. THANK YOU
  30. 30. GET ALL OUR LAST DISCOVERIES AND UPDATES Vincent TERRASI @vincentterrasi Remi BACHA @remibacha Data-seo.com Remibacha.com

×