Taming the Hummingbird - #PMIEUR Berlin

Taming the Hummingbird Performance Marketing Insights Berlin, 2014

Jan-Willem Bobbink - @jbobbink #PMIEUR http://bit.ly/pmi-hummingbird

WHO SAW IT COMING?

ALGORITHM UPDATE

“ACCORDING TO GOOGLE, THIS NEW ALGORITHM IMPACTS 90% OF ALL QUERIES”

90% OF

ALL SEARCH QUERIES!

GOOGLE STARTED USING HUMMINGBIRD ABOUT A MONTH AGO, IT SAID. GOOGLE ONLY ANNOUNCED TODAY.

ADJUSTMENTS WERE ALREADY LIVE FOR MULTIPLE WEEKS

Performance Marketing Insights Berlin, 2014

Director of SEO @ Acronym Media – Blogging at www.notprovided.eu

WHAT CHANGED?

WHY SO MUCH CONFUSION?

Back to the basics

CONVERSATIONAL SEARCH EVEN MORE AWESOME!

That’s the knowledge graph!

"Hummingbird is focused more on ranking information based on a more intelligent understanding of search requests, unlike its predecessor, Caffeine, which was targeted at better indexing of websites."

GOOGLE: “TRY TO UNDERSTAND” RELATIONS

WHAT IS THE DIFFERENCE?

REWRITING THE QUERIES

HOW ABOUT ORGANIC RESULTS? [how old is the president of the United States, Barack Obama]

Biography is not in the query!

HOW ABOUT ORGANIC RESULTS? [age obama]

Keyword based!? [age obama]

RESULTING IN MORE RELEVANT SERPS

Semantically, related phrases will be those that are commonly used to discuss or describe a given topic or concept, such as “President of

the United States” and “White House.”

MONITOR SEO PATENTS

FIFA, ARE YOU WATCHING?

SEO PATENTS, NOT ONLY FOR THE NERDS!

- Learn from the past

- Predict future changes

- Get an idea about the inner working of search engines

- Get an idea what certain features take to exist

WHAT CAN YOU LEARN FROM PATENTS?

PREDECESSOR - 2004

PHRASE-BASED VS SINGLE KEYWORDS

KEYWORD BASED SCORING

“A document is retrieved in response to a query containing a number of

query terms, typically based on having some number of query

terms present in the document.”

Google research has shown that on more difficult queries, people start to type their searches as natural language questions. They also searched longer queries on average. This study also stated that, at the time of the study (2010), most of the time the question queries failed to give users the information they were looking for and they would revert back to keyword queries.

WHY DEVELOP HUMMINGBIRD?

DISTRIBUTION OF WEB SEARCH QUERIES [Lin et al. 2011]

THE HUMMINGBIRD PATENT?

REVISING SEARCH QUERIES

http://www.google.com/patents/US8538984

IT’S ALL ABOUT CONCEPTS

“The goal is that pages matching the meaning do better, rather than pages

matching just a few words.”

CAR VERSUS AUTO

CAR VERSUS AUTO

FULL QUESTION NOT NEEDED

Already filed by Google in 2005

DETERMINING QUERY TERM SYNONYMS WITHIN QUERY CONTEXT

http://www.google.com/patents/US7636714

HOW IS THE KNOWLEDGE GRAPH WORKING?

HOW ABOUT SCALE?

-YAGO: 10 million entities and 120 million facts -Freebase: 37 million topics, 1,998 types, and more than 30,000 properties

- DBpedia: 3.77 million things, 2.35 million classified in Ontology, including:

- 764,000 persons, 573,000 places, - 333,000 creative works, 192,000 organizations, - 202,000 species and 5,500 diseases. -111 languages, together 20.8 million things

Source: WSDM’14 conference, http://ejmeij.github.io/entity-linking-and-retrieval-tutorial/

INTERNATIONAL DIFFERENCES

CURRENT STATUS

GOOGLE.DE (GERMAN)

GOOGLE.ES (SPANISH)

GOOGLE.NL (DUTCH)

KNOWLEDGE GRAPH LOCALISED?

German Spanish Dutch

GOOGLE.DE (GERMAN)

GOOGLE.NL (DUTCH)

PRERENDERED QUERIES?

HOW CAN YOU DEAL WITH HUMMINGBIRD?

AS AN AFFILIATE?

HUMMINGBIRD MYTHS

GOOGLE KNOWS WHAT QUALITY IS

ADD MORE TEXT TO YOUR PAGES

LITERALLY ADD MORE QUESTIONS & ANSWERS

TEXTS IN THE FORM OF QUESTIONS

PUT MORE FOCUS ON LONGTAIL!

OK, LETS TURN IT AROUND

SRC: Searchmetrics.com 2014 US Ranking factors study

DETERMINE TARGET AUDIENCE

DEFINE INTENTION PER PERSONA

BUILD ENTITY SPECIFIC PAGES Using natural and semantically rich language:

Use entities in copy: 5 facts you have to know about Jan-Willem Bobbink being in Berlin at Performance marketing Insights

Entity attributes: 1987, Utrecht, SEO, Acronym, Physics etc.

So start with attribute stuffing instead of keyword stuffing

ORGANISE YOUR PAGES TOPICALLY

ENTITY BASED KEYWORD RESEARCH

GENERATE SYNONYM LISTS

http://www.performancemarketinginsights.com/14/europe/agenda/2/

SRC: http://www.alchemyapi.com/

Use Google’s Freebase API

https://developers.google.com/freebase/

RELATE CONTENT TO ENTITY

SRC: http://www.blindfiveyearold.com/knowledge-graph-optimization

sameAS EXAMPLE

GET LINKS FROM SEMANTICALLY RELEVANT SOURCES

http://semantic-link.com/#/berlin

FILL FREEBASE WITH RELEVANT INFORMATION

Google may present better SERPs also in terms of better ads

FROM ADVERTISING PERSPECTIVE

How to get in their?

HOW DOES GOOGLE DECIDE ON SOURCE?

SRC: Bill Slawski http://www.seobythesea.com/2013/05/google-knowledge-graph-results/

HOW TO DEAL WITH KNOWLEDGE GRAPH?

NO PATTERN FOUND YET

GOOGLE GETTING YOUR TRAFFIC?

https://class.coursera.org/nlangp-001

https://www.coursera.org/course/nlp

PDF: HTTP://NLP.STANFORD.EDU/IR-BOOK/PDF/IRBOOKPRINT.PDF

If you are interested in Natural language processing, read it:

AN INTRODUCTION TO INFORMATION RETRIEVAL

WANT TO KNOW MORE ABOUT

ENTITY RETRIEVAL AND LINKING?

http://ejmeij.github.io/entity-linking-and-retrieval-tutorial/

Questions? Don’t hesitate to ask! Or find me at the bar

http://bit.ly/pmi-hummingbird & @jbobbink

Image Credits Thanks for the images!

http://community.qlik.com/blogs/theqlikviewblog/2013/02/21/visualizations-the-tip-of-the-iceberg-of-understanding http://asset5.instanthumour.com/wp-content/uploads/2013/11/is-google-boy-or-girl-2.jpg https://encrypted-tbn2.gstatic.com/images?q=tbn:ANd9GcQovn4u6HtBbDB6uceJQLk7WyBNvRFKQLt2lYxF3gy94HRzosYG Images: http://img3.wikia.nocookie.net/__cb20130815124007/transformers-legends/images/c/c4/Triple-facepalm.jpg http://www.verticalresponse.com/blog/how-to-find-your-target-market/ http://footage.shutterstock.com/clip-2399522-stock-footage-reading-a-book-or-bible.html http://www.searchengineguide.com/matt-bailey/keyword-strategies-the-long-tail.php Europe: http://www.ezilon.com/maps/images/Europe-physical-map.gif Devices: http://www.google.com/insidesearch/features/search/assets/img/devices-preview.png Postcard: https://plus.google.com/u/0/+AmitSinghal/posts/AtndBA1pzNg Entity: http://www.entitythemovie.com/gallery Cutts meme: http://www.jacobking.com/affiliate-link-cloaking