Contenu connexe Similaire à What is phrasing - An explorative approach to improved user manipulation (19) What is phrasing - An explorative approach to improved user manipulation 1. © BSI – We grow businesses
AN EXPLORATIVE APPROACH TO
IMPROVED USER MANIPULATION
WHAT IS PHRASING?
Brand Science Institute, January 2019
2. AGENDA
© BSI – We grow businesses 2019 | Seite: 2
Core Issue
Procedure Model Phrasing
3. © BSI – We grow businesses 2017 | Seite: 3
“IT AIN’T WHAT YOU KNOW THAT GETS
YOU INTO TROUBLE. IT’S WHAT YOU
KNOW FOR SURE THAT JUST AIN’T SO.”
Mark Twain
4. SEARCH FOR RELEVANCE IN THE LONG-TAIL
Short Head:
General Keywords
High search volume
High competition
Low conversion rate
Long-Tail:
Specific keywords (chains)
Low search volume
Low competition
Higher conversion rate
© BSI – We grow businesses 2019 | Seite: 4
5. On the basis of identified keywords
and logical findings, chains of words
are formed and it is thought that the
user can put himself in a position to do
so. We see what we want to see...
The learned procedure is therefore
highly inefficient and leads to
numerous problems:
We don't know which question the
user really asks us.
We don't know what his (deep
psychological) motive is.
We do not know which structures
arise from millions of questions per
year.
We don't know which topics from the
user's questions are really relevant.
We do not know how the questions
and topic areas are connected.
Depth psychological and
structural drivers from
search behavior
PROBLEM WE DO NOT CREATE SUSTAINABLE RELEVANCE
© BSI – We grow businesses 2019 | Seite: 5
6. MODEL PHRASING
1 - EXPLORATORY DATA CUSTOMER 2 - ITERATIVE KEYWORD TEST 3 - TESTING KEYWORDS QS 4 - SCRAPING PHRASES
Structuring of screening contents of
the customer
Iterative testing of keywords
mediumKeyword Search Tools & Testing
for stable synonyms
Explorative Testing aller Keywords
in Search Suggest Methods Google
Identification of dominant phrases in
relevant markets (Scraping)
Clustering relevant topic areas Extraction of related keywords /
synonyms
Primary test of granular questionnaire
components and their composition
Test for stability of extracted questions
using Search Suggest Scraper
Extraction keywords & preparation for
analysis
Derivation of first relevant question
modules
Drilling Question building on
Keyletters questions
Summary & evaluation of the most
relevant questions
5 - QUANTITATIVE PATTERN 6 - QUALITATIVE MOTIVE 7 – RELEVANT QUESTIONS
Identification of structures (patterns)
in extracted questions
Analysis of question structure data on
underlying motives
Clustering Topics Phrases
Evaluation of structures (Patern)&
expressiveness for content building
alternatives
Identification of basic emotional states
and dominant motives
Descriptive presentation of relevant
issues
Explanation of Pattern Analysis Plot
to clarify relevant question structures
Comparison & evaluation of motifs in intra
brand competition
Derivation of recommendations from
phrasing for motif users, content
creation, etc..
8 – VERIFIZIERUNG STRUCTURE
Checking structures with the help of
semantic networks
Checking the validity of identified
clusters and topic areas
Identification of relationships between
topic areas and clusters
© BSI – We grow businesses 2019 | Seite: 6
7. RESULT TYPE I MOST RELEVANT QUESTIONS
welches xxx bei
welches xxx für
welches xxx bei
welches xxx hat am meisten
welches xxx bei
welches xxx schmeckt
welches xxx hat am meisten
welches xxx hat viel
welches xxx für
welches xxx ist gut bei
welches xxx hat viel
welches xxx für
welches xxx ist gut
wo kommt unser xxx her
wo kommt xxx vor
xxx wo kommt es her
das gesündeste xxx der
das beste xxx für
das richtige xxx
das beste xxx still
das beste xxx ohne
das beste xxx medium
das teuerste xxx
das gesündeste xxx
das beste xxx
das xxx der
das xxx mit den meisten
welches xxx für
welche xxx haben viel
welche xxx sind unbedenklich
welche xxx sind basisch
welche xxx wurden getestet
welche xxx sind nicht verunreinigt
welche xxx gehören zu
welche xxx gibt es in
welche xxx haben viel
welche xxx sind verunreinigt
welche xxx sind für geeignet
welche xxx sind natriumarm
welche xxx enthalten
welche xxx für
welche xxx sind für geeignet
7.
xxx
7.4
xxx bei
7.1
xxx ohne
7.2
xxx mit
7.3
xxx für
7.5
xxx gegen
7.6
Mixed
© BSI – We grow businesses 2019 | Seite: 7
8. ERGEBNISTYP II ABLEITUNG PSYCHOLOGISCHER MOTIVE
Within the sections, different types of questions and techniques were used:
1. General Associations
2. Query of positive and negative associations
3. Selective control of questions and matching
Question clustering
after multiple iterations
Analytical techniques
Psychology – Meaning of Words
Analysis of the emotional significance and underlying motives through the
creation of emotional perceptual spaces and the derivation of meaningfulness:
1. Analysis of the questions on the basis of individual words and breakdown of word
frequencies along different language dimensions
2. Derivation of multidimensional motif structures of over 8 million content components
and spaces of meaning
3. External validation of analysis results by sampling and experimental comparisons
Which question clusters arise in general?
How large are the clusters in terms of volume and degree of
differentiation?
Which patterns can be identified from the questions asked?
How can initial motives be derived from mere manual
classification?
How are the questions related to each other and how large is
the intercorrelation between the questions?
…
WC Sixltr Dic Negate Assent Affect Posemo Posfeel Optim Negemo Anx Anger Sad Swear
1073 30,75 68,59 3,91 0,47 7,46 5,13 0,47 0,37 2,33 0 1,49 0,37 0
1282 25,59 66,46 3,51 0,94 4,91 1,64 0,23 0,23 3,28 0,31 1,64 0,47 0,08
690 38,55 66,52 1,88 0,43 6,38 4,35 0,14 0,72 2,03 0,43 1,3 0,14 0
197 32,99 59,9 1,52 1,52 4,57 3,05 0 0 1,52 0 1,02 0 0,51
377 36,07 64,72 3,18 1,59 4,51 0,8 0 0 3,71 0 2,92 0,27 0
847 29,04 66,82 3,19 0,47 4,72 3,54 0,24 0,71 1,18 0 0,35 0,24 0
833 31,45 64,95 3,84 0 8,16 4,2 0,12 0,6 3,96 0,48 2,4 0,12 0
221 28,51 71,04 1,81 0,9 7,24 4,07 0,9 0,9 3,17 0,45 1,81 0,45 0
855 27,49 69,24 3,39 0,47 4,21 2,92 0,35 0,58 1,29 0 0,82 0,12 0
18 22,22 72,22 11,11 0 5,56 0 0 0 5,56 0 5,56 0 0
© BSI – We grow businesses 2019 | Seite: 8
9. RESULT TYPE II STRUCTURE RECOGNITION TOPIC FIELDS
© BSI – We grow businesses 2019 | Seite: 9