Etuma 
Customer 
Feedback 
Analysis 
Making sense of 
customer emotions 
Matti Airas, CEO and Co-Founder, Etuma 
1
2 
Companies Are 
Facing Ever 
Increasing 
Competition 
• Brick vs. click 
• Competition crossing borders 
• Consumer trends are hard to 
predict and short lived 
• Consumers tend to be less loyal
There are things that your 
customers would like to tell 
you: selection, emerging 
trends, operational 
problems…
If you only 
listened 
4
5 
80% of all Data is 
Unstructured 
“Vos emballages 
sont trop fragiles” 
Please, send my 
purchase in a 
stronger container 
next time. 
My sunglasses 
were broken." 
“Ensinnäkin 
toimitus kesti yli 
kolme viikkoa ja 
lisäksi tilaamani 
aurinkolasit olivat 
vahingoittuneet." 
“Your website does 
not have the 
information to 
make a purchase 
decision”
Etuma turns 
open text 
into 
structured 
intelligence 
6
7 
Etuma has solved the key 
problems in customer 
feedback analysis 
Real-time 
Reporting 
Actionable 
Information 
Ambiguous 
Language 
Results can be used 
directly in decision 
making and are 
consistent over time for 
trend analysis. 
Results are available 
immediately after the 
customer submits the 
feedback. 
Language is a living thing. Etuma 
uses a combination of machine 
learning and manual optimization 
work by Etuma computational 
linguistics experts to keep the 
system always up to date.
8 
Text to 
Statistics
9 
language the same way a 
TOPIC-LEVEL 
SENTIMENT 
SEMANTICS 
Etuma understands 
SYNTAX 
human does 
MORPHOLOGY 
PART OF 
SPEECH 
LANGUAGE 
RECOGNITION 
Optimized for 
customer 
experience! 
Relevant and consistent 
Optimized for customer 
experience 
Relevant and consistent 
analysis results 
Fast adaptation to any 
feedback channel and type 
analysis results! 
Fast adaptation to any 
feedback channel and type!
LANGUAGE 
RECOGNITION 
MORPHOLOGY& 
PART OF SPEECH SYNTAX 
“Please, send my purchase in a stronger container next 
time. My sunglasses were damaged.” 
ENGLISH 
please (ADVERB) 
send (VERB) 
my (PRONOUN) 
purchase (NOUN) 
in (PREPOSITION) 
a (ARTICLE) 
strong 
(ADJECTIVE) 
container (NOUN) 
send (VERB PHRASE) 
my purchase 
(NOUN PHRASE) 
stronger container 
(NOUN PHRASE)
11 
Connect 
Feedback 
Channel(s) 
Analyze and 
Enrich With 
Structured 
Information 
Correlate, Model 
and Recognize 
Patterns 
Visualize, 
“Taskify”
Easy to 
implement 
cloud service 
12 
1. 
Optimize 
analysis based 
on historical 
data. 
2.3. 
Define roles 
and scenarios. 
Create users, 
design 
reports. 
Connect 
feedback 
channels 
using web 
interfaces. 
All you need to do is 
participate in one workshop 
and implement the 
connectors.
13 
Lower Churn - 
Increased loyalty 
NEED 
• Lot of customer data but inability 
to predict when customers are 
about to churn or reduce buying 
SOLUTION 
• Predicting which customers are 
about to leave the service or reduce 
buying and do personalized 
communication. 
RESULTS 
• Reduce churn first year 1%, next 
year 5%, 3rd year 10%
14 
Thank You! 
To continue discussion please contact 
matti.airas@etuma.com 
All picture under https://creativecommons.org/licenses/by-nd/2.0/

Etuma Customer Feedback Analysis - how to keep your customers loyal

  • 1.
    Etuma Customer Feedback Analysis Making sense of customer emotions Matti Airas, CEO and Co-Founder, Etuma 1
  • 2.
    2 Companies Are Facing Ever Increasing Competition • Brick vs. click • Competition crossing borders • Consumer trends are hard to predict and short lived • Consumers tend to be less loyal
  • 3.
    There are thingsthat your customers would like to tell you: selection, emerging trends, operational problems…
  • 4.
    If you only listened 4
  • 5.
    5 80% ofall Data is Unstructured “Vos emballages sont trop fragiles” Please, send my purchase in a stronger container next time. My sunglasses were broken." “Ensinnäkin toimitus kesti yli kolme viikkoa ja lisäksi tilaamani aurinkolasit olivat vahingoittuneet." “Your website does not have the information to make a purchase decision”
  • 6.
    Etuma turns opentext into structured intelligence 6
  • 7.
    7 Etuma hassolved the key problems in customer feedback analysis Real-time Reporting Actionable Information Ambiguous Language Results can be used directly in decision making and are consistent over time for trend analysis. Results are available immediately after the customer submits the feedback. Language is a living thing. Etuma uses a combination of machine learning and manual optimization work by Etuma computational linguistics experts to keep the system always up to date.
  • 8.
    8 Text to Statistics
  • 9.
    9 language thesame way a TOPIC-LEVEL SENTIMENT SEMANTICS Etuma understands SYNTAX human does MORPHOLOGY PART OF SPEECH LANGUAGE RECOGNITION Optimized for customer experience! Relevant and consistent Optimized for customer experience Relevant and consistent analysis results Fast adaptation to any feedback channel and type analysis results! Fast adaptation to any feedback channel and type!
  • 10.
    LANGUAGE RECOGNITION MORPHOLOGY& PART OF SPEECH SYNTAX “Please, send my purchase in a stronger container next time. My sunglasses were damaged.” ENGLISH please (ADVERB) send (VERB) my (PRONOUN) purchase (NOUN) in (PREPOSITION) a (ARTICLE) strong (ADJECTIVE) container (NOUN) send (VERB PHRASE) my purchase (NOUN PHRASE) stronger container (NOUN PHRASE)
  • 11.
    11 Connect Feedback Channel(s) Analyze and Enrich With Structured Information Correlate, Model and Recognize Patterns Visualize, “Taskify”
  • 12.
    Easy to implement cloud service 12 1. Optimize analysis based on historical data. 2.3. Define roles and scenarios. Create users, design reports. Connect feedback channels using web interfaces. All you need to do is participate in one workshop and implement the connectors.
  • 13.
    13 Lower Churn- Increased loyalty NEED • Lot of customer data but inability to predict when customers are about to churn or reduce buying SOLUTION • Predicting which customers are about to leave the service or reduce buying and do personalized communication. RESULTS • Reduce churn first year 1%, next year 5%, 3rd year 10%
  • 14.
    14 Thank You! To continue discussion please contact matti.airas@etuma.com All picture under https://creativecommons.org/licenses/by-nd/2.0/