using ai to make sense of customer feedback

52
Using AI to Make Sense of Customer Feedback Alyona Medelyan @zelandiya

Upload: alyona-medelyan

Post on 24-Jan-2018

399 views

Category:

Data & Analytics


2 download

TRANSCRIPT

Page 1: Using AI to Make Sense of Customer Feedback

Using AI to Make Sense of

Customer Feedback

Alyona Medelyan

@zelandiya

Page 2: Using AI to Make Sense of Customer Feedback

Correct Understanding of Customer Feedback

Can Save Millions

Page 3: Using AI to Make Sense of Customer Feedback

2015: Tens of Thousands of New Zealanders

were Surveyed About the new Flag

Page 4: Using AI to Make Sense of Customer Feedback

Government Reported

the Results of Manual Feedback Analysis

Page 5: Using AI to Make Sense of Customer Feedback

Actual Responses

Two costly & unnecessary referendum followed. Outcome: NZ kept the current flag

Millions could have been saved!

People wanted to ”keep the current flag”

Page 6: Using AI to Make Sense of Customer Feedback

1. Types of customer feedback

2. Why analyzing customer feedback is important

3. Why is it hard

4. Approaches

5. Applying AI to customer feedback analysis

6. Demo

Page 7: Using AI to Make Sense of Customer Feedback

Different Types

of Customer Feedback

Page 8: Using AI to Make Sense of Customer Feedback

Types of Customer Feedback

one-on-one interviews / focus groups

call centre logs / complaints

social media

open-ended survey questions / reviews

quantitate survey questions

UX tests / analytics

unstructured

structured

Page 9: Using AI to Make Sense of Customer Feedback

Collection Analysis Insight

one-on-one interviews / focus groups hard hard good

call centre logs / complaints easy hard limited

social media easy hard limited

open-ended survey questions / reviews easy medium good

quantitate survey questions easy easy limited

UX tests / analytics medium easy limited

unstructured

structured

Comparing Types of Customer Feedback

Page 10: Using AI to Make Sense of Customer Feedback

Why Understanding

Customer Feedback

is More Important than Ever

Page 11: Using AI to Make Sense of Customer Feedback

Customer Experience

is the New Marketing

It’s Measured Using

Net Promoter Score Surveys

Image credit

Page 12: Using AI to Make Sense of Customer Feedback
Page 13: Using AI to Make Sense of Customer Feedback

The number of “Net Promoter Score”

searches on Google since 2004

1. Growing Number of

Satisfaction Surveys and Reviews

Page 14: Using AI to Make Sense of Customer Feedback

v

¯\_(ツ)_/¯

2. The Need to Explain

the Why’s Behind the Scores

Net Promoter Score by month over time

Page 15: Using AI to Make Sense of Customer Feedback

3. Scores can be Cheated

Unstructured Feedback, not so Much

Page 16: Using AI to Make Sense of Customer Feedback

Why Analyzing

Customer Feedback is Hard

Page 17: Using AI to Make Sense of Customer Feedback

Common Misconception:

Sarcasm Makes Analysis Hard

One of Many Sarcastic Tui Beer Adverts

Page 18: Using AI to Make Sense of Customer Feedback

Sarcasm is Hard: Even People Struggle

I’ll keep it in

mind

They’ll do itI’ve

forgotten

already

Page 19: Using AI to Make Sense of Customer Feedback

Sarcasm is Rarer Than You Think

Dataset Sarcasm Example

NPS Survey 1%I’m so disappointed! What a great

customer service you have!

Social Media

comments5% Very helpful answer. Troll.

Page 20: Using AI to Make Sense of Customer Feedback

The Actual Challenges

With Customer Feedback

Page 21: Using AI to Make Sense of Customer Feedback

Challenge 1: Messy Data

Page 22: Using AI to Make Sense of Customer Feedback

How many ways there are to say

‘wet paper’?

Challenge 2: Synonyms and Paraphrases

Page 23: Using AI to Make Sense of Customer Feedback

Hundreds of

possible variations

of the same theme

wet

dripping

soaking

soaked

damp

drenched

paper

papers

newspaper

news paper

newspapers

news papers

+

Paraphrasing the Same Theme

Page 24: Using AI to Make Sense of Customer Feedback

Challenge 3: Negation

Positive or Negative?

My coffee was great positive

My coffee was awful negative

My coffee was not great negative

My coffee was not that great neutral?

I did not think my coffee was great negative

I did not expect my coffee to be this great positive

I was disappointed with the quality of the coffee negative

I was not disappointed with the quality of the coffee positive

Page 25: Using AI to Make Sense of Customer Feedback

Approaches to

Customer Feedback Analysis

Page 26: Using AI to Make Sense of Customer Feedback

Manual Coding

1.

Page 27: Using AI to Make Sense of Customer Feedback

Figure out the Code Frame, Apply, Repeat

What is the meaning of life?

1 2 3 4 5

What is the meaning of life?

42

Friends and family

Making a difference in the world

Happiness

Finding happiness

To achieve, to conquer

Family

What is the meaning of life?

42

Friends and family

Making a difference in the world

Happiness

Finding happiness

To achieve, to conquer

Family

1

2

3

4

4

5

2

Page 28: Using AI to Make Sense of Customer Feedback

Sentiment in a Manual Code Frame

Customer Service

Positive Negative

Timely Nice Helpful Didn’t fix issue Rude

Page 29: Using AI to Make Sense of Customer Feedback

Word Clouds

2.

“Every time I see a word cloud presented as insight,

I die a little inside.”

– J. Harris, journalist

Page 30: Using AI to Make Sense of Customer Feedback

Word Clouds Lack

Interpretation, Context, Meaning

“Overall the language

focuses on sweeping

statements focusing on

the state of the nation.”

Kalev Leetaru (Forbes)

Page 31: Using AI to Make Sense of Customer Feedback

You wouldn’t create a Word Cloud from your Numbers,

why is it ok from Text?

Page 32: Using AI to Make Sense of Customer Feedback

Rule-based Approaches

3.

Page 33: Using AI to Make Sense of Customer Feedback

It’s Hard to Find a Rule That Works Well

I was impressed by how friendly the person

on the other end of the line wasStaff friendliness ✔

The lady who helped me was friendly Staff friendliness ✔

Friendliness of staff Staff friendliness ✔

Your website is very user friendly Staff friendliness ✘

The young man on the phone was very pleasant Other ✘

friendly OR friendliness –> Staff friendliness

Page 34: Using AI to Make Sense of Customer Feedback

Text Categorization

4.

Page 35: Using AI to Make Sense of Customer Feedback

old

customer

responses

categories

new

customer

responses

Machine

Learning

Algorithm

Predictive

Model categories

Need for Sufficient Training Data,

and Clear Categories

Page 36: Using AI to Make Sense of Customer Feedback

Customer Feedback Analysis

Needs to be ‘Unsupervised’

Page 37: Using AI to Make Sense of Customer Feedback

Thanks to an unsupervised approach, Facebook found

Candi Crash Saga causes low App Store reviews

Page 38: Using AI to Make Sense of Customer Feedback

Topic Modeling

5.

Page 39: Using AI to Make Sense of Customer Feedback

21

3

A Topic can be Hard to Interpret

2

???ok

Source: Ben Fields

Page 40: Using AI to Make Sense of Customer Feedback

Sentiment

Page 41: Using AI to Make Sense of Customer Feedback

1. Rule-based (dictionary)

2. Text categorization (positive / negative)

Two Sentiment Detection Approaches

Page 42: Using AI to Make Sense of Customer Feedback

Advances in AI > Customer Feedback

Page 43: Using AI to Make Sense of Customer Feedback

Messy Data

Paraphrases

Negation

AI > Challenges

Word2vec*

Deep Learning

*See also: Conceptnet.io

Knowledge Representation

Page 44: Using AI to Make Sense of Customer Feedback

Word2Vec

Image source: ericbern.com

Best Intro: Word2Vec Udacity Youtube

Page 45: Using AI to Make Sense of Customer Feedback

Knowledge Representation

Page 46: Using AI to Make Sense of Customer Feedback

Deep Learning

Precision Recall F-Measure Errors

People 84 73 75 <1

Dictionaries 61 57 54 8

Linear Regression 65 56 47 3

Deep Learning 62 57 49 2

Sentiment Analysis is not about maximizing F-Measure,

it’s about reducing true Errors: positive confused with negative

Page 47: Using AI to Make Sense of Customer Feedback

Theme Extraction

6.

Page 48: Using AI to Make Sense of Customer Feedback

From Words to Complex Themes

Page 49: Using AI to Make Sense of Customer Feedback

Applying Customer Feedback Analysis

Page 50: Using AI to Make Sense of Customer Feedback

Google: Sentiment by Theme

Page 51: Using AI to Make Sense of Customer Feedback

Thematic Demo

Page 52: Using AI to Make Sense of Customer Feedback

Thanks

@zelandiya

[email protected]

getthematic.com

linkedin.com/in/medelyan