53
Advanced Text Analytics and Strategies That Influence Consumers WORDS OF WISDOM: David Richeson Chief of Digital Innovation & Influence Marina Maher Communications MARINA MAHER COMMUNICATIONS #SMW2017

Words of Wisdom

Embed Size (px)

Citation preview

Advanced Text Analytics and Strategies That Influence Consumers

WORDSOF WISDOM:

David Richeson

Chief of Digital Innovation & Influence

Marina Maher Communications

M A R I N A M A H E RC O M M U N I C A T I O N S

#SMW2017

Can we use data from social media conversation

to understand how to motivate your audience by

knowing exactly what they crave, fear, hate or

love?

Yes. We can. (With a little help from big data.)

How most brands market laundry detergent.

How consumers really feel about laundry.

By the end of this session, you’ll understand:

What advanced text analytics is.

How it can improve the way you market your

product, service or idea.

How you might apply these principles to drive

real influence and behavior change.

The next 45 mins (Prepare to nerd out a bit.)

1. The Fundamentals Of Influence

2. Advanced Text Analytics:

The Basics

3. A Look At Two Brands That

Implemented This Strategy

4. Rules Of Thumb

Let’s dial it back to the fundamentals.

How are people influenced?

1

World Of Communications Has Changed:

First Principles:

WHAT IS

OUR PURPOSE

AS MARKETERS?

In Short: To Get Person “A” To Think/Do/Say “B”

Our job is to influence people to

take a specific, desired action.

BA

But barriers exist between A and B. And that’s where we come in.

BA

But it’s not so easy anymore...

Welcome to theOpt-In World

TODAY’S REALITY

HOW DO WE GET INSIDE THE CIRCLE?

CREATED BRAND CONTENT

CO-CREATED CONTENT

USER-GENERATEDCONTENT

PAID SOCIAL AMPLIFICATION

PAID SEARCH

NATIVE ADVERTISING

EARNEDPLACEMENTS

AND BREAK DOWN

BARRIERS TO ACTION.

A B

Okay, so how do you begin?

How do you decipher what

your audience really believes?

2

What does this tell you?

We need to use statistical analysis to find themes in

the data.

Turning unstructured

data into structured

data.

WHAT CAN WE

DISCOVER FROM THIS

TYPE OF ANALYSIS?

We can

understand desires and interests.

Target Share of Needs States (based on conversation volume)

General Shampoo User Share of Needs States

Adventure 52%

Fun & Excitement

22%

Self 8%

Fashion 18%

Family & Home Life

34%

Self & Relationships

29%

Health 10%

Beauty 17%

We can

find what they are interested in.

Inspiration Driven Information Driven

Spas

Family

FashionTrends

BeautyTrends

HomeDecor Health & FitnessTravel

OutdoorActivities

DiningOut

Apparel/AccessoriesShopping

Yoga PersonalGroomingLuxury Autos

PersonalTechnology

BestsellerNovels

High Content Engagement

Gardening

We can

find what may make them switch brands.

Why they’re switching away from our brand.

100 110 120 130 140 150 160 170I don’t believe HE does what it says

I don’t really trust the HE brand

Isn’t the competitor brand better

I always use the highly rated brands

I don’t think the HE scent is that good

I’m looking for something better

I use HE once in a while

It’s not really a brand I care to use

I’m ready for a change

I only buy HE on sale

It’s too expensive

My needs have changed

Dandruff came back

Effectiveness

Still itchy hair

Price

Long-lasting bottle/package

Brand I trust

Scent

Recommended to me

Extra volume

Thicken my hair

Dry scalp

Natural ingredients

Healthier scalp

Long-lasting scent

MEN WOMEN181

171

144

140

138

133

122

121

117

114

114

114

109

109

177

154

148

121

109

139

127

129

130

98

129

120

119

114

Why they’re switching away from our brand.

We can

discover barriers to trial and advocacy.

-How do I get rid of dandruff?

-What are the best products?

3500

3000

2500

2000

1500

1000

500

0

Men Women Men Women Men Women

Research Comparison Purchase Trial Advocacy

Dandruff Shampoo Path to Advocacy, Men vs. Women

3500

3000

2500

2000

1500

1000

500

0

Men Women

3500

3000

2500

2000

1500

1000

500

0

Men Women

3500

3000

2500

2000

1500

1000

500

0

3500

3000

2500

2000

1500

1000

500

0

-Will this replace my regular

shampoo?

-Is it good for anti-itch?

-Do I need to go to a

dermatologist?

-I want to try it first

-This shampoo is expensive

-My dandruff came back

-I hate the scent

-Makes my scalp dry

-I recommend/love the product

-I will always use it

HOW DO WE

UNDERSTAND THEIR

UNDERLYING EMOTIONS?

Share of Top Emotions in the Laundry

Conversation, 2015

2%3%5%

8%

11%

18%

22%

31%

Stress

Fear

Satisfaction

Joy

Boredom

Acceptance

Displeasure

Sadness

Acceptance26%

Satisfaction39%

Fear

Joy7%

Unconventional14 %

Fun22%

Emotional Shifts Along YBF

Decision Path or Pods

100%

80%

60%

40%

20%

0%

Problem Recognition InformationSearch

Comparison ofCompetitive Set

Purchase Post-Purchase

Fear

Apprehension

Acceptance

Trust

Surprise

Annoyance

Anger

Anticipation

Joy

Cool.

So how can brands use this data?

3

The data helps tell us

how to talk in ways that overcome

barriers and even prompt advocacy.

BA

ENGAGEMENTSAMPLIFY

ENDORSEMENTS

Let’s share an example of

how it all works.

For our client’s Olympic program,

we needed to understand the

emotional drivers that could instigate

sharing to dominate the conversation.

Why they shared similar content.

2%3%8%

11%

16%

26%

34%

Thanking my mom/remembering mom

Beauty of perseverance/courage

Made me cry

Provide inspiration to someone else

Reconnecting with a family member

What is the music

Can relate on personal level

Emotions tied to sharing similar content.

300

250

200

150

100

50

0

Be

st

Gre

at

Am

azin

g

Pro

ud

Str

on

g

Go

od

Hard

Alive

Aw

eso

me

Str

on

ge

r

Be

au

tifu

l

Hap

py

Insp

irin

g

Tru

e

Yo

un

g

Irra

tio

nal

To

ug

hlo

ve

Th

an

kfu

l

Excit

ing

Cra

zy

Incre

dib

le

Po

ssib

le

Ind

ep

en

dan

t

Nic

e

Re

al

Sp

ecia

l

Un

co

nd

itio

nal

Co

nfi

de

nt

Wo

nd

erf

ul

Imp

ort

an

t

Pe

rfe

ct

Em

oti

on

al

Co

mp

ass

ion

ate

We also looked at Mother’s Day content.

Thanking Mom

What I am doing with mom on MD

What I am buying mom for MD

Remembering mom

First-time mom acknowledgements/advice

Describing how mom inspired me

Talking about childhood memory

Sharing cartoon/video that mom appreciates

Thanking dad/grandparents

Comparing mom to a celebrity

Sharing famous quote about motherhood

Discussing why mom punished me

We created an emotional connection with our demographic,

introducing them to the video in a meaningful way that

resulted in mass sharing and positive sentiment.

300

250

200

150

100

50

Be

st

Gre

at

Am

azin

g

Pro

ud

Str

on

g

Go

od

Ha

rd

Ali

ve

Aw

eso

me

Str

on

ge

r

Be

au

tifu

l

Ha

pp

y

Insp

irin

g

Tru

e

Yo

un

g

Irra

tio

nal

To

ug

hlo

ve

Th

an

kfu

l

Excit

ing

Cra

zy

Incre

dib

le

Po

ssib

le

Ind

ep

en

dan

t

Nic

e

Re

al

Sp

ecia

l

Un

co

nd

itio

nal

Co

nfi

de

nt

Wo

nd

erf

ul

Imp

ort

an

t

Pe

rfe

ct

Em

oti

on

al

Co

mp

ass

ion

ate

Thank You Mom Conversation Emotions

The content went viral and the client became the top Olympic

sponsor in conversation volume and sentiment at 100 days out,

with a 52% share of voice among global Olympic sponsors.

Olympics Sponsor Conversation Landscape During Sochi

Net Sentiment

Comp

Comp

Comp

Avg

. Daily

Me

ntio

ns

800

700

600

500

400

300

200

100

0

-100

Olympics Sponsor Conversation Landscape,

April 27 - May 10 (11am)

-5.0% 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0%

Previous Olympics

Conversation

Sentiment

Ho

urly

Me

ntio

ns

200

150

100

50

0

Comp

Client

Client

Current Olympics

Conversation

Comp Comp Comp Comp

0% 1% 2% 3% 4% 5% 6%

#1 branded conversation in the Olympics conversation (52% share of voice) among all global sponsors

73% more efficient: goal $.15 CPV; we delivered $.04 CPV

406% increase in sentiment

3,621% increase in conversation volume

This technique has yielded

incredible results on other

programs:

Leading Deodorant:

• Content engagement rate average: 15 - 20%

• Increased sales by 20% while blunting growth

of new competitor in the market

Leading Shampoo:

• Content engagement rate average: 14%

• Increased sales by 7%

Comprehending language

yields consistent results.

So how can I apply this thinking

to my brand’s social media

marketing strategy?

4

Start every program with some type of social listening. Keep an

open mind about what you’ll learn- you may find things that are

counterintuitive to what you thought.

Spend 70% of your time listening. 30% talking. Adapt as the

conversation changes. Be willing to change your strategy

accordingly.

Be willing to go outside your social channels to find your

audience - wherever they are.

Let’s wrap this up.

5

Understanding language is the key to influence in the modern world.

1. Find me afterwards in the Speaker AMA

Lounge

2. Email me for the slides if interested

3. Follow me on Twitter

4. I’ll be at SXSW. Let’s talk over a beverage

@dricheson

[email protected]

What can you do?

M A R I N A M A H E RC O M M U N I C A T I O N S

LET’S TALK: I’M LISTENING

M A R I N A M A H E RC O M M U N I C A T I O N S