19
Can Digital Metrics be a Substitute for Sales Data? Satya Menon Chief Scientist Millward Brown @satya_menon Ciju Nair VP Analytic Consulting Millward Brown @cijutr

Double-Take-Digital V02

Embed Size (px)

Citation preview

Page 1: Double-Take-Digital V02

Can Digital Metrics be a Substitute for Sales Data?

Satya MenonChief ScientistMillward Brown@satya_menon

Ciju NairVP Analytic ConsultingMillward Brown@cijutr

Page 2: Double-Take-Digital V02
Page 3: Double-Take-Digital V02
Page 4: Double-Take-Digital V02
Page 5: Double-Take-Digital V02

Sound familiar?

“What impact does social, digital and marketing in general, have on sales and brand equity?”

Page 6: Double-Take-Digital V02

Marketing Eco-SystemWhat is the Short-term Impact on business?What is Longer term Impact on brand equity?

EARNED OWNED

PAID

SHAREDEXPERIENCE

ATL/BTLOnline/Mobile adsBrand PartnershipsSearch

PR/NewsWOM/BuzzSocial MediaViral Video

Brand Website(s)Social MediaFan PagesCorporate EventsEmployees

Dealers/staffTest drivesAuto ShowsLoyalty programs

Page 7: Double-Take-Digital V02

Connected Data Helps Drive Modeling Insights….

SALES DATAIncludes Inventory, Incentives & Pricing

COMPETE PANEL

MACROECONOMIC INDICATORS

2 million panelistsOnline clickstream

behavior

SOCIAL DATA

Page 8: Double-Take-Digital V02

How does Digital influence Sales?

Digital & Social Media

Direct sales impact

In-direct sales impact

Owned Media

Earned Media

Other Sales Drivers

Brand HealthPaid Media Paid Media Halos from brand variantsPOE AmplificationMedia SynergiesCreative Wear-outMagic moments/Tent pole eventsFighting strategy & WeightsTrade activityTrade Halos from brand variantsProduct launches / innovation / cannibalizationSampling / Test drive etc.Events & SponsorshipsCo-Marketing Competitor TradeCompetitor MediaMacroeconomic factorsSeasonalityHolidays

Sales

Page 9: Double-Take-Digital V02

So what is new – Digital as leading indicator of Sales?

• Leading metrics

• Direct and indirect effects

• Multiplier effect

• Noise, threshold and other drivers

In summary, digital metrics can act as leading indicators of sales and hence also act of sales proxy in many categories

Page 10: Double-Take-Digital V02

Cleaning – Social dataSpam cleaning used to filter out non-US content and mentions of handles not belonging to the cosmetic brand.

CLEAN

OFF TOPIC

SPAM

TOTAL3,385,502

279,323(8%)

520,997(15%)

2,585,182

MATERESA JUAN TALISICReimbursement are APPLICABLE only in the Philippines and some other braches abroad.... Loreal Carlos Paris added Victor Speed Jackson to the conversation

GreenPetProducts@diez_maybelline thanks for the follow

TY MARIEL'oreal Blur foundation and concealer is AWESOME!!!! I love it!!! I bought mine at Target xoxo

GoCelebrityUSRT @msleamichele A signed bookplate, @LOrealParisUSA Colour Riche lipcolour AND an ambition card! http://t.co/IohqOU0fst

my-safest-havenutosnimavic:Nyx, Goddess of the night #Nyx #Goddess of Night #Mythology

Selflessism#ugh computer pls #i can't handle the torment #someone save me from my virus infected laptop #or place me in hybernation until christmas bc mac #ugh

Page 11: Double-Take-Digital V02

Since it was launched in early October 2013, Dodge Durango’s Ron Burgundy campaign has generated a True Reach® of 10 million views. A vast majority of the viewership has come from shared content as earned media amounting to roughly 10 times the media that was paid for by Chrysler.

Dodge Durango’s Ron Burgundy campaign

Page 12: Double-Take-Digital V02

Durango Campaign

* POE is Paid, Owned and Earned media. Figures in brackets are pre vs post launch of campaign

Media Contributions

5.2%

POE Amplification

2.6%(1.1% to 4.3%)

Online Engagement

12%

Durango TV Advertising

4.2%

Durango TV Advertising

2.3%

Durango Organic Search2.1%

Seasonality

1.7%

Durango Shoppers

3.8%

Gas Price-2.7%

Durango Online

Ads0.7%

Durango Print Ads

0.3%

Durango Online

Ads0.05%

Durango Print

Advertising0.02%

Other Dodge TV Advertising

0.07%

Other Dodge Online

Advertising0.2%

Durango Online News 2.9%

Dodge Durango Social2.1%

Incentives17.0%

(16.7% to 17.4%)

Segment Shoppers

1.0%

Category and Macroeconomic

-1.1%

Incremental Sales

37%

Base 63%

Paid media impact increased from 1.5% to 10.9%POE Amplification increased from 1.1% to 4.3% Online Engagement had high sales contribution at 12%

Page 13: Double-Take-Digital V02

Impact on Sales Pre & Post Campaign06

/25/

1207

/11/

1207

/27/

1208

/12/

1208

/28/

1209

/13/

1209

/29/

1210

/15/

1210

/31/

1211

/16/

1212

/02/

1212

/18/

1201

/03/

1301

/19/

1302

/04/

1302

/20/

1303

/08/

1303

/24/

1304

/09/

1304

/25/

1305

/11/

1305

/27/

1306

/12/

1306

/28/

1307

/14/

1307

/30/

1308

/15/

1308

/31/

1309

/16/

1310

/02/

1310

/18/

1311

/03/

1311

/19/

1312

/05/

1312

/21/

1301

/06/

1401

/22/

1402

/07/

1402

/23/

1403

/11/

1403

/27/

1404

/12/

1404

/28/

14

-400

-200

0

200

400

600

800

1000

1200

1400

1600

Durango Shoppers Leading 4 Weeks

Durango Incentives

Durango TV

Durango Online

Durango Print

Average Gas Price

Dodge Durango Social

Seasonality

Base

Estimated Sales

Contribution

Ron Burgundy campaign

Durango Shoppers was a leading indicator (4 weeks) predicting future sales

Page 14: Double-Take-Digital V02

Pre-launch social and online activity were leading indicators for PS4 sales

0

100000

200000

300000

400000

500000

600000

700000

800000

900000

Online Ad

Holidays

Facebook +_Twitter + YouTube

TV

Shoppers

Base

Estimated Sales

Contribution

PS4 launchElectronics Entertainment ExpoConference

PS4 TV followed by Online shoppers and Online media (comparable to TV) were the biggest drivers of PS4 sales. However online ads generated twice the shopper demand as TV

TV GRPs and Social conversation volume in FB, Twitter & Youtube during E3 conference had a lagged impact on PS4 launch sales

Page 15: Double-Take-Digital V02

Online behaviors act as lead indicators of Sales outcomes in other categories too…

0

2

4

6

8

10

12

14

16

18

0

100

200

300

400

500

600

700

0.000

0.000

0.000

0.000

0.000

0.001

0.001

0.001

0.000

50000.000

100000.000

150000.000

200000.000

250000.000

300000.000

350000.000

400000.000

7/3/

2010

7/17

/201

07/

31/2

010

8/14

/201

08/

28/2

010

9/11

/201

09/

25/2

010

10/9

/201

010

/23

/20

10

11/6

/201

011

/20

/20

10

12/4

/201

012

/18

/20

10

1/1/

2011

1/15

/201

11/

29/2

011

2/12

/201

12/

26/2

011

3/12

/201

13/

26/2

011

4/9/

2011

4/23

/201

15/

7/20

115/

21/2

011

6/4/

2011

6/18

/201

17/

2/20

117/

16/2

011

7/30

/201

18/

13/2

011

8/27

/201

19/

10/2

011

9/24

/201

110

/8/2

011

10/2

2/2

01

111

/5/2

011

11/1

9/2

01

112

/3/2

011

12/1

7/2

01

112

/31

/20

11

1/14

/201

21/

28/2

012

2/11

/201

22/

25/2

012

3/10

/201

23/

24/2

012

4/7/

2012

4/21

/201

25/

5/20

125/

19/2

012

6/2/

2012

6/16

/201

26/

30/2

012

7/14

/201

27/

28/2

012

8/11

/201

28/

25/2

012

9/8/

2012

9/22

/201

210

/6/2

012

10/2

0/2

01

211

/3/2

012

11/1

7/2

01

212

/1/2

012

12/1

5/2

01

212

/29

/20

12

2010 2012

Health & Beauty Brand Sales (offline)

Organic Search

Q4 2012Q1 2012

Brand Sales (Baseline)

Social Sentiment

Alcoholic Beverage

POE Amplification also varies across categories • 2-5% in Automotive• 10-15% in CPG• 20-30% in Gaming & technology products

Page 16: Double-Take-Digital V02

Sharing some thoughts on Social data and what next…

Every time we look at comparisons of social versus sales the same questions come up:

1. How much sales lag is due to category dynamics

2. How much of the sales lag is artificial and due to the different data reporting time frames

3. Understand threshold levels at which Social and other digital metrics start having a Sales / Brand Equity impact

4. Online vs Offline impact of social conversations

We would love to know the answers! “We are further exploring some partial truths we already know from experience (limited case studies) and academic work. “

Diminishing Return Optimal RangeWastage

Steep increase next $ more profitable than

previous $ spent

Slow increase and next $ more profitable than

previous $ spent

Media reaches saturation levels and next $ not as

profitable as last

Guiyang Xiong, Sundar Bharadwaj (2014) Prerelease Buzz Evolution Patterns and New Product Performance. Marketing Science http://dx.doi.org/10.1287/mksc.2013.0828

Page 17: Double-Take-Digital V02

…Re-Think Predictive Value in Social Conversations of key consumer segments over time

It is critical that we understand the Social/WOM “contagion effects” of advertising via sharing and conversations.

Inferring drivers of Future Brand Performance and Growth“We are also exploring how consistency of:1. Brand Love Scores (derived from social

conversations among consumers) and/or2. an intermediary metric like state dependence

behavior can also be used to classify and separate - loyals from switchers

and their flow/evolution patterns over time can be used to infer impact of social/WOM on sales and predict how different brands will fare in future”

Overwhelmedstrugglers

Price-sensitiveChillers

Stylish Fun-seekers

Go-with-the-Flow

TraditionalPerformers

Conscious Living

Page 18: Double-Take-Digital V02

Key Take AwaysThere is significant value in digital metrics as:

• Leading indicator of future Sales/Brand Equity

• As a Proxy for Sales and as a supplement to competitor intelligence

Key intermediary digital metrics we find to be of immense value in signaling intent to purchase and driving sales are:

• Online shoppers (Volume of shoppers engaging in any digital activity e.g. MB Digital)

• Organic search (Google, other search engines and site search e.g. Amazon, Ebay)

• Social buzz (Twitter, Facebook, Youtube etc.)

• Website visits (content/information sharing and ecommerce )

Estimation Bias when computing Digital ROIs

• TV ROIs are higher when including the amplification effects, it has in driving social and online shoppers

• Digital ROIs could look inflated at low levels of spend and expected to come down when scaled

Page 19: Double-Take-Digital V02

Can Digital Metrics be a Substitute for Sales

Data?

DOUBLE TAKE

Satya MenonCHIEF SCIENTISTMillward Brown Analytics@satya_menon

Ciju NairVP ANALYTIC CONSULTING

Millward Brown Analytics@cijutr