How to avoid being a BIG DATA ZOMBIE www.brandwatch.com |: @brandwatch | Tel: +44 (0)1273 234 290
May 18, 2015
How to avoid being a BIG DATA ZOMBIE
www.brandwatch.com |: @brandwatch | Tel: +44 (0)1273 234 290
© 2012 Brandwatch | www.brandwatch.com 2
Brandwatch in 1 slide
Session structure
© 2012 Brandwatch | www.brandwatch.com
• Introduction – evaluating your reporting
• Workshop – approaching analysis
• Summary – relevant data not big data
4
Introduction
© 2012 Brandwatch | www.brandwatch.com
5© 2012 Brandwatch | www.brandwatch.com
When was the last time you
updated your reporting?
6© 2012 Brandwatch | www.brandwatch.com
Do your metrics match your
objectives?
7© 2012 Brandwatch | www.brandwatch.com
What have you done as a
result of reporting insights?
© 2012 Brandwatch | www.brandwatch.com 8
Workshop
iStrategy Zombie Data workshop
© 2012 Brandwatch | www.brandwatch.com
Objective of next 30 minutes is to explore some big data about the social media impact of M&S 2012 Christmas advert
This is big data: 10,000+ data points in 3 weeks with lots of associated data (location of sender, time of day, language used, response/retweets, etc)
Imagine you are a Marketing Executive in M&S and you are trying to assess whether the TV campaign is working, what to tweak/change and what lessons to learn
Objective is not to judge M&S Xmas campaign. (They are not a client and all data is available publicly - although you’d need to a tool like Brandwatch to collect and make sense of it)
A1: Number of social media web pages (tweets, blogs, forums, etc) that mention M&S
© 2012 Brandwatch | www.brandwatch.com
01 Nov 1203 Nov 12
05 Nov 1207 Nov 12
09 Nov 1211 Nov 12
13 Nov 1215 Nov 12
17 Nov 1219 Nov 12
21 Nov 12
0
1,000
2,000
3,000
4,000
5,000
6,000
Men
tio
ns
per
day
of
M&
S
Advert first broadcast
Question Your answer
What does the chart tell you? Mentions rose after advert was launched, then declined
What objective does this help to measure? Campaign Awareness
What action could you take based on this data?
Not much
What would make it better? Some context, work out how much of overall volume is about advert, did it link to sales, how well did we expect it to do, etc, etc
Zombie Score
50 = Useful data5 = Zombie slide
A2: Topic cloud of social media webpages that mentions the M&S advert
© 2012 Brandwatch | www.brandwatch.com
Question Your answer
What does the chart tell you? People talked on social media about M&S adverts
What objective does this help to measure? Campaign Awareness and reason why
What action could you take based on this data?
Investigate angle of Down’s Syndrome, GemmaAnneStyles
What would make it better? Get rid of the mentions of M&S/adverts/Xmas and get more topics in this + have it coloured by sentiment + have verbatim / explanation of main points
Zombie Score (0-5)
4
A3: Most used twitter hashtags associated with M&S adverts
© 2012 Brandwatch | www.brandwatch.com
0
10
20
30
40
50
60
70Most used hashtags
Tw
eets
@annefoxycoxy
303 RTs, 804k followers
“ It'll be here before we know it RT @GemmaAnneStyles: Always enjoy a good M&S advert. Bit early for Christmas but I'll let it slide. ”
Most retweeted tweet
@carolineplayle (‘Seb’s Mum’)
60 (M&S ad) tweets, 605 followers
“ M&S in store mag is out today! Not just Christmassy, Seb is in it and it earned him his place in the TV Ad! A little piece of BIG history! X ”
Most prolific tweeter
Question Your answer
What does the chart tell you? Hashtags were quite well used but a wide spread
What objective does this help to measure? Engagement with campaign
What action could you take based on this data?
Campaign using hashtags or respond to some hashtags
What would make it better? More holistic view of engagement (rather than focus on just twitter hashtags)
Zombie Score (0-5)
3
B1: Number of social media web pages (tweets, blogs, forums, etc) that mention M&S
© 2012 Brandwatch | www.brandwatch.com
01 Nov 1203 Nov 12
05 Nov 1207 Nov 12
09 Nov 1211 Nov 12
13 Nov 1215 Nov 12
17 Nov 1219 Nov 12
21 Nov 12
0
1,000
2,000
3,000
4,000
5,000
6,000
Men
tio
ns
per
day
of
M&
S
Advert first broadcastM&S announce 10% fall in profits
Question Your answer
What does the chart tell you?
What objective does this help to measure?
What action could you take based on this data?
What would make it better?
Social Competition ended
Increased showing of ads
Zombie Score (0-5)
B2: Example mentions of Seb White within the M&S advert
© 2012 Brandwatch | www.brandwatch.com
TWITTER@freyamulderrig
FACEBOOKSally Hewerdine
FORUMAdifferentopinion.net
TWITTER@caroline_s
“ I'm well happy m&s have a child with Down's syndrome modelling their clothes awwwww he's so sweet”
“ LOVE this! You're a star @carolineplayle - M&S TV advert to feature disabled boy Seb White bit.ly/tt.34687 via @Independent”
“ So chuffed and proud of M&S, they have ditched the usual celebrities for their 'Christmas Advertising' and have chosen regular kids, including a staring role for a gorgeous little boy with Downs Syndrome. Brilliant, will definitely be buying some Crimbo presents from their store now ... ”
“ Good on M&S for putting Seb RIGHT in the public face. A lot of people had their cloised minds opened with the Para-Olympics. Seb can play his part too! ”
Zombie Score (0-5)
Question Your answer
What does the chart tell you?
What objective does this help to measure?
What action could you take based on this data?
What would make it better?
B3. Number of social media webpages that mention the adverts of 5 different UK retailers
© 2012 Brandwatch | www.brandwatch.com
John Lewis Asda Marks and Spencer Waitrose Littlewoods0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
45,000
50,000
Question Your answer
What does the chart tell you?
What objective does this help to measure?
What action could you take based on this data?
What would make it better?
Zombie Score (0-5)
B4. Number of tweets about adverts of 5 UK retailers and potential reach of those tweets
© 2012 Brandwatch | www.brandwatch.com
John Lewis
Asda Marks and Spencer
Waitrose Little-woods
0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
Tw
eets
John
Lew
isAsd
a
Mar
ks a
nd S
penc
er
Wai
trose
Little
woods
0
5,000,000
10,000,000
15,000,000
20,000,000
25,000,000
30,000,000
35,000,000
Po
ten
tial
rea
ch(t
ota
l fo
llo
wer
s)Question Your answer
What does the chart tell you?
What objective does this help to measure?
What action could you take based on this data?
What would make it better?
Zombie Score (0-5)
B5. Ratio of positive/negative mentions about adverts for 5 UK retailers using auto sentiment
© 2012 Brandwatch | www.brandwatch.com
Asda John Lewis Littlewoods Marks and Spencer
Waitrose0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Positive
Negative
Question Your answer
What does the chart tell you?
What objective does this help to measure?
What action could you take based on this data?
What would make it better?
Zombie Score (0-5)
C1. Summary of how promoting adverts worked on Twitter, YouTube and Facebook
© 2012 Brandwatch | www.brandwatch.com
AdvertTwitter YouTube Facebook
Re-tweets
Favour-ites Views Comments Likes Dislikes Page
likesVideo post comments
Video post likes
Asda n/a n/a 126k 0 311 433 555k No post No post
John Lewis 555 105 1,878k 1,631 9,966 304 543k 2,401 553
Littlewoods 2 0 7k 9 9 1 141k 14 5
Marks and Spencer
30 3 177k disabled disabled disabled 917k 669 91
Waitrose 31 2 88k 14 39 9 n/a n/a n/a
Question Your answer
What does the chart tell you?
What objective does this help to measure?
What action could you take based on this data?
What would make it better?
Zombie Score (0-5)
C2. Comments on video when posted on Facebook
© 2012 Brandwatch | www.brandwatch.com
Question Your answer
What does the chart tell you?
What objective does this help to measure?
What action could you take based on this data?
What would make it better?
Zombie Score (0-5)
John Lewis 2012 Xmas ad
Marks and Spencer 2012 Xmas Ads
Littlewoods 2012 Xmas Ad
Typical Social Brand Video*
Marks and Spencer 2010 Xmas Ads
0
500
1000
1500
2000
2500
3000
2401
669
14 10 67
* Based on 250 companies that entered the 2012 Social Brands Top 100 awards run by Headstream and Brandwatch
D1. Your turn…
© 2012 Brandwatch | www.brandwatch.com
Please use the next slide to draw your perfect chart M&S chart (free of zombie-insight)
Here are some examples of the information it could include:
• Include other sales: e.g., ad spend, website visits, store visits, panel data, etc
• Get more verbatim and top mentions (e.g., by celebrities)
• More demographics on who talked about the adverts
• More detail on what aspects of the advert people liked
• More detailed historic comparison with previous Christmas’
D2. <Insert title>
© 2012 Brandwatch | www.brandwatch.com
Zombie Score (0-5)
O
22
Summary
© 2012 Brandwatch | www.brandwatch.com
23© 2012 Brandwatch | www.brandwatch.com
‘BIG’ DATA?
24© 2012 Brandwatch | www.brandwatch.com
SIZE ISN’T THE ONLY
THING THAT MATTERS
* SORRY.
*
25© 2012 Brandwatch | www.brandwatch.com
BIG = RELEVANT
26© 2012 Brandwatch | www.brandwatch.com
“…data helps us answer questions; big data also helps us conceive new questions.”*
Philip Sheldrake, Euler Partners, and author ‘The Business of Influence’