CONTEXT: WHAT’S BIG DATA?
WELCOME TO DATA OBESITY!
5
http://www.datasciencecentral.com/profiles/blogs/basic-understanding-of-big-data-what-is-this-and-how-it-is-going
CONTEXT: WHAT’S BIG DATA?
HOW BIG IS BIG?
6
http://www.domo.com/blog/2013/05/the-physical-size-of-big-data/
in 1 year!
creates enough data to fill
CONTEXT: WHAT’S BIG DATA?
BIG IN GROWTH, TOO.
7
http://www.infosysblogs.com/brandedge/2013/04/20130419Infographc.html https://studentforce.wordpress.com/2013/09/21/umuc-big-data-revolution-is-here/
CONTEXT: WHAT’S BIG DATA?
9 SOURCES
8
https://studentforce.wordpress.com/2013/09/21/umuc-big-data-revolution-is-here/
CONTEXT: WHAT’S BIG DATA?
6 TYPES
9
{
"created_at": "Thu Sep 15 16:29:08 +0000 2016",
"id": 776457834095644700,
"id_str": "776457834095644672",
"text": "I love @glip because it makes me more productive and
reliant on far fewer tools! #gliplove #goglip #gliptastic :)",
"truncated": false,
"entities": {
"hashtags": [
{
"text": "gliplove",
"indices": [
82,
91
]
},
{
"text": "goglip",
"indices": [
92,
BUILDING DASHBOARDS
SCHEDULED SEARCHES + INDICES
20
by sourceoforder by af_type
by af_source by af_name
by af_name
+ ppc_s
SUMMARY INDEX INDEX_MAIN_SOURCES INDEX_A.COM INDEX_AFFILS INDEX_PAID_SEARCH INDEX_SHOPPING_ENGINES
12M+ TRANSACTIONSFULL DB
METRICS
SAVED SEARCHES
DATA DELTAS
METRICS DATA DELTAS
METRICS DATA DELTAS
METRICS DATA DELTAS
METRICS DATA DELTAS
BUILDING DASHBOARDS
AUTOMATED BIG DATA FLOW (EXAMPLE 1)
28
RAW DATA EXTRACT LOADTRANSFORM
Traffic Sources
& Session Stats
RAW DATA
Behavioral Segments,
Funnels, Retention &
LTV
EXTRACT
Additional aggregation
and data refinement
Core Website
Social Engagement Footprint
Unified social
footprint metrics
Enrichment of
email addresses
CRM data store for
easy segmentation +
analysis
Additional context
on Twitter followers
More flexible segments,
funnels + retention
metrics
BUILDING DASHBOARDS
AUTOMATED BIG DATA FLOW (EXAMPLE 2)
29
LOAD
Custom
dashboards
synced with
70+ APIs
Traffic Sources
& Session Stats
Realtime (RT)
TRANSFORMRAW DATA EXTRACT
Core Website
Social Engagement Footprint
heavy-duty query tools already in place
App Databases
Custom aggregation scripts
Postgres or
Redshift DB
Daily
Pull
Internally-reported metrics
summarized for triangulation
Daily CSV
Behavioral segmentation
+ in-app messaging
RT
Behavioral Segments,
Funnels, Retention & LTV
RT
Unified social
footprint metrics
Unified app downloads
& ratings metrics
App Store Activity
email address
enrichment
BUILDING DASHBOARDS
AUTOMATED BIG DATA FLOW (EXAMPLE 3)
30
LOAD
Custom
dashboards
synced with
70+ APIs
Traffic Sources
& Session Stats
Realtime (RT)
TRANSFORMRAW DATA EXTRACT
Core Website
MongoDB
Custom aggregation scripts
MySQL
Presence Table
Internally-reported metrics
summarized for triangulation
Weekly
CSV
Behavioral segmentation
+ in-app messaging
RT
Behavioral Segments,
Funnels, Retention & LTV
RT
Unified app downloads
& ratings metrics
App Store Activity
email address
enrichment
Daily Dump
Instant dashboards
for Intercom
INFERRING SEGMENTS FROM BIG DATA 58
X11 X21 X31 X41
X12 X22 X32 X42
X13 X23 X33 X43
X14 X24 X34 X44
X15 X25 X35 X45
X51
X52
X53
X54
X55
High Frequency
High Recency
Low Frequency
Low Recency
Still Loyal
Once Loyal
New
Old
F + R = LOYALTY INSIGHTS
INFERRING SEGMENTS FROM BIG DATA
$0
$1,500
$3,000
$4,500
$6,000
59
Average total spent ($) by new MFR quantiles rerun for non-outlier M1 + M2 customers
M1 M2 M3 M4 M5
percent: top 20% of
M1+2
2nd 20% 3rd 20% 4th 20% Bottom 20%
segment
size:
93,134 93,139 92,861 93,406 93,143
avg. $
spent:
$3,337 $1,137 $642 $412 $276
total $
spent:
$345,234,826 $105,573,528 $59,348,459 $38,398,553 $25,537,936
% of total
revs:
53% 32% 18% 11% 8%
High-Value
Customers
Low-Value Customers
M = VALUE
INSIGHTS
INFERRING SEGMENTS FROM BIG DATA
PRICE-POINT
CUTOFFS
61
Best Camera/Lens Purchased
DSLR Body DSLR Lens DSLR Body + Lens Point-and-Shoot
Segment Name
Relationship
to Photography
Memory
Keepers
Use cameras to record
family memories and
milestones
less than
$650
less than
$300
less than
$950
less than
$450
Hobbyists
Enjoy the picture-taking
process; understand and
use camera controls
$650 - $1725 $300 - $750 $950 - $2300 $450 - $700
Prosumers
Advanced skills, but do
not make a living from
photography
$1725 - $2750 $750 - $3000 $2300 - $4200 $700 - $2500
Pros
Rely on photography as a
profession $2750+ $3000+ $4200+ $2500+
INFERRING SEGMENTS FROM BIG DATA
CROSSING RFM
W/ CATEGORIES
62
Low Value High Value
Still
Loyal
Once
Loyal
New Old
Still
Loyal
Once
Loyal
New Old
Memory Keepers 1 2 3 4 5 6 7 8
Hobbyists 9 10 11 12 13 14 15 16
Prosumers 17 18 19 20 21 22 23 24
Professionals 25 26 27 28 29 30 31 32
3. Cross-Tabulate
Top customers and categories to
create behavioral and
loyalty-based segments
9
key categories
account for 81% of sales
2. Isolate
the top customers and
categories by total dollars
spent, frequency, and
recency (RFM) measures
465,683
top customers
account for 88% of sales
1,164,927 customers 807 categories
1. Aggregate
72 months of Internet channel
transaction data, organizing by
key variables
2,246,094 Internet Channel transactions
4. Generate
Segment-specific marketing
recommendations which can
be further targeted by brand
YIELDS SOLID TARGETS
FOR TACTICAL PLANNING
FINAL THOUGHTS
A NEW TYPE OF
KNOWLEDGE
WORKER
66
http://www.doclens.com/87922/think-issue-7-2014/
FINAL THOUGHTS
AN INCREDIBLY VALUABLE SKILL
67
https://studentforce.wordpress.com/2013/09/21/umuc-big-data-revolution-is-here/
FINAL THOUGHTS
THE CORNERSTONE OF A DAUNTING FUTURE?
68
https://studentforce.wordpress.com/2013/09/21/umuc-big-data-revolution-is-here/
FINAL THOUGHTS
DATA AS INTERFACE
69
for
using
Made Visual
BACKGROUND TITLES + BUTTONS TEXT + LINES
Your data + brand
up to
100,000
objects
Anywhere on the Web
using
1 line
of code