Transcript Advanced Analytics Digital Distro - Hippocampus

Advanced Analytics Based on
Digital Data
© 2015 Blueocean Market Intelligence
1
Setup and integrated multiple data sources to provide
unified customer profiling and targeting platform
Client:
Leading PC Manufacturer in the US
Study Objective
•
•
Industry:
Technology
Business Impact:
Seamless data integration. Scaled
up customer Database.
Highly accurate and robust
platform acting as single source of
truth
•
•
The client wanted a unified view of their customers across their different business lines – B2C, B2B, EPP
and B2V
Insufficient customer details and incorrect data leading to inaccurate customer targeting causing low
customer conversions
Lack of integration among different databases and inability to link data among them hampered plans to
create a central database and “one view of the customer”
Inability to measure efficacy of marketing plans
Approach
•
•
Initial workshop:
− Blueocean Market Intelligence conducted detail and in-depth analytics workshop to build initial
understanding of business requirements and priorities
Data integration and storage
−
−
−
−
•
Setup the cloud based technology stack to deliver global and scalable solution
Developed host of APIs and other data bridges to integrate data from different client owned and 3rd party
platforms
Created metadata and schemas to be leveraged for scale
Leveraged industry MDM best practices to drive effective data storage practices
Business insights and deployment
−
Developed visualizations by leveraging industry leading platform Tableau™ and created various analytics
applications to meet business requirements
Result
•
•
•
•
•
•
Integrated data from 45+ sources and scaled up unified customer database
9000+ intelligence output delivered weekly
425+ business users leveraging the solution drive business strategies
Lift of 17.5% on the upsell and cross sell programs
Improved customer retention and enhanced customer experience by improving NPS at 20 basis points
Consolidated 360 view of customer information
© 2015 Blueocean Market Intelligence
2
Integrated multi channel data for optimized attribution
modeling and accuracy
Client:
Leading PC Manufacturer in the US
Industry:
Technology
Business Impact:
Study Objective
•
•
•
•
To understand the contribution of online channel and marketing spends and measure the ROI thereof
This analysis will help optimize marketing strategies/ digital offer policies
To identify distinguished segments within overall customer pool
This understanding would help employ customized marketing and targeting policies to maximize sales and
revenue
Approach
•
•
•
Improved channel attribution and
accuracy in reports. Better ROI
level insights driven by custom BI
solution. Improved brand advocacy
and loyalty driven by actionable BU
solutions.
Blueocean conducted series of workshops to understand business drivers, activities and also define the
actual list of hypothesis which need to be solved
Prepared and integrated several data sources such as client CRM, web metrics, promotion and product
data and publically available demographic data
Blueocean team build a multi regression based attribution model using to determine allocation for different channels to maximize
the ROI
Historic Sales
Social Media
Industry Demand
Seasonality /
Promotions
Secondary Data
Result
•
•
•
Attribution Model
Integrated data from 35+ sources and scaled up unified marketing database
A 70% accurate attribution model which improved the decision making process from a previous model at
15%
Effective online strategies resulting in improving brand advocacy by 15%
© 2015 Blueocean Market Intelligence
3
Collaborated with the client to implement big data
architectural roadmap to drive multi channel insights
Client:
Leading PC Manufacturer in the US
Study Objective
•
•
Industry:
Technology
•
One of the leading technology clients in the US had envisioned a roadmap to integrate multi channel, multi
source and different types of data (structure and unstructured) in a single platform to ensure business
stakeholders believed one single source of truth for all business KPIs and Metrics
Till the time this initiative was planned and launched – different GEOs , regions and teams were using their
own reporting and insights techniques and methodologies to keep track of changing business patterns
To drive more accountability and consistent reporting practices, client sponsors specifically required data
from different sources to reside in a single platform based on which reporting, BI and Insights practice
would be set up
Approach
Business Impact:
•
Highly scalable and flexible
solution. Seamless data
integration. World class data
architecture and governance plans.
•
•
•
Blueocean team conducted a series of workshop to understand different data sources, significance of BU
wise KPIs and metrics, closely watched numbers by GEOs, workflow management for all the varied data
sources and a common primary identifier to connect
Post month long workshop with all GEO level stakeholders, Blueocean team created data source and
process blueprint – which would be used as the reference guide for implementation and deployment of
the big data platform
Blueocean team set up a global delivery center with team of big data specialists and business analysts to
drive seamless transition between business to technical requirements
Post sign off, Blueocean team set up proprietary technology stack based on different technology arms (e.g
– APIs to drive different extraction procedures, Amazon S3 for storage, Talend solution for data
classification and ETL processing and combination of Tableau / Qlikview for reporting business KPIs
and metrics
Result
•
•
•
•
Achieved $85MM in incremental revenue
Seamless integration of 40 disparate data sources, which currently process over 250 data
transformations daily
64 standardize dashboards with 8300 instances of ad hoc analysis and insights delivered to client in first
4 months
World class data architecture and governance practice, empowering data enrichment, visualization and
multichannel analytics
© 2015 Blueocean Market Intelligence
4
Gender prediction using supervised learning algorithms
for a fortune 500 company
Client:
Leading online ecommerce
company in the US
Study Objective
•
•
The client is a pioneer in measurement of mobile subscriber behavior and wanted to understand the
gender of the subscribers based on installed Apps
The metering application installed on smart devices captures behavior of the device accurately and this
information was to be used by advertisers in order to ensure focused and targeted marketing
Industry:
eTail
Business Impact:
Accurate prediction algorithms.
Real time GUI for end users driving
higher degree of solution usage.
Approach
•
•
•
•
Initial data provided by the client was a set of user IDs along with the application names
Data cleansing and transformations were performed in order to ensure data can be fed to a supervised
learning algorithm
The data provided was highly imbalanced and skewed towards males as it was the dominant class to be
predicted; applied weighted measures to give more importance to the minority class
Support Vector Machines Learning Algorithm was applied to predict gender of the subscribers
Result
•
•
Achieved accuracy close to 80% for both classes of interest
Developed an integrated solution with a GUI to enable real time results to be obtained based on real time
data feeds to the learning algorithm
© 2015 Blueocean Market Intelligence
5
Extracted insights from various forums using advanced
text mining for a technology major
Client:
Leading PC Manufacturer in the US
Study Objective
•
•
The client was currently classifying issues manually (cost ineffective) which are posted by customers on the
company’s multitude of Tech Forums and call center agent transcripts
The client wanted to automate classification of text data to achieve accuracy close to what was being
achieved manually
Industry:
Technology
Business Impact:
Better segmentation strategy.
Automated issue classification.
Enabled targeting marketing efforts
based on data science.
Approach
•
•
•
•
Merged and cleaned forum and agent transcript data to create 130,000 rows
Data received had a breadth of topics and had 6 levels of classification for each row. Level 5 has the highest
number of distinct nodes at 4277 followed by Level 3 at 616 nodes
Team built a comprehensive dictionary based on a thorough understanding of the entire dataset
Business rules were created combined with the synonym-sets to build the model
Result
•
•
Achieved accuracy close to 60% in the 1st iteration
Provided a roadmap to enable classification of issues on general blogs and forums
© 2015 Blueocean Market Intelligence
6
Identified future attrition using a predictive scorecard
and reduced attrition by 25%
Client:
Leading bank in APAC region
Study Objective
•
•
•
Industry:
Financial Industry
Business Impact:
Insights into measures to reduce
attrition. Customer level credit
scorecarding. Propensity scores
prepared using customer
behaviour and attitudinal analyse.
A leading bank had a high attrition rate
Management wanted to know the root cause of this attrition
Need road map for retention strategies
Customers that have higher likelihood to attrite
Lesser frequency of ATM
usage in last quarter
Lower number of
products customer is
holding
Have not been active on
Bill payment or POS in
last quarter
Approach
Result
•
•
Attrition Scorecard to predict the propensity
of customers to attrite given his current
demographic & behavioural characteristics
Decreasing SA Average
Balance or dormant in
last quarter
From the model, we targeted top 3 deciles
contributing to 68% of the total attriters and
helped address attrition
Top
2nd
30
3rd
68%
4th
22
5th
6th
16
25% decrease in Attrition
7th
8th
9th
12
6
4
Bottom
Targeting
Model deciles
6%
% of
targeted clients
Top 3 deciles include over 68% of
the total attriters
$ XX K increase in Net Income over 1 year
© 2015 Blueocean Market Intelligence
7
Implemented analytical measurement system to track
ROI from B2B marketing initiatives
Client:
Leading PC Manufacturer in the US
Study Objective
•
•
Industry:
Technology
Business Impact:
Seamless data integration. Scaled
up customer Database.
Better prospecting and profiling of
B2B segment. Higher ROI from
targeted campaigns
•
This client was struggling to attribute campaign measurement success KPIs across the board and wanted
immediate help with proper reporting, attribution and analysis of ROI arriving from their B2B marketing
initiatives.
The existing CRM database used by the relationship marketers and B2B team had a good deal of nonstandardized / unstructured data formats that needed to be cleaned.
After completing the first two objectives, they also requested that the cleansed data to be used as input to
launch targeted digital campaigns
Approach
•
•
•
•
The Blueocean Market Intelligence team set up a analytics hub consisting of datawarehouse specialists,
data visualization specialist and digital scientists. A series of workshops were conducted to understand
business activities to define KPIs for success parameters.
The team integrated several data sources such as Eloqua,CRM ( Salesforce), web metrics,
( Adobe Marketing Cloud) , Sales data (SAP)and agency Data (media spends and budgets). Consulting was
provided to re-structure the digital tracking to capture required metrics and the existing BI and data
integrations process was redesigned for more accurate tracking on RM / B2Bmarketing initiatives.
The team conceptualized, designed and implemented a big data platform with proprietary technology
stack to drive on demand, real-time dashboards and insights
Result
•
•
•
•
Integrated 21+ data from multiple sources, and scaled up the existing unified marketing database
Implemented highly effective predictive models to forecast marketing spends and expected revenue at
confidence levels of 85% and R2 values close to 75%
Improved accuracy in tracking digital marketing data with close to 80% CL of attributing KPIs to campaigns
Enabled a very seamless automated mechanisms of budget allocation for each B2B campaign –
basis propensity models and visualization techniques
© 2015 Blueocean Market Intelligence
8
Implemented proprietary lead analytics engine using
advanced analytical models (1/2)
Client:
Leading Cloud Solutions Company
in the US
Industry:
Technology
Business Impact:
Better lead scoring. Improved
marketing to sales lead flows.
Reduced lead leakage powered by
more robust lead nurturing
process.
Study Objective
•
•
•
This client’s stakeholders were worried about issues related to poor lead quality and sales pipeline and
funnel indicating lack of a robust lead scoring mechanism.
Specifically, the client requested that Blueocean Market Intelligence create a propensity model at
marketing team level based on the attributes of the identified leads to segment/tear them into high and
low propensity of conversion.
The team was tasked with creating a decision sciences model to analyze at risk and dormant leads, identify
the factors leading to the non-conversion of leads at sales team level and identify the possible reasons or
factors leading to a high sales follow up rate.
Approach
•
•
•
The Blueocean Market Intelligence team deployed a team of analytics SMEs and domain experts to analyze
various metrics indicating prospect touch points on the funnel.
Post exploratory data analysis (EDA), the team created a detailed modeling approach using logistic
regression techniques to arrive at probabilistic scores highlighting prospect propensity to convert. They
deployed cluster analysis to arrive at a decision sciences mode highlighting at risk vs. dormant customers.
Prospect segmentation was then performed to identify and conduct RCA on non conversion of certain
leads. A combination of quantitative methods and qualitative analysis were carried out to lock on reasons
for a high sales follow up
Result
•
•
•
•
•
Accurate decision making with customer micro-segmentation improved the overall funnel efficacy
Targeted marketing - 25% increase in converted leads at the sales team level post implementation of
recommendations
TAT for lead conversion (aggregated) reduced by 15%
Robust propensity score carding model achieved with weekly / monthly score carding provided to key
stakeholders
Follow up rate reduced by 20% QoQ Productized an insights process using attributes which highlight risky
and dormant leads (e.g – months / days since last purchase, follow up rate, funnel conversion rate etc.)
© 2015 Blueocean Market Intelligence
9
Implemented proprietary lead analytics engine using
advanced analytical models (2/2)
Features
• Multi attribute based model
highlighting quality of leads
coming via the sales
• Comparison of lead quality as
they flow into marketing and
sales funnel, during the
qualification process
• Helped clients streamline the
entire flow of lead generation
using recurring reports and alert
mechanism
Project
Advanced modeling project to create intelligence around a lead funnel
© 2015 Blueocean Market Intelligence
10
360 Transformation
For more information:
Anees Merchant
[email protected]
Bhaskar Dey
[email protected]
United States | United Kingdom | India | United Arab Emirates
www.blueoceanmi.com
© 2015 Blueocean Market Intelligence
11