Making Online Shopping Smarter with Advanced Analytics

• Cognizant 20-20 Insights
Making Online Shopping Smarter
with Advanced Analytics
Merging offline and online data can help retailers to customize targeting,
resulting in incremental increases in e-commerce revenues.
Executive Summary
Online shopping has emerged as one of the most
popular Internet activities, providing a variety of
products for consumers and a multiplicity of sales
challenges for e-commerce players. Research
suggests that online sales has grown 16.1% yearover-year from March 2010 through March 2011. 1
Both brick and mortar and pure play online
retailers are competing for the attention of
the online consumer. A real opportunity for
companies lies in applying advanced digital
analytic techniques that integrate offline and
online data sets to optimize on-site store inventories. One example from retail is how Best Buy
is leveraging data to become more customercentric. When Best Buy determined that 7% of
its customers were responsible for 43% of its
sales, the company segmented its customers
into several archetypes and redesigned stores
to address the buying habits of these particular
customer segments, thereby enhancing the
in-store experience and increasing same-store
sales by 8.4%.2
While optimization and other analytical
techniques like A/B/multivariate testing, visitor
engagement, behavioral targeting and audience
segmentation can point toward a high likelihood
cognizant 20-20 insights | june 2012
of a prospect’s willingness to buy, transactional
data points such as offline sales are key indicators
of actual sales. This paper provides insights on
how merging offline and online data can support
customized targeting, resulting in incremental
increases in e-commerce revenues.
Constraints and Opportunities
In most companies, unfortunately, online and
offline sales are processed in technical silos.
Offline shopping data is very important because
many people still want to touch and see products
first-hand before they buy. Conventional wisdom
suggests that about 70% of all consumers will
make an offline purchase as a result of online
marketing. Thus, offline data is extremely
beneficial, and mining it along with clickstream
data provides a valuable resource for improving
customer satisfaction, product development,
sales forecasting, merchandising and visibility
into the profit margin for goods and services.
The ability to identify and reach consumers at the
most crucial point in their buying decision-making
process, as reinforced by Google’s “Zero Moment
of Truth”3 study, demonstrated that in-store or
online transactions are heavily influenced by
insights gained in the moment before a purchase
decision is made. Thus, steering consumers
Applying Advanced Analytics
Data
ETL
Profiling
Association
Rules
Logistic
Regression
Variables
Creation
Statistical
Control Process
Weighted
Average
Segmentation
Statistical
Control Process
Classification
Time Series
Linear
Regression
Advanced Analytics
Output
Figure 1
toward a particular product or service and then
capturing them at the point of sale is a very
powerful solution.
An e-commerce player
that can capture the
consumer’s attention
and activity further
up the funnel and
combine it with
advanced analytics
techniques can begin
to assume the mantle
of an analytics leader.
This is why Google’s search
marketing is so appealing
to marketers; consumers’
desires are instantly telegraphed by their actions.
Creating a holistic customer
picture is not a new idea but
very few companies have
achieved it. Constraints
have included database
and hardware technology
that could not effectively
support the solution; data
management issues; Web
analytic systems that are not focused on the availability of the underlying data; not getting the right
people to focus on the data; the division of online
marketing and traditional marketing skill sets; etc.
Leading the Charge
For e-commerce companies, the challenge is
that customers are often relatively far into the
purchase funnel when they arrive at a particular
site. An e-commerce player that can capture the
consumer’s attention and activity further up the
funnel and combine it with advanced analytics
techniques can begin to assume the mantle of an
analytics leader.
cognizant 20-20 insights
Joining geographic, creative, time of day (e.g.,
morning, afternoon, evening, etc.) data and
mapping these attributes against time and
location-based sales data can highlight hidden
interactions between online and offline sales
activity. This means e-commerce players can
develop more effective multichannel strategies.
The end result is that real additional sales can
be increased. Of course, e-commerce companies
have an option to use advanced Web analytics to
create a specific segment for visitors who arrived
online via offline campaigns. Web analytic4 tools
provide similar solutions, but they are not by any
means a complete set of offerings nor do they
create a single view of your customer. Organizations need to pull all the data attributes, offline
and online, into a single database, which would be
further refined by advanced analytics techniques,
and use the unified data for precision targeting as
depicted in Figure 1.
Figure 2 illustrates that by combining Website
data (i.e., clickstream information, etc.) with
loyalty card insights, sales data and third-party
information, e-commerce players can gain critical
understanding into customer behavior. To reach
the forefront of the data-driven digital revolution,
e-commerce players need to acquire key tools,
analytic prowess and necessary skills. Not only
do e-commerce players need to invest in the right
advanced analytics tools, but they must also gain
access to mathematicians and statisticians to
create models and interpret findings.
2
Blending Multiple Sources of Sales Data for Optimal Positioning
1
3
2
Online + Offline
Customer Behavior
We Reach — Right Customer at Right
Time with Right Products
Completes the Jigsaw Puzzle
Enable Complete View
of Purchase Cycle
Integrate Promotion, Web Visit, Sales
Transaction, Customer Data
Leverage Digital Analytics to
Gain Customer Insights
ONLINE
Quality
k
eedbac
F
g
Demo
Score
Model
Social
Media
y
t
egistr
t Accoun
Gift R
Paymen
tion
Promo
Sales
Predict the
“Right Time to Sell”
Understand Customer
Buying Choices
Privacy
ss
Addre
g
Catalo
Item
Segmentation
Multi
Channel
Optimization
Predictive
Modeling
Market Basket
Analytics
Factor
Analysis
Probability
to Buy
Store Locator
Analytics
ROI
Probability
to Drop Off
on
Locati
isit
bV
We
Assessing Impact of
Online Over Offline
OFFLINE
Figure 2
Organizations need to apply more complex data
calculations to generate a more accurate picture
of customer behavior and transactional activity.
Measuring offline marketing campaigns statistics
is not as easy as many people think. There are
numerous different tools that can help organizations improve data accuracy; however, many are
third-party solutions that are not integrated into
a Web analytics solution.
Advanced Analytics Framework
in Action: E-Commerce Gold
While e-commerce sales will continue to grow
over the next four years, eMarketer5 estimates
that the rate of growth will decline steadily. This
makes it even more vital for e-commerce players
to be strategic about how they use clickstream or
Web log data to reach consumers.
The Art and Science of Connecting Offline and Online Data
Data Preparation
Segmentation
Campaign Data
Profiling
1.0
Web Visit Data
0.8
Demographic Data
Raw Analysis
Data
0.6
0.4
0.2
0.0
Sales Transaction Data
Finalize models by taking
inputs from business teams
Lift Chart*
Rank Ordering
Predictive
Modeling
Data
High Score
Model
Medium Score
Low Score
Optimize
Targeting
Figure 3
cognizant 20-20 insights
3
Business
Input
Output
Benefits of Merging Online and Offline Data: Scenario One
Situation: Client has huge online and offline data which needed integration for actionable analysis.
Solution : Profile customers based on their online purchasing/browsing behavior with offline transactions.
Visit
freq.
Online Data
Buyer
Visitor Profiling
Web Visit
Non
Buyer
Web
Registration
Email
Campaign
Regular
Seasonal
Regular
Seasonal
Regular
Seasonal
Regular
Seasonal
Low
High
Low
% Customer
Large Frequent Buyer
Channel Preference
Purchase
Behavior
Online
Transactions
Banner Ads
Sales
Transaction
Demography
Visitor and
Session
Recognization
Store Location
Address
11.7%
Small Buyer
21.0%
Modern
Utilitarian
Organic
Classic
Brand A
Brand B
Brand C
All Brands
17%
16%
2.7%
0.8%
0.6%
3.5%
6.5%
2.0%
0.5%
2. Channel preference
helped client in
reaching out to
customers with their
preferred channel.
Both
2.0%
5.4%
3.0%
0.4%
0.8%
1.5%
1.9%
% of
Customers
% of Sales
40.82%
17.55%
10.76%
8.37%
43.75%
11.10%
5.19%
11.04%
3. Taste of a buyer
helped client in
reaching them with
targeted offers.
% of Identified visitors
% Identified Sessions
Dec-10
Jan-11
16%
15%
22%
19%
Concept
Retail
5.0%
30.7%
Small Frequent Buyer
1. Profiling helped client
in recognizing the
potential buyers and
up sell seasonal buyers.
44
26
5
4
38
26
10
8
Channel Preference
eCom
Large Buyer
Frequent Buyer
Visits per
Visitor
15.98%
1.04%
0.10%
4.59%
1.71%
0.11%
1.45%
1.03%
Medium Buyer
Design Preference
Segment
Purchasing
Preference
Offine Data
% of
Visitors
Seasonality
High
Dec-10
13%
14%
4. Visitor and session
recognization was
improved by 30%
to 50%.
Jan-11
8%
6%
11%
8%
7.8%
9%
6%
6.3%
Figure 4
Figure 3 illustrates a typical process/methodology
followed by a digital analytics center of excellence
(CoE) to merge offline and online data in order to
build predictive models that optimize customer
targeting.
In the near future, organizations will likely deploy
more sophisticated and integrated Web/digital
analytics tools to more easily combine online
campaign data with offline insights. This will
include:
• Getting the Web log or clickstream data from
We have applied this methodology at several of
our clients resulting in the scenarios presented in
Figures 4 and 5.
Web analytics tools.
• Merging it with point-of-sale data.
Benefits of Studying Online Customer Behavior: Scenario Two
Situation: : Unstructured huge clickstream data with not much knowledge of what to do with it.
Solution : Web path analysis revealed browsersʼ intent and needs which helped to customize offerings.
Browser
Web Visit
Log
Web
14
40
30
42
33
10
Web Path
Online Purchasing
After E-mail
Click-through
Online
Transactions
Web Page
Email
View
Campaign
Banner Ads
Web
Registration
Offine Data
Browser Store
Distance
Analysis
Sales
Transaction
Item View
Demography
search
Address
50
20
Registration
Store
Location
Onsite
Thousands
Express Checkout
Analysis
Online
Web Data
18
e
Onsite Search
Effectiveness
0
Category 1
1. Checkout analysis helped client
in estimating the return on new
development on products Web site.
Store Area Distance from browsers
% Customers with % Order Value
in 50 Miles
with in 50 Miles
Grove
96%
88%
Westchester
96%
82%
Whitman
95%
79%
3. Browser behavior around a
store helped client in providing
customized offer on specific
stores.
Figure 5
cognizant 20-20 insights
4
100%
80%
60%
71%
62%
65%
61%
48%
70%
40%
20%
0%
Retail
Category 2
CHECKOUT SIGN IN
EXPRESS CHECKOUT SIGN IN
Location
Name
% of Unique Customers
Purchased with web Visit
30% people are using Express Checkout
60
Catalog
eCommerce
Transaction Channel
30 Days prior
90 Days prior
2. Customer behavior of browsing
before buying helped client in
focusing more on information
on their website about products.
Activities
# of Visits
Total Product Search
Exit after Search
NULL Search
Product Search
Using Search Again
Top or Left
Navigations
Homepage
700,000
150,000
7,000
260,000
160,000
% Activities
after onsite
search
22.5%
1.0%
39.3%
24.1%
50,000
7.1%
20,000
2.4%
4. Onsite search effectiveness helped
client in customizing product
information for better search results.
Generating Collective Intelligence
Conclusion
Marketplace (Online + Offline)
Filtering
Group of Experts
Leaders, Scientists
and Experience Facilitators
Validation and
Application
Synthes
is
Sales, Web, Demography, Store Location, E-mail,
Banner Mobile, Social Media, Data, etc.
Analytical Tools
Omniture, Webtrends, SAS, SQL,
Coremetrics, etc.
have created a unique solution framework (see
Figure 6).
Highly Specialized
Information Silos
Collective Intelligence
New Technologies, New Knowledge,
New Philosophy
Insight Builder Tool
Figure 6
• Studying online and offline customer behavior.
• Refining that data with the use of advanced
analytic techniques, set methodology and
algorithms, applied by the right domain
experts.
• Then, validating/refining the data for pinpoint
targeting.
The end result: Enabling companies to reach the
right customer with the right product offerings
at the right time and place. With this in mind, we
The solution framework shown in Figure 6 can be
customized to meet the unique requirements of
each company. Because it is technology agnostic,
it can seamlessly integrate with any ERP system,
the Web, social media, DW/BI, CRM and POS
systems. Also, it takes data in whatever form is
given and then it works toward synthesizing each
of the data sets in a format that can be further
sliced and diced using our proprietary algorithms
and models. It thereby brings offline and online
data together to enable business users to make
insightful decisions to move their businesses
forward.
For example, an organization selling laptops or
mobile devices that captures the correct 1% of
its prospects can generate an additional 3% in
increased revenues. It can do so by applying a
tried and true process/methodology (as illustrated in Figure 3), followed by a digital analytics
center of excellence to merge offline and online
data and build predictive models to optimize
customer targeting.
Thus, identifying the most valuable customers
using online and offline data allows companies to
more accurately identify those who are transaction minded and can help boost their revenues.
A data-driven optimized strategy is the key, and
that is possible only with offline and online data
integration.
Footnotes
1
http://www.ipaydna.biz/Online-sales-witness-16.1-percent-growth-year-over-year-n-43.htm
2
http://www.fico.com/en/FIResourcesLibrary/Best_Buy_Success_2271CS_EN.pdf
3
http://www.thinkwithgoogle.com/insights/library/studies/the-zero-moment-of-truth-macro-study/
4
http://en.wikipedia.org/wiki/Web_analytics
5
http://eMarketer
About the Author
Ashish Saxena is the Leader of the Digital Analytics CoE within Cognizant’s Enterprise Analytics
Practice. He has over 15 years of experience in e-commerce, multichannel retail, digital and advanced
analytics space. He also holds several patents in the e-commerce/mobile space. Ashish can be reached at
Ashish.Saxena3@cognizant.com.
cognizant 20-20 insights
5
About Cognizant’s Enterprise Analytics Practice
Cognizant’s Enterprise Analytics Practice (EAP) combines business consulting, in-depth domain
expertise, predictive analytics and technology services to help clients gain actionable and measurable
insights and make smarter decisions that future-proof their businesses. The Practice offers comprehensive solutions and services in the areas of sales operations and management, product management
and market research. EAP’s expertise spans sales force and marketing effectiveness, incentives
management, forecasting, segmentation, multichannel marketing and promotion, alignment, managed
markets and digital analytics. With its highly experienced group of consultants, statisticians and industry
specialists, EAP prepares companies for the future of analytics through its innovative “Plan, Build and
Operate” model and a mature “Global Partnership” model. The result: solutions that are delivered in a
flexible, responsive and cost-effective manner www.cognizant.com/enterpriseanalytics.
About Cognizant
Cognizant (NASDAQ: CTSH) is a leading provider of information technology, consulting, and business process outsourcing services, dedicated to helping the world’s leading companies build stronger businesses. Headquartered in
Teaneck, New Jersey (U.S.), Cognizant combines a passion for client satisfaction, technology innovation, deep industry
and business process expertise, and a global, collaborative workforce that embodies the future of work. With over 50
delivery centers worldwide and approximately 140,500 employees as of March 31, 2012, Cognizant is a member of the
NASDAQ-100, the S&P 500, the Forbes Global 2000, and the Fortune 500 and is ranked among the top performing
and fastest growing companies in the world. Visit us online at www.cognizant.com or follow us on Twitter: Cognizant.
World Headquarters
European Headquarters
India Operations Headquarters
500 Frank W. Burr Blvd.
Teaneck, NJ 07666 USA
Phone: +1 201 801 0233
Fax: +1 201 801 0243
Toll Free: +1 888 937 3277
Email: inquiry@cognizant.com
1 Kingdom Street
Paddington Central
London W2 6BD
Phone: +44 (0) 20 7297 7600
Fax: +44 (0) 20 7121 0102
Email: infouk@cognizant.com
#5/535, Old Mahabalipuram Road
Okkiyam Pettai, Thoraipakkam
Chennai, 600 096 India
Phone: +91 (0) 44 4209 6000
Fax: +91 (0) 44 4209 6060
Email: inquiryindia@cognizant.com
­­© Copyright 2012, Cognizant. All rights reserved. No part of this document may be reproduced, stored in a retrieval system, transmitted in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise, without the express written permission from Cognizant. The information contained herein is
subject to change without notice. All other trademarks mentioned herein are the property of their respective owners.