Learning to Leverage Your Web Data - PowerPoint PPT Presentation

1 / 44
About This Presentation
Title:

Learning to Leverage Your Web Data

Description:

The Web Site: A Competitive Advantage ' ... How Does Portfolio Management Apply to Web Analytics? ... Operational metrics will prevent portal and Web site failures ... – PowerPoint PPT presentation

Number of Views:94
Avg rating:3.0/5.0
Slides: 45
Provided by: penton
Category:
Tags: data | learning | leverage | web

less

Transcript and Presenter's Notes

Title: Learning to Leverage Your Web Data


1
Learning to Leverage Your Web Data
Internet World Web Seminar Series March 20, 2003
2
PresentersKurt Schlegel, Program Director,
Enterprise Analytics Strategies, Meta Group Jay
Henderson, Senior Product Manager, SPSS Inc.
ModeratorPhil Ripperger, Publisher Internet
World Magazine
3
Todays Agenda
  • Introduction Web Analytics
  • The Role of Web Analytics in Todays IT
    Portfolio
  • Organizational Imperatives - The Bottom Line
  • Available Solutions/Key Considerations
  • Going Beyond The Basics
  • Fully Leveraging Customer Value - The Return
  • Web Analytics_at_Work - Case Studies
  • Q A

4
The Web Site A Competitive Advantage
  • Ever expanding Channels of Communication
  • Know Thy Customer
  • You Can Only Improve What You Can Measure

5
The Web Site A Competitive Advantage
Anytime you can increase your sales activity or
improve your customer service one percentThat is
a big number. It relates to large
dollars Internet World, September 2002
6
FLASH POLL
Are you able to tell your CEO who your most
valuable customers are? Yes No
7
FLASH POLL
Are you actively looking for ways to improve how
you measure and track online customer
data? Yes No
8
Valuing Web Sites Metrics Required
Enterprise Analytic Strategies META Group March
20, 2003
9
How Does Portfolio Management Apply to Web
Analytics?
  • High risk/reward spending such as correlating
    online and offline data (e.g., links a customer
    on Web to offline points of interaction) is done
    by few organizations, but will spread throughout
    the G2000 over 4-5 years
  • Moderate risk/reward spending such as advanced
    analytics for customer segmentation, user session
    analysis, operational metrics, and
    quality-of-experience metrics will be more
    heavily adopted 2002/03
  • Low risk/reward spending such as basic
    clickstream usage statistics is virtually
    ubiquitous throughout the G2000

IT Investment Portfolio
Venture
Invest
Discretionary Projects
Value/Timing
Growth
Risks
Grow
Discretionary Enhancements
Risk
Non- Discretionary
Non-Discretionary Costs
Run
Core
High-risk/reward investments will become
discretionary enhancements and eventually core
investments in the next 4-5 years as Web sites
expand their role as a corporate channel
10
What Metrics Are Required?
  • Simple metrics
  • of page views
  • of unique sessions
  • Session duration
  • Advanced KPIs
  • Process completion
  • Customer loyalty
  • Reduced sales cycle
  • Challenges
  • Data integration
  • Report customization
  • Defining process

Build a Business Case Through Web Metrics
KPIs
Data Integration
Process
Business Process Definition
Tech- nology
People
Privacy Policies
Data Warehouse
Key Performance Indicators
Just as dot-com companies without valid business
models collapsed, Global 2000 (G2000) Web sites
without identifiable benefits will fail
11
Aligning Measures for ROI Metrics
  • Increased Customer
  • Acquisition/Retention
  • Less elapsed time(prospect to client)
  • More products consumed (deeper penetration)
  • Higher up-sell/cross-sell rates (broader
    penetration)
  • Greater loyalty/satisfaction
  • Improved renewal rates
  • Higher conversion ratios
  • Improved Customer
  • Knowledge/Efficiency
  • Increased effectiveness/moreutilization of site
    resources
  • Greater customer reach beyond direct channels
  • Time on the site is more productive/tasks are
    streamlined

Build and Assemble Dynamic Data Elements
Back-End Infrastructure
Integrate Offline Context
Define E-Metrics
Identify Key Business Issues
Take Action to Improve Business
Front-End Analytics
Optimize Business Inv.
Although metrics may vary, categories related to
customer loyalty, reduced sales cycle,
self-service, etc. will map to page types menu
(directory type) form (input/data collection)
content (informational) application
(transactional) and collaborative (interactive
service, community)
12
Generating Sales Leads
Touch the Masses With Marketing Campaigns
Promoting Web Site
Depending on Campaign 0.5-5 (Typically) Will
Visit
It Is Up to the Sales Force to Close These Leads
Percentages Vary, But Typically gt1/3
Of the Visitors, a Higher Percentage (e.g.,
10-30) Will Become Qualified Leads
Note Actual vary widely based on environmental
variables (e.g., quality of original list)
Operational metrics will prevent portal and Web
site failures by quickly identifying underlying
problems of usage and efficacy
13
Laying Out a Clickstream Analysis Architecture
Extract, Transform, Load
Analytics
High-end solutions must provide a flexible data
model, scalable data processing, and domain
knowledge (e.g., metrics by vertical)
14
FLASH POLL
Which data collection technology do you plan to
use for Web analytics? Log file
analysis Server plug-in Network
collector Scripted page Dont know
15
Examining Business Practice Privacy Web Site
Interaction
Crossing the Privacy ChasmBuilding Trust
  • Web sites and privacy
  • Brand issues
  • Consumer trust
  • Regulatory rulings
  • Channel credibility
  • Tracking technology
  • Cookies, Web bugs, page beaconing
  • Data collection concerns
  • Role of third parties
  • Velocity of change

Automation remains critical to enabling
systematic support for privacy policy and risk
management efforts
16
Web Analytic Vendor Solutions
  • Web Analytic Outsourcing
  • Primary value data collection
  • Require tracked pages to be tagged
  • Traditional BI (Tools)
  • Build mentality
  • Integrates well with other data sources
  • Some vendors need
  • tool to capture Web data

Enterprise Web Analytic Segments
  • Clickstream Analysis Software
  • Typically log file method for traffic analysis
  • Data collection limited compared to page beacons
  • CRM and App Platforms
  • (Operational Applications)
  • Provide actionable analytics
  • Limited focus on Web analytics

Market growth has come mostly from Web analytic
service providers, typically those using Web
beacon data collection process
17
Web Analytic Architectural Choices
  • Bottom Line
  • Low end Counts basic statistics (e.g., hits,
    page views, unique visits) in a single-server
    environment with precanned reports (i.e., no
    customization)
  • Midtier Collects and analyzes data from Web
    sites that span more than one Web server reports
    are customizable two-dimensional analysis
    begins to do some customer segmentation
  • High end Includes data warehouse (OLAP) and data
    mining capabilities integrates data from
    targeted sources external to the Web server
    (e.g., content management server, application
    server, advertising server) can offer personally
    identifiable info
  • MSP Service providers collect clickstream data
    from log files or via scripted Web pages that
    extract, transform, and load usage with reports
    viewable from a Web site MSPs are the
    fastest-growing segment of the industry due to
    speed of deployments and reduced project risk

Business Impact CRM requirements to exploit
clickstream data portend growth in high-end (and
midtier) clickstream analysis
18
ROI ApproachesGoing Beyond the Basics
  • Jay Henderson
  • Senior Product Manager

19
Agenda
  • Six Key Business Problems
  • Embracing the Customer Lifecycle
  • Segmenting Users
  • Integrating Data
  • Enabling Infrastructure
  • Conclusion
  • The leader in Predictive Analytics
  • 35 year heritage as an innovator in analytic
    technologies
  • Publicly traded NASDAQ SPSS
  • 1,200 employees
  • 200,000 customers

20
Six Key Business Problems
Marketing
Commerce
Loyalty
Channels
Users
Content
21
Six Key Business Problems
Marketing
Commerce
Loyalty
Channels
Users
Content
22
R O I P E R S P E C T I V E SSix Key ROI
Business Questions
Which banner ads generate the most interest?
Where is the best site location to put an ad?
What ads drive customers to purchase? How does
this vary by customer segment? Which online
marketing programs produce the most
customers? How are my email promotions impacting
Web behavior?
What are the profiles of my best customer
segments? What products are customers buying?
What are the key drivers of purchase behavior?
What site paths are optimal for purchasing? How
are browsers different from buyers? How do online
buyers differ from offline ones? Who does
research online and then buys offline?
Marketing
Commerce
Loyalty
Channels
Users
Content
How many unique users am I getting? Where do they
come from? How long do they stay? What are the
optimal navigational paths? Which can be
improved? Where do users leave my site? Why?
What types of visits are occurring, and how
are they distinguished?
What content/areas are attractive to each
segment? What content is being overlooked? By
whom? How often should content be refreshed? To
what degree is content being personalized? How
can this be improved? Which content providers
are effective? Which are not?
23
ROI ApproachesGo Beyond the Basics
  • Embrace The Customer Lifecycle Model

24
Lifecycle Understanding
  • Identify the goals of the site
  • Map the steps to achieve that goal into the
    Customer Lifecycle

5,000
10,000
100,000
20,000
10,000
200,000
80,000
1,000,000
Time
25
Activity Analysis
  • Events indicate
  • Visit objective
  • Visit outcome
  • Simple Events - Confirmation page in
    registration process
  • Complex Events

Acquisition event
Conversion event
26
Case Study Financial Times
  • Problem
  • Both free and paid content
  • Multinational audience
  • Limited high value real-estate
  • Effectiveness of marketing campaigns
  • Benefits
  • Optimize subscription conversion
  • Informed decisions selecting content placement
  • Streamlined registration process

27
ROI ApproachesGo Beyond the Basics
  • Segmentation

28
Business logic the 90/10 rule
Percentage of Value
Percentage of visitors
29
Customer SegmentationFirst Steps
  • Identify preliminary list of attributes that
    relate to your problem at hand
  • For each attribute, establish Segments and
    determine the number of people in each
  • Look at the series of attributes and analyze the
    data to determine first-cut actionable steps

30
Customer SegmentationLong-Term
  • Long-term, move to multiattribute segmentation to
    develop sophisticated segmentation models.
  • More holistic, complex analysis of customers
  • Highlights interrelationships more effectively
  • Demonstrates striking insights
  • Sophisticated segmentation models frequently
    require assistance with data mining.
  • Highest benefit is reached from integration of
    multiple data sources.

31
Case Study 1-800-flowers
  • Problem
  • Complex, high traffic Web site
  • Sophisticated analytical and database marketing
    requirements
  • Multiple online and offline customer touch-points
  • Benefits
  • Understand, profile and segment e-customer base
  • Better target high value customers
  • UUID cookie strategy

32
ROI ApproachesGo Beyond the Basics
  • Data Integration

33
Data Integration is an ROI Accelerator
34
Case Study Movielink
  • Problem
  • Contractual obligation to report ROI to partners
  • Analysis paralysis / Information overload
  • Disparate data sources
  • Benefits
  • Integration with Financial Data
  • Link online and offline metrics
  • Partner confidence in results
  • Focused set high value ROI reports
  • Single view of customer

35
Major Phases of E-Metrics Adoption
Operational Systems
Business Unit D
Business Unit C
Business Unit B
Business Unit A
36
Enabling Infrastructure
37
N E T G E N E S I S Platform Highlights
38
N E T G E N E S I S Privacy Enablement
  • Online Privacy a growing consumer concern
  • Web sites use cookies to track users
  • Privacy policies hard to understand
  • Users not aware of 3rd party Web sites
  • NetGen Includes
  • Opt-out and anonymous capability
  • Allows clients to deploy P3P opt-out compact
    policy

39
Predictive Web Analytics
40
Predictive Web Analytics
41
Prediction Provides ROI Beyond Reports
  • Example
  • Without prediction
  • 100 of customers targeted with offer
  • 100 of responders found
  • 50 Targeted with offer
  • 50 of responders found
  • With prediction,
  • best 50 of customers targeted
  • 88 of responders found
  • Each customer can get a score between 0 and 1 for
    churn, purchase, response propensity

42
SPSS and NetGenesis Strengths
  • Innovative Web metrics experience
  • Roadmap to ROI, advanced analytics and predictive
    Web analytics
  • Total solutions approach analytic applications
    plus E-Metrics consulting services
  • Domain expertise for addressing Web data
    challenges
  • Vertical industry knowledge and experience
  • Highest scale and most reliable solution for
    complex enterprise Web environments
  • Open, extensible and cross-platform Web-based
    architecture

43
QA Web Seminar Attendee Survey
44
Thank You For Joining Us
Contact Information jhenderson_at_spss.com Kurt.Sc
hlegel_at_MetaGroup.com Phil Ripperger,
pripperger_at_iw.com
Write a Comment
User Comments (0)
About PowerShow.com