Web Analytics Discussion

1
Web Analytics Discussion

• OPA Marketing and Research Day
• April 18, 2006

2
Discussion Outline

• Purpose
• How do we get the most value out of our
  investment in web analytics?
  
• Survey Results
• Issues / Challenges / Concerns
• ROI / Level of Investment
• Benefits
• Success Factors
• Best Practices
• Discussion

3
Survey Results

• 24 survey participants
• ABC AllBusiness.com
• Belo BusinessWeek
• CBS CNet
• CondeNet Cox
• DisneyInteractive Dow Jones
• Edmunds ESPN
• IBS iVillage
• KnightRidder NYTimes.com
• Reuters Scripps
• Sporting News Tribune
• Turner USAToday
• Weather.com WashingtonPost Newsweek Interactive
  
  

4
Half of Us Use Omniture
Other CoreMetrics, SurfAid,
FireClick, Revenue Science
42 use 2 tools (Tacoda, Site Census, NetTracker,
Personify, Urchin)
5
Most (80) Are Satisfied with Their Vendor
Interestingly, Omniture did not receive any
dissatisfied ratings
6
Most of Us Expect to Renew with Current Vendor
7
Staffing Levels Vary Widely
Q. Please indicate the number of dedicated web
analytics resources in each applicable dept.
8
Resources are Spread Across Departments
Q. Please indicate where web analytics
responsibilities reside in your organization.
(multiple responses allowed)
9
Issues / Challenges - Staffing

• A new discipline web analytics
• Find and retain the right people
• Create career paths
• Web Analytics Association useful
• Skills are very technical
• Need dedicated resources from IT
• Inadequate resources to meet demand
• Analysts, managers stuck doing implementation

10
Issues / Challenges - Data

• Trust and confidence in the data
• Will you get the same answers with a different
  tool
  
• Trending with other sources
• Looking-forward instead of rear-view mirror
• Parallel flight-testing, A/B testing
• Amount of data can overwhelm
• Roles and responsibilities
• Whose responsibility is it when data is
  questionable
  
• Semantics
• Consistent ways of referring to things

11
Issues / Challenges - Implementation

• Unrealistic expectations
• Will never be done or just run itself
• Needs constantly grow and evolve
• Caught up in implementation
• Not enough time for analysis
• Research support may stay consistent, but IT
  support may not
  
• Final word is for Research to validate

12
Issues / Challenges - Vendor

• Post-sale customer service
• Level of support from acct mgr, engineering
• Switching costs
• Lose years of legacy data
• Implementation costs in time

13
Issues / Challenges Information-Sharing

• Getting right info to the right people
• Right format (dashboards, emails, paper)
• Self-service
• Customizing published reports
• Ensuring the information gets used
• With Editorial, avoid insulting journalistic
  integrity
  
• With Sales, avoid sharing too much information
• Tying data to ROI

14
Issues / Challenges - Tracking

• Tracking video streams
• Moving away from pageviews to other metrics
• Segmentation
• Cumbersome
• Time spent
• Accuracy
• ClickMap
• Reliability
• Doesnt work with flash

15
Concerns

• Solutions are cookie-based
• Cookie deletion a real problem
• Alternate solutions needed from measurement
  companies
  
• Internal numbers do not match with panel data
• Panel data is about ½ of internal
• Are PVs and UVs the metrics that matter most?
• Or is it loyal users or engagement
• Looking ahead to future revenue models

16
ROI / Level of Investment

• Determining appropriate level of investment
• Prioritizing vs other research needs
• Investment in the software/service contract
• Investment in staff to use it
• Using 5 of the capacity?

17
Benefits

• Decisions more strategic
• Based on data, not just gut
• Marketing campaign management
• Product development
• Resource allocation
• Redesigns / Usability
• Search campaigns
• Track distribution partners
• Sales for client reporting
• Drill-down capability not available in NNR

18
Success Factors

• Executivelevel support for analytics
• Make full commitment
• Both for initial deployment and for widespread
  use
  
• Goals tied to metrics
• Make it part of fabric of whole company
• Hire and staff appropriately
• Headcount
• Continuity of IT resources as well
• Focus on KPIs (Key Performance Indicators)
• Avoid getting buried in too much data
• Start by setting business goals build KPIs
  around them
  
• KPIs should be something people can identify
  with
  

19
Success Factors (continued)

• Reporting structure
• Report into Operations vs Marketing vs. Sales
• Widespread usage
• Dont silo analytics only in Research
• But have central responsibility with 1-2 people
• Otherwise danger of conflicting evars,
  under-the-radar activity
  
• Dont use sampling

20
Best Practices

• Demystify
• Training across organization (ongoing)
• Brown-bag lunches for info-sharing
• Issue regular reports from analysts
• Transparency of prioritization process
• Input form for submitting requests
• Can view each others requests
• Triage based on business value (2 times/week)
• Set expectations internally
• Is a toolkit, not a magic wand
• Engage business to help prioritize

21
Discussion

• What other issues do we need to discuss?

22
Web Analytics Discussion

• OPA Marketing and Research Day
• April 18, 2006