The Future of Software Development - Devin AI Innovative Approach.pdf
Seth Duncan/Context Analytics 2009 Institute for Public Relations Research Measurement Summit
1. Measuring the Impact of Earned
Online Media on Business Outcomes:
A Methodological Approach
Presentation to the IPR Measurement Summit
October, 2009
Seth Duncan, Research Manager
sduncan@context-analytics.com
2. Itinerary
1. Brief overview of how web analytics work and how they can benefit PR professionals
2. Practical steps for how communications teams can use “out-of-the-box” web analytics
3. How more advanced statistics can be used to integrate web analytics and other forms
of media measurement to help communications team target the correct online
audiences and optimize messaging strategies
4. Shortcoming of web analytics and emerging uses
5. Web Analytics: A Brief Primer
4. JavaScript code
1. Visitor types URL into executed: collects data
browser (or clicks on link) and sends to collection
server (e.g., Google
Analytics, Omniture, etc.)
2. Request sent to 3. Server sends page
website server with JavaScript code
6. Web Analytics: A Brief Primer
Direct Traffic
Search Results
Website
Email Landing
Sale
Campaigns Page
Landing
Earned Media Download
Page
Paid Search Landing
Registration
Page
Content
Network Ads •Number of Unique Visitors
•Avg. Time Per Visit
Social Media •Bounce Rate
Advertisements •Number of Page Views
•Conversions
7. Web Analytics: A Brief Primer
Earned Media
Unpaid Traffic from:
•Mainstream Media (e.g., NYTimes.com)
•Online Media (e.g., CNET)
•Blogs •Number of Unique Visitors
•Avg. Time Per Visit
•Forums
•Bounce Rate
•Twitter
•Number of Page Views
•Social Networking Sites •Conversions
•Other Corporate Websites
8. Why Are Web Analytics So Important to PR?
Uses same metrics to measure earned and paid media
Use fact rather than intuition when addressing questions such as:
•Is our corporate Twitter account driving traffic to the right web pages?
•Is Key Message A more effective at driving sales than Key Message B?
•Should we invest more resources in social or traditional media?
•What audiences should corporate communications be targeting?
Can help optimize overall communications strategy by matching the right
messages with the right audiences
Sites that Demand
Effective
refer a lot generation
messages
of traffic and sales
10. Why PR Is Not Using Web Analytics
Web Analytics
Dashboard Traditional
Media Report
Social Media Report
Raw Referral Data from
Web Analytics Solution
11. Why PR Is Not Using Web Analytics
Web Analytics
Dashboard Traditional
Media Report
Social Media Report
Raw Referral Data from
Web Analytics Solution
14. Using Web Analytics for PR
Basic Analytics Advanced Analytics
Pulled directly from solution Integrated with other data
Address basic questions Address strategic questions
Who visits your website? Which audiences respond best to
campaigns and product offerings?
Which sites drive most traffic?
Which messages are most effective
Which sites drive most sales? at driving traffic and engagement?
Should we invest more resources in How can we match the right
social or traditional media? messages to the right audiences?
15. Basic Analytics
What sites drove the most traffic and engagement?
Raw Earned Media Report
From Web Analytics
Clean
16. Basic Analytics
What types of sites drove the most traffic and referrals?
•Traditional
•Blog
Raw Earned Media Report Media Type •Forum
Categorize •Video
From Web Analytics •etc…
•General News
Site •Lifestyle
Content/Vertic •Gaming
al •etc…
Sites with
Outreach relationships/
contact
17. Basic Analytics
Just by spending a little time to categorize/segment sites (hypothetical data)…
Media Type Site Content Outreach
Conversion
Rate for Ad
Words
18. Practical Tips for Basic Analytics
1. Web metrics depend on PR goals
Generating Demand/Leads Sales
1.Unique Visitors 1.Revenue
2.Registrations 2.Orders
3.Downloads 3.Conversion Rate
4.Avg. Time on Site
5.Goal Page Visits
6.Top Exit Pages
19. Practical Tips for Basic Analytics
2. Take the time to download and clean the data
Results from search and
email campaigns could
make earned media
appear less effective
20. Practical Tips for Basic Analytics
3. Look at both totals and averages
Averages reveal “hidden gems” (hypothetical data):
Total Sales Average Sales
21. Advanced Methods
What types of stories and posts drive
action?
Assign different sites and
stories attributes that
will later be ranked to
better understand what’s
most effective
22. Tying It All Together
Web Analytics
Dashboard Traditional
Media Report
Social Media Report
Raw Referral Data from
Web Analytics Solution
23. Where to Get Data About the Site
Knowing who is visiting your site and where they are coming from
What sort of data is useful? Where do you get it?
Site Category •Human categorization
Site •Social media monitoring tool (Radian6,
Media Type Techrigy, Buzzmetrics, etc.)
Content
•Business intelligence tools
Referring Site •Search engine/panel data (Google Ad
Demographics Planner, Microsoft Adcenter, Quantcast,
etc.)
Traffic At Referring Site
•Panel-based data (e.g., comScore, Compete,
etc.)
24. How to Measure Story/Post
Content
These are the components of media coverage that drive perceptions and action
Similar to what is usually found in traditional and social media monitoring reports:
•Sentiment
•Spokespeople/Quote
ds
•Key Messages
•Competitor Mentions
•Industry Issues
•Story/Post size
•Brand Prominence
•Product Mentions
25. Data Integration and Analysis
Audience Data Site-Level Attributes: Site
type, site content,
audience, traffic, number
of posts, etc.
Unpaid Referral Data
From Web Analytics
Integrated Data
Integrated,
Causal Model
Post-Level Attributes:
Online Media
Sentiment, product
Content mentions, messages,
story length, etc.
26. Ranking Attributes by Importance
It’s like multivariate testing, only for earned media instead of advertisements
Apply multiple regression, hierarchical linear modeling, or any statistical analysis that
can provide an estimate of “effect size”
Audience/ Site Types/Categories
Site Attributes Site Demographics
Site Traffic
Regression Web
Analytics
Sentiment
Metric
Key Messages
Content/
Message Competitor Mentions
Attributes Industry Issues
Story/Post Size
Product Mentions
27. Ranking by Coverage Attributes by
Importance
Statistical Output Helps:
What Drives Visitors? What Drives Sales?
•Understand what types
of articles/posts are
effective
•What messages are
most effective?
•What types of media
outlets are effective?
•Ultimately, helps to
prioritize PR efforts
28. Tying It All Together
Creating a Causal Model Using Path Analysis or Structural Equation Modeling
31. Playbook
Match messages
Key Message:
that drive
“Product makes you
engagement with smarter”
audiences that are
likely to visit the site
Sites With High Income
Audiences
32. Playbook
Match messages “Product makes you
that drive smarter”
In
engagement with Sites With High Income
audiences that are Audiences
likely to visit the site
36. Future Directions
Integrating Advertising and Earned Media
•Are ads more effective when
they appear alongside unpaid
media?
•Is unpaid media more effective
when paired with ads?
37. Takeaway Messages
1. Web analytics allow communications teams to use the same types of measurement as
other forms of marketing
2. Web analytics can show what sites and stories/posts are driving the most traffic and
engagement on a corporate website “out of the box”
3. When integrated with other types of media measurement, web analytics can help
communications teams match the right types of messages with the right audiences to
increase engagement and revenue - this allows communications teams to use
sophisticated segmentation and targeting methods used in paid media marketing
4. Web analytics are not perfect: Imperfections with data collection, reliance on click-
throughs, and difficulty integrating different data sources are barriers to widespread
PR adoption of web analytics
38. About Context Analytics
Context Analytics helps marketing and communication
teams gain a competitive edge by identifying and Service offerings
assessing perceptions, positioning strategies, and • Global Media Research & Measurement
emerging relevant issues in mainstream and social • Social Media Analytics & Influence Mapping
media • Consulting & Research Program Management
Our research adds strategic insight to campaign • Primary Research
planning and is critical to assessing and demonstrating • Competitive Research
the value of marketing programs to executives. • Business Impact Analysis
Context Analytics is a subsidiary of Text 100.
Groups served within our clients’
Representative Clients: organizations:
• Corporate Communications/PR
• Product PR
• Pricing Strategy
• Marketing
• Branding
• Advertising
• Sponsorship Marketing
• Investor Relations
• Customer Service
• Legal/HR