14. Experimentation as a continuous process
• Generate hypotheses
• Scope requirements
• Prioritize experiments
Hypothesis
generation
• Execute experiments or
personalized campaigns
• Real-time performance
monitoring
Execution
• Build new features or experiences
• Determine target audiences
• Develop creative assets
Design and
development
• Reporting and analysis
• Determine next phase
of experiments based
on results
Analysis
46. opticon2017
Results • 1.5% increase in engagement
• 113% increase in conversions to the
products page
• 117% increase in conversions on the “Test
it Out” CTA to start the account creation
process
57. What is preventing us from running
more experiments?
• Many experiments require developer
involvement
• Experiments go untested due to
resource constraints
• Experimentation is not part of a user's
content creation workflow
58.
59. Share and update content across
the entire customer journey
Knowledge Base
Academy Developers
Marketing
Community
60. Module - Hero Split
Module - Media Showcase
Module - Media Showcase
Module - Media Showcase
61. Module - Icon Cards
Module - Center Content
Module - Media Showcase
62.
63. Split Page Testing, using Fullstack
• Launch experiments from the
CMS
• No developer needed
• User is not redirected to a new
page
• Optimizely Web is great for small
changes
• Optimizely Full stack enables us
to experiment with completely
different layouts
67. opticon2017
Recap
1. Get everyone involved
2. Use a partner to help you along the way
3. Define clear objectives for your site
4. Segment your visitors
5. More Experiments = More Wins