Generating  High-­Impact  Experiments
Hazjier  Pourkhalkhali
Senior  Strategy  Consultant
@Hazjier
Optimizely is the world’s largest digital laboratory
We researched the world’s top companies
Interviewed experts
§ Spoke with thought
leaders in the industry
§ Met with top customers
Built maturity model
§ Developed
comprehensive
maturity model
§ Assessed 100
customers
§ Surveyed 50+
customers
Held 100+ Business Reviews
§ Met with executives
at over 100
Optimizely clients
Analyzed Top Revenue Experiments
§ Analyzed the 100 best
experiments run on
Revenue Per Visitor
§ Asked customers to
describe test and process
User
acquisition:
1.5m
User
retention
33%
Add-to-Cart:
4.00%
Cart
Checkout:
50.0%
Average
quantity:
2.00
Average per
unit price:
€25.00
Average
Order Value:
€50.00
Conversion
Rate:
2.00%
Revenue Per
Visitor:
€1.00
Visitors:
2.0m
Revenue:
€2.0m
Understand the revenue drivers
Goals
Strategies
Tactics
Cart
Checkout
Rate
Wording /
messaging
Location /
size
Design /
color /
iconograp
hy
Establish a
visual
hierarchy
Remove
non-
critical
content
Limit
number of
choices
Select/
target the
right USP’s
Refine the
messaging
Crystallize
visually
Clear and
consistent
UI
Structure,
naming,
and order
of sections
Describe
visually
Test ideal
product
visuals
Use the
right
review
systems
Descriptions
Reviews
Images
Emphasize
the Primary
Call to Action
Minimize
Distractions
Communicate
Unique
Selling Points
Design
Intuitive
Navigation
Market
Products
Effectively
Build a Sense
of Urgency
Show
quantity
left
Push
temporary
offers
Promote
buying
before set
times
Testing requires freedom
LOW  FREEDOM HIGH  FREEDOM
“Every change we make to our
website has to be approved
by our senior staff”
-Financial Services company
Testing requires freedom
LOW  FREEDOM HIGH  FREEDOM
“Our Brand Styleguide is key.
We cannot make changes that
do not fit our design book”
-Financial Services company
Testing requires freedom
LOW  FREEDOM HIGH  FREEDOM
“We would change our
company logo if it
converted better”
-Travel Company
NO  BEARD CAVEMAN
.30% .29% .28% .32% .25% .74%
.11% .28% .24% .32% .09% .59%
.74%
.59%
Minimize Distractions:
Remove Non-Critical Content
+14%
Revenue Per
Visitor
BASED ON KEYWORDCURATED
+24%
Click to Sale
+30%
Add-to-Cart
+210%
Conversions
§ The user is reminded
repeatedly of price
§ The user must remember the
product benefits
§ The user must scroll down to
the CTA when their basket is
3 or more
Problem Solution Result
§ Reduce size of price
§ Highlight discounts
§ Add product visuals
§ Duplicate CTA at top of
page
§ Increase Conversion Rate
§ Increase Conversion Rate
§ Increase CTR to checkout
Build powerful hypotheses
§ Analyze one section of your
application / website
§ Document 10 different
problems for your users
§ Group them by common root
causes
§ Prioritize by urgency
Problem Solution Result
Optimizely Hypothesis Workshop
§ Analyze one section of your
application / website
§ Document 10 different
problems for your users
§ Group them by common root
causes
§ Prioritize by urgency
Problem Solution Result
Optimizely Hypothesis Workshop
§ Select 1 concrete
problem
§ Design 10 solutions per
theme
§ Choose 4 preferred
solutions
The problemwith “A/B Testing”
2
A B
4
A B C D
Most people test only 2 variations
0
10,000
20,000
30,000
40,000
50,000
60,000
2 3 4 5 6 7 >8
Tests
65% 17% 17%
# of Variations
Most people test only 2 variations
0
10,000
20,000
30,000
40,000
50,000
60,000
2 3 4 5 6 7 >8
Tests
# of Variations
Test 5 variationsand nearly double your win rate!
Win Rate
# of Variations
14% 18% 20% 27% 20% 22% 23%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2 3 4 5 6 7 >8
+90%
Running more variationswill significantlyimprove your testing!
§ Analyze one section of your
application / website
§ Document 10 different
problems for your users
§ Group them by common root
causes
§ Prioritize by urgency
Problem Solution Result
Optimizely Hypothesis Workshop
§ Select 1 concrete
problem
§ Design 10 solutions per
theme
§ Choose 4 preferred
solutions
§ Primary Goals:
What do you seek to impact
immediately
§ Secondary Goals:
What would you like to
impact, but do not have to
§ Monitoring Goals:
What are metrics you want
to track
Optimizely funnel analysis
Homepage
Flight
Results
Book
Flight
Passenger
Details
Flight
Options
Payment
Page
Step
45%
30%
65% 65%
90%
70%
40%
40%
15%
20%
5%
25%
15%
30%
20%
15%
5% 5%
Exit Rate
Divergence
Completion
Optimizely funnel analysis
Homepage
Flight
Results
Book
Flight
Passenger
Details
Flight
Options
Payment
Page
Step
Primary /
Secondary
Goals
§ Searches
Submitted
§ Category tabs
§ Deals
§ Frequent Flyer
§ Flight Selection
§ Filter Usage
§ Refine Search
§ Sort Order
§ Continue
button
§ Edit search
§ Upsells
§ Field
completion
rate
§ Continue
button
§ Continue
button
§ Upsells
§ Seat
selection
§ Conversion
rate
§ Form field
completion
§ Payment
Type
Monitoring
Goals
§ Checkboxes
§ Flight tracker
§ Home page
clicks
§ Menu
interactions
§ Site search
§ Home page
clicks
§ Menu
interactions
§ Site search
§ Field error
rate
§ Home page
clicks
§ Menu
interactions
§ Site search
§ Shopping
cart
pageviews
§ Home page
clicks
§ Return to
previous page
§ Edit address,
information
§ Error rate
§ Analyze one section of your
application / website
§ Document 10 different
problems for your users
§ Group them by common root
causes
§ Prioritize by urgency
Problem Solution Result
Optimizely Hypothesis Workshop
§ Select 1 concrete
problem
§ Design 10 solutions per
theme
§ Choose 4 preferred
solutions
§ Primary Goals:
What do you seek to impact
immediately
§ Secondary Goals:
What would you like to
impact, but do not have to
§ Monitoring Goals:
What are metrics you want
to track
Optimization Summer Games - Test Ideation

Optimization Summer Games - Test Ideation

  • 2.
    Generating  High-­Impact  Experiments Hazjier Pourkhalkhali Senior  Strategy  Consultant @Hazjier
  • 3.
    Optimizely is theworld’s largest digital laboratory
  • 5.
    We researched theworld’s top companies Interviewed experts § Spoke with thought leaders in the industry § Met with top customers Built maturity model § Developed comprehensive maturity model § Assessed 100 customers § Surveyed 50+ customers Held 100+ Business Reviews § Met with executives at over 100 Optimizely clients Analyzed Top Revenue Experiments § Analyzed the 100 best experiments run on Revenue Per Visitor § Asked customers to describe test and process
  • 6.
    User acquisition: 1.5m User retention 33% Add-to-Cart: 4.00% Cart Checkout: 50.0% Average quantity: 2.00 Average per unit price: €25.00 Average OrderValue: €50.00 Conversion Rate: 2.00% Revenue Per Visitor: €1.00 Visitors: 2.0m Revenue: €2.0m Understand the revenue drivers
  • 7.
    Goals Strategies Tactics Cart Checkout Rate Wording / messaging Location / size Design/ color / iconograp hy Establish a visual hierarchy Remove non- critical content Limit number of choices Select/ target the right USP’s Refine the messaging Crystallize visually Clear and consistent UI Structure, naming, and order of sections Describe visually Test ideal product visuals Use the right review systems Descriptions Reviews Images Emphasize the Primary Call to Action Minimize Distractions Communicate Unique Selling Points Design Intuitive Navigation Market Products Effectively Build a Sense of Urgency Show quantity left Push temporary offers Promote buying before set times
  • 8.
    Testing requires freedom LOW FREEDOM HIGH  FREEDOM “Every change we make to our website has to be approved by our senior staff” -Financial Services company
  • 9.
    Testing requires freedom LOW FREEDOM HIGH  FREEDOM “Our Brand Styleguide is key. We cannot make changes that do not fit our design book” -Financial Services company
  • 10.
    Testing requires freedom LOW FREEDOM HIGH  FREEDOM “We would change our company logo if it converted better” -Travel Company
  • 12.
    NO  BEARD CAVEMAN .30%.29% .28% .32% .25% .74% .11% .28% .24% .32% .09% .59% .74% .59%
  • 13.
    Minimize Distractions: Remove Non-CriticalContent +14% Revenue Per Visitor
  • 14.
    BASED ON KEYWORDCURATED +24% Clickto Sale +30% Add-to-Cart +210% Conversions
  • 17.
    § The useris reminded repeatedly of price § The user must remember the product benefits § The user must scroll down to the CTA when their basket is 3 or more Problem Solution Result § Reduce size of price § Highlight discounts § Add product visuals § Duplicate CTA at top of page § Increase Conversion Rate § Increase Conversion Rate § Increase CTR to checkout Build powerful hypotheses
  • 18.
    § Analyze onesection of your application / website § Document 10 different problems for your users § Group them by common root causes § Prioritize by urgency Problem Solution Result Optimizely Hypothesis Workshop
  • 19.
    § Analyze onesection of your application / website § Document 10 different problems for your users § Group them by common root causes § Prioritize by urgency Problem Solution Result Optimizely Hypothesis Workshop § Select 1 concrete problem § Design 10 solutions per theme § Choose 4 preferred solutions
  • 20.
    The problemwith “A/BTesting” 2 A B 4 A B C D
  • 21.
    Most people testonly 2 variations 0 10,000 20,000 30,000 40,000 50,000 60,000 2 3 4 5 6 7 >8 Tests 65% 17% 17% # of Variations
  • 22.
    Most people testonly 2 variations 0 10,000 20,000 30,000 40,000 50,000 60,000 2 3 4 5 6 7 >8 Tests # of Variations
  • 23.
    Test 5 variationsandnearly double your win rate! Win Rate # of Variations 14% 18% 20% 27% 20% 22% 23% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 2 3 4 5 6 7 >8 +90%
  • 24.
    Running more variationswillsignificantlyimprove your testing!
  • 25.
    § Analyze onesection of your application / website § Document 10 different problems for your users § Group them by common root causes § Prioritize by urgency Problem Solution Result Optimizely Hypothesis Workshop § Select 1 concrete problem § Design 10 solutions per theme § Choose 4 preferred solutions § Primary Goals: What do you seek to impact immediately § Secondary Goals: What would you like to impact, but do not have to § Monitoring Goals: What are metrics you want to track
  • 26.
    Optimizely funnel analysis Homepage Flight Results Book Flight Passenger Details Flight Options Payment Page Step 45% 30% 65%65% 90% 70% 40% 40% 15% 20% 5% 25% 15% 30% 20% 15% 5% 5% Exit Rate Divergence Completion
  • 27.
    Optimizely funnel analysis Homepage Flight Results Book Flight Passenger Details Flight Options Payment Page Step Primary/ Secondary Goals § Searches Submitted § Category tabs § Deals § Frequent Flyer § Flight Selection § Filter Usage § Refine Search § Sort Order § Continue button § Edit search § Upsells § Field completion rate § Continue button § Continue button § Upsells § Seat selection § Conversion rate § Form field completion § Payment Type Monitoring Goals § Checkboxes § Flight tracker § Home page clicks § Menu interactions § Site search § Home page clicks § Menu interactions § Site search § Field error rate § Home page clicks § Menu interactions § Site search § Shopping cart pageviews § Home page clicks § Return to previous page § Edit address, information § Error rate
  • 28.
    § Analyze onesection of your application / website § Document 10 different problems for your users § Group them by common root causes § Prioritize by urgency Problem Solution Result Optimizely Hypothesis Workshop § Select 1 concrete problem § Design 10 solutions per theme § Choose 4 preferred solutions § Primary Goals: What do you seek to impact immediately § Secondary Goals: What would you like to impact, but do not have to § Monitoring Goals: What are metrics you want to track