Increasing User Engagement: How one company tested its site navigation and increased clickthrough by over 35%

Increasing User Engagement:
How one company tested its site navigation and increased
clickthrough by over 35%

We’re sharing on Twitter!
#WebClinic

#WebClinic
Ben Filip
Senior Manager,
Data Sciences
MECLABS Institute
Today’s speakers
Austin McCraw
Senior Director,
Content Production
MECLABS Institute
Taylor Kennedy
Senior Manager,
Optimization & Training
MECLABS Institute

#WebClinic
Experiment: Background
Background: A tourism commission seeking to enhance visitor interaction with its website
content in order to boost appeal for choosing its city over other destinations.
Goal: Test the site navigation to increase visitor engagement with key site content.
Primary Research Question: Which navigation type will increase site engagement?
Secondary Research Question: Which navigation type will lead to the highest lead generation
rate?
Test Design: A/B/C split test
Experiment ID: TP 11211
Record Location: MECLABS Research Library
Research Partner: (Protected)

#WebClinic
Experiment: Control
The Control has:
• Six main categories in the
navigation
• A drop-down structure with
numerous options under each
heading
• A relatively narrow pixel-width
compared to site header
*Anonymized

#WebClinic
Experiment: Treatment A
*Anonymized
In Treatment A, we:
• Utilized a new color to bring
contrast to the navigation
menu
• Reduced the number of
categories from six to four
• Increased the font size and
legibility of the navigation

#WebClinic
Experiment: Treatment B
*Anonymized
In Treatment B, we:
• Utilized a new color to bring
contrast to the navigation menu
• Reduced the number of categories
from six to three
• Increased the font size and
legibility of the navigation
• Employed images within the
navigation to communicate key
content groups (visual navigation)

#WebClinic
Experiment: Quick comparison
Audience Question:
Which navigation structure will increase site engagement?
?
*Anonymized
Treatment B
Control
Treatment A

#WebClinic
Design Clickthrough Rate Level of Confidence % Relative Change
Control 19.39% ---
Treatment A 26.26% 99% 35.39%
Treatment B 19.58% 55% 0.99%
Experiment: Results
Relative increase in navigation CTR35%Treatment A resulted in a 35.39% relative increase in navigation clickthrough rate.

#WebClinic
Design Pages/Unique Visitors Level of Confidence % Relative Change
Control 3.57 ---
Treatment A 3.75 89% 5.05%
Treatment B 3.63 55% 0.99%
Experiment: Results
No difference in pageviews per visitor
There was no statistical difference in the number of pages viewed per unique visitor.
*Percent relative change and LOC numbers calculated from raw unique visitor and pageview data.

#WebClinic
Design Pages/Unique Visitors Level of Confidence % Relative Change
Control 3.57 ---
Treatment A 3.75 89% 5.05%
Treatment B 3.63 55% 0.99%
No difference in Page Views per Visitor
There was no statistical difference in the number of pages viewed per Unique Visitor
*Percent relative change and LOC numbers calculated from raw unique visitor and pageview data
How do we determine what constitutes
meaningful engagement?

#WebClinic
Today’s metric overload

#WebClinic
Today, we are going to help you define
meaningful engagement and walk you
through three tactics for helping you boost
engagement on your site.
Meaningful engagement

#WebClinic
What is “meaningful engagement”?
1. Engagement – First, we must understand that the term “engagement” is a
noun that represents an action set. At its essence, engagement is the
fundamental relationship between the subject (the who?) and the predicate
(the what?).
2. Meaningful – Next, we must understand that in order to be considered
“meaningful,” the relationship between the subject and the object must
inform you about the behavior of the visitor on your site (the why?).
“Meaningful Engagement”

#WebClinic
Example: BuzzFeed
BuzzFeed’s Stated Objective:
“To provide the most shareable
breaking news, original reporting,
entertainment and video across the
social web to a global audience”

#WebClinic
Example: BuzzFeed
*Anonymized
Who: Visitors
What: Pageviews per session
(average page depth).
The Why: The visitor is (or is
not) interested in the type of
content.

#WebClinic
Example: BuzzFeed
*Anonymized
Who: Visitors
What: Video events per
unique visitor.
The Why: The visitor does (or
does not) value video as a
medium of communication.

#WebClinic
Example: BuzzFeed
*Anonymized
Who: Visitors
What: Social interaction per
page (story).
The Why: The content is (or is
not) relevant to the
motivations of the visitor.
Enables better targeting for
additional content
or advertising.

#WebClinic
Example: BuzzFeed
*Anonymized
Who: Visitors
What: Social interaction per
page (story)
The Why: The content is (or is
not) relevant to the
motivations of the visitor.
Enables better targeting for
additional content
or advertising.
The goal is not to discover metrics with
meaning, but rather to discover the meaning in
the metrics. The value in metrics is not the
“what.” It is the “why.”

#WebClinic
Metrics to focus on:
• Number of unique visits
• Frequency of visit
• Recency of visit
• Depth of visit
• Time on site
• Bounce rate (inverse)
• Avg. time on page
• Open rate
• Etc.
Common “engagement” metrics
• Click tracking for key
on-page elements
• Comments
• Email sign-ups
• Feed subscribes
• Chats
• Social followers
• Likes, +1’s
• Clickthrough
• Etc.
• Total conversions
• Lead submissions
• Social shares
• Retweets
• Contact initiations
• Direct references
• Back-links
• Etc.
Interested
in Content
Interactive
with Content
Influenced
by Content
1 2 3

#WebClinic
How can we optimize our website to
increase these kind of engagement
metrics?
?

#WebClinic
Tactic #1:
Actively curate your content

#WebClinic
Background: A tourism commission seeking to enhance visitor interaction with their website
content in order to boost appeal for choosing their city over other destinations.
Goal: Test the site navigation to increase visitor engagement with key site content.
Primary Research Question: Which navigation type will increase site engagement?
Secondary Research Question: Which navigation type will lead to the highest lead generation
rate?
Test Design: A/B/C split test
Research Partner: (Protected)

#WebClinic
Experiment: Control
• Navigation increases friction by
including an almost-overwhelming
amount of options.
• As marketers, we must determine
which content or category of
content is most valuable to the
visitor.
*Anonymized

#WebClinic
Experiment: Treatment creation
• To create each treatment, our
research team used analytics
data to identify the most popular
content on the site.
*Anonymized

#WebClinic
Experiment: Treatments
*Anonymized
Treatment A
Treatment B
• Primary navigation tabs were
streamlined based upon visitor
interaction data with site content
over a calendar year.
• The most engaged-with category
content and secondary content were
prioritized and emphasized, while
less popular content was removed in
order to reduce friction.

#WebClinic
F
Design Lead Gen Rate Level of Confidence % Relative Change
Control 1.01% ---
Treatment A 1.36% 99% 34.93%
Treatment B 1.19% 91% 18.67%
Experiment: Results
Relative increase in total leads34%Treatment A resulted in a 34.93% relative increase in total leads.
What you need to understand:
Navigation content was curated based upon visitor behavior with the control nav.

#WebClinic
F
Design Lead Gen Rate Level of Confidence % Relative Change
Control 1.01% ---
Treatment A 1.36% 99% 34.93%
Treatment B 1.19% 91% 18.67%
Experiment: Results
Relative increase in total leads34%Treatment A resulted in a 34.93% relative increase in total leads
What you need to understand:
Navigation content was curated based upon visitor behavior with the control nav.
Do not simply rely on your marketer’s intuition to
determine which content is valuable to your visitors.
Actively leverage existing “interest” data to deliver
the most relevant content to your visitors.

#WebClinic
Tactic #2:
Clarify the presentation of your
content

#WebClinic
Background: Dr. McGlaughlin's personal website, which utilizes email to increase viewership
and audience engagement.
Goal: To utilize responsive email to increase audience engagement with email sends.
Research Question: Which email type will generate the highest read rate?
Test Design: A/B split test
Experiment ID: Internal
Research Partner: FlintsNotes.com

#WebClinic
Experiment: Control
• The copy of the email is barely
legible on mobile phones,
requiring excessive resizing by the
recipient.
• Is this the ideal experience for a
message type that is supposed to
be read?

#WebClinic
Experiment: Treatment
• The treatment utilizes a responsive
format that simplifies the
presentation of the core email
component — the message.
• Font size automatically adjusts to the
screen width of the mobile device.
• Secondary elements are moved into
the single column and remain at the
bottom of the email.

#WebClinic
Experiment: Side-by-side comparison
Control Treatment

#WebClinic
Design Read Rate % Relative Change
Control 55.67%
Treatment 72.33% 29.9%
Experiment: Results
Relative increase in read rate29%The Treatment resulted in a 29.9% relative increase in email read rate.

#WebClinic
Control 55.67%
Experiment: Results
Relative increase in read rate29%The treatment resulted in a 29.9% relative increase in email read rate
Having relevant content is not enough. We must
learn to not only optimize the essence (focus) of our
content, but also it’s expression (form).

#WebClinic
Tactic #3:
Connect your content to the
next logical action

#WebClinic
Background: Migraine Treatment Centers of America offers an innovative long-term
migraine treatment solution to people suffering from migraines.
Goal: To increase leads from the microsite.
Primary Research Question: Which value exchange strategy will result in a higher
conversion rate?
Approach: A/B multifactor split
Record Location: MarketingExperiments Research Library
Research Partner: Migraine Treatment Centers of America
Research Notes:


#WebClinic
Experiment: Control
• There are several things wrong
with this Control, but one of the
biggest problems is that this site
does not effectively connect the
momentum created by its
content to the next logical step.
• There are CTAs on the site, but
they are disconnected from
content. There is content on the
site, but it is disconnected from
the CTAs.
Control: Site Map

#WebClinic
Experiment: Treatment
Notice a few key things in the
Treatment:
• The site is not simply organized
by content types, but rather by
action-oriented engagement.
• The main CTA is now directly
connected to the momentum
created by the content.

#WebClinic
Control Treatment
Experiment: Side-by-side

#WebClinic
Experiment: Results
Control 1.167%
Relative increase in lead rate331%The Treatment generated 330.6% more conversions than the Control.

#WebClinic
Experiment: Results
Control 1.167%
Relative increase in Lead rate331%The Treatment generated 330.6% more conversions than the ControlThe goal of all marketing is to influence choice.
Every piece of content on your site should clearly
lead the visitor to the next logical step in the
cognitive journey.

#WebClinic
Maximizing subscription revenue
Tactics for Increasing Engagement:
1. Actively curate your content: Do not simply rely on your marketer’s intuition to determine
which content is valuable to your visitors. Actively leverage existing “interest” data to
deliver the most relevant content to your visitors.
1. Clarify your presentation: Having relevant content is not enough. We must learn to not
only optimize the essence (focus) of our content, but also it’s expression (form).
2. Connect your content to the next logical action: The goal of all marketing is to influence
choice. Every piece of content on your site should clearly lead the visitor to the next logical
step in the cognitive journey.

#WebClinic
Live Optimization: Scottrade
Primary Audience: Prospective customers
of Scottrade investment products.
Page Purpose: Open a Scottrade
investment account.
Top of Page
Middle of Page

#WebClinic
Live Optimization: Skillsera
Primary Audience: Visitors looking to
enhance their career skills using online
training.
Page Purpose: Request more information
on a specific training or certification.
Homepage

#WebClinic
Live Optimization: Medgate
Primary Audience: Prospects interested in
more information around safety
management software.
Page Purpose: Request more information
directly via phone or contact form.
Services Page

#WebClinic
Live Optimization: Strayer University
Primary Audience: Visitors searching for
online education services.
Page Purpose: Obtain contact and other
important information about prospective
students.
Landing Page

#WebClinic
Live Optimization: Home Science Tools
Primary Audience: Visitors searching for
safe, accessible science tools and materials.
Page Purpose: Purchase science
equipment.
Top of Page
Bottom of Page

#WebClinic
Interested in optimizing your online business?
MECLABS conducts rigorous experiments in the new science of optimization. We
apply our discoveries to help leaders optimize the financial performance of
their sales and marketing programs.
Learn more about how you may be a fit
for a MECLABS Research Partnership:
• Select Research Partnership
Opportunities on the post-webinar
survey
• Contact us directly:
info@MECLABS.com
1-877-635-0565
x

Increasing User Engagement: How one company tested its site navigation and increased clickthrough by over 35%

More Related Content

What's hot (20)

Similar to Increasing User Engagement: How one company tested its site navigation and increased clickthrough by over 35% (20)

More from MarketingExperiments (12)

Recently uploaded (20)

Increasing User Engagement: How one company tested its site navigation and increased clickthrough by over 35%