How Slow Pages Hurt: The Entire User

Submitted By jeffsmrctest21
Words: 784
Pages: 4

How Slow Pages Hurt the Entire User
Understanding
the neurological impact
Experience
of poor performance

Tammy Everts
RWD Summit – April 2014

It’s a mobile-first world. 55% of all time spent on retail sites takes place on a mobile device.

comScore, October 2013

Slide 3

Slide 4

Stuart McMillan, Schuh’s Journey to RWD (Conversion Conference 2013)
Slide 5

Four all-too-common mobile assumptions

Assumption #1

My site isn’t slow on mobile.

Slide 7

Radware, 2013 State of the Union: Mobile Ecommerce Performance
Slide 8

Radware, 2013 State of the Union: Mobile Ecommerce Performance
Slide 9

Two things are slowing down your site for mobile (and they’re completely beyond your control).
• Latency – can range from 35 milliseconds to 350+ milliseconds per resource (e.g. images,
CSS files)
• Connection – 3G can be up to 15 times slower than broadband
Slide 10

RWD is awesome, but it *can* come with performance penalties.

Slide 11

Assumption #2

Mobile users expect pages to be slow.

Slide 12

Keynote, 2012 Mobile User Survey
Slide 13

Assumption #3

Mobile users want to browse, not buy.

Slide 14

By 2017, retail mcommerce is expected to hit
$113 billion –
26% of total ecommerce sales.

eMarketer, September 2013
Slide 15

Mobile shopping cart abandonment rate is 39% greater than desktop rate.

2013 Google I/O
Slide 16

Assumption #4

Users will stick around, even if pages are slow, if they really want to buy.

Slide 17

Skava/Harris Interactive, 2013

Slide 18

Case study: The impact of HTML delay on mobile business metrics
Slide 19

Why neuroscientific mobile testing? Slide 20

• 2010 EEG study of desktop users • Throttled connection from
5MB to 2MB
• Found that participants had to concentrate up to 50% harder • Afterward, participants reported negative brand associations Slide 21

Slide 22

Slide 23

What is emotional engagement research?

“95% of the consumer’s decisions are made at the subconscious level.”
Dr. Gerald Zaltman, Harvard UniversityExecutive Committee of Harvard
University’s
Mind, Brain and Behavior Interfaculty Initiative

Slide 25

The problem with surveys… Traditional research relies on eliciting postcognitive responses.
But thinking and talking about emotions changes and distorts them.

Slide 26

Five benefits of neuroscientific testing 1
2
3
4
5

Evaluates think/feel (not say)
Quantified data
Moment-by-moment interaction
Cause-and-effect triggers
Fresh, deeper insights

Slide 27

EEG Emotional Engagement
Study

Our research team
• Seren – leaders in customer experience & service design • NeuroStrata – expert consultants in blending neuromarketing applications
• Neurosense – global leader in implicit methodologies Slide 29

The brands we tested

Slide 30

Our testers
• 24 testers (12 male and 12 female)
• Pre-screened to ensure normal cognitive functioning • Experienced mobile device users
• Did not know they were part of a performance study Slide 31

Methodology
• Standardized set of shopping tasks (browsing and checkout) • Testers served sites over one of two speeds:
– normal Wifi
– artificial 500ms delay

• Using EEG headset and eyetracker, measured moment-by-moment responses

Slide 32

Slide 33

Why test a 500ms delay?

Case study: The impact of HTML delay on mobile business metrics
Slide 34

We focused on the metrics most affected by the
500ms delay.

Frustration
Emotional engagement

Slide 35

Slide 36

Normal speed

2.66s

2.92s

2.83s

4.24s
Slide 37

Frustration levels across sites (normal speed) Slide 38

Engagement levels