Top Banner
93

Ioannidis, anthony salzburg 2015

Aug 18, 2015

Download

Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Ioannidis, anthony   salzburg 2015
Page 2: Ioannidis, anthony   salzburg 2015

User Experience

What is it?

Why should I care?

How does it affect us?

Page 3: Ioannidis, anthony   salzburg 2015

- Why user experience matters

In the next hour or so…

- How you can make $300 million in 1 year…

- Good (and bad) practice examples

…and help a few people in the process

Page 4: Ioannidis, anthony   salzburg 2015

My job

Help my clients improve

4

their customers’ user experience

Page 5: Ioannidis, anthony   salzburg 2015

Why does it matter?

Page 6: Ioannidis, anthony   salzburg 2015
Page 7: Ioannidis, anthony   salzburg 2015
Page 8: Ioannidis, anthony   salzburg 2015
Page 9: Ioannidis, anthony   salzburg 2015

Do you use Facebook?(let me see hands, please!)

Have you looked at the privacy settings?(hands again, please!)

Have you changed any of the privacy settings?(Really?)

Page 10: Ioannidis, anthony   salzburg 2015

What the **** can go wrong with that?

(Remember this question, we’ll come back to it later)

Page 11: Ioannidis, anthony   salzburg 2015

Manipulating opinions

Page 12: Ioannidis, anthony   salzburg 2015

The obvious way

Page 13: Ioannidis, anthony   salzburg 2015

The obvious way

Page 14: Ioannidis, anthony   salzburg 2015

The obvious way

Page 15: Ioannidis, anthony   salzburg 2015

The obvious way

Page 16: Ioannidis, anthony   salzburg 2015

The less obvious way

U.S. Marines help an

Iraqi soldier with

water from a canteen

in Iraq

March 21, 2003

AP

Page 17: Ioannidis, anthony   salzburg 2015

Reportedly… (I found no evidence)

Page 18: Ioannidis, anthony   salzburg 2015

The not-at-all obvious way

For 1 week, in 2012, Facebook hid certain

elements from 689,003 people’s news feed

The aim was to see whether this manipulation

would affect users’ emotions

No informed consent was requested from or

given by the participants

Page 19: Ioannidis, anthony   salzburg 2015

The not-at-all obvious way

The experiment showed that emotions

expressed by others on Facebook do

influence our own emotions

In other words:

Yes, Facebook can manipulate people’s

emotions by cherry-picking what it shows

on their news feed

And no, it doesn’t have to tell you

Page 20: Ioannidis, anthony   salzburg 2015

The ‘innocent’ way

Extra gentle for the

most sensitive skin.Start with ultra sensitive skin, add the chemicals

and moisture and you have diaper rash.

Baby diaper’s unique high-absorbency natural

blend cotton padding provides extra thick,

gel-free protection for your baby.

20

Page 21: Ioannidis, anthony   salzburg 2015

The ‘innocent’ way

Extra gentle for the

most sensitive skin.Start with ultra sensitive skin, add the chemicals

and moisture and you have diaper rash.

Baby diaper’s unique high-absorbency natural

blend cotton padding provides extra thick,

gel-free protection for your baby.

21

Page 22: Ioannidis, anthony   salzburg 2015

The ‘innocent’ way

22

Extra gentle for the

most sensitive skin.Start with ultra sensitive skin, add the chemicals

and moisture and you have diaper rash.

Baby diaper’s unique high-absorbency natural

blend cotton padding provides extra thick,

gel-free protection for your baby.

Page 23: Ioannidis, anthony   salzburg 2015

The ‘innocent’ way

23

Extra gentle for the

most sensitive skin.Start with ultra sensitive skin, add the chemicals

and moisture and you have diaper rash.

Baby diaper’s unique high-absorbency natural

blend cotton padding provides extra thick,

gel-free protection for your baby.

Page 24: Ioannidis, anthony   salzburg 2015

The ‘innocent’ way

24

Extra gentle for the

most sensitive skin.Start with ultra sensitive skin, add the chemicals

and moisture and you have diaper rash.

Baby diaper’s unique high-absorbency natural

blend cotton padding provides extra thick,

gel-free protection for your baby.

Page 25: Ioannidis, anthony   salzburg 2015

Which one can sell more diapers?

1 2

3 4

25

Page 26: Ioannidis, anthony   salzburg 2015

Images can influence us

Composites from actual eye-tracking data

26

Page 27: Ioannidis, anthony   salzburg 2015

Celebrities can influence us

27Post made: September 19, 2014

Joan Rivers passed away on September 4, 2014

Page 28: Ioannidis, anthony   salzburg 2015

What others do can influence us

28

!

Page 29: Ioannidis, anthony   salzburg 2015

The world according to Google

Page 30: Ioannidis, anthony   salzburg 2015

Actual Google UK suggestions

July 20th 2014

The world according to Google

Page 31: Ioannidis, anthony   salzburg 2015

But, I hear you say…

…how do we know these things affect us?

Page 32: Ioannidis, anthony   salzburg 2015

Tools of the trade

Eye-tracking

User testing

A/B testing

Analytics

Page 33: Ioannidis, anthony   salzburg 2015

Eye-tracking…

What it used

to be like

What it is like

these days

33

Page 34: Ioannidis, anthony   salzburg 2015

…what we learn from it…

34

Page 35: Ioannidis, anthony   salzburg 2015

…and how it translates to $

And, of course, don’t forget our diaper ad…35

More expensive

Less

expensive

Page 36: Ioannidis, anthony   salzburg 2015

Tools of the trade

Eye-tracking

User testing

A/B testing

Analytics

Page 37: Ioannidis, anthony   salzburg 2015

User testing

A friendly chat

with users about a

product or website

What they don’t

realise…

37

Page 38: Ioannidis, anthony   salzburg 2015

Does user testing work?

Air France flight 447 disappeared over the

Atlantic Ocean in July 2009

228 dead38

Page 39: Ioannidis, anthony   salzburg 2015

Does user testing work?

Old-style cockpit:

If pilot is pushing forward and co-pilot is

pulling back, their forces are averaged and

the stronger ‘wins’ 39

Page 40: Ioannidis, anthony   salzburg 2015

Does user testing work?

New-style cockpit uses joysticks – no need to

apply any significant force to fly the plane40

Page 41: Ioannidis, anthony   salzburg 2015

Now consider this:

• Pilot is pushing joystick forward

• Co-pilot is pulling joystick back

Should the plane...

• Go down (obey pilot – he knows best)

• Go up (obey co-pilot – pilot is drunk!)

• Average the two, as in the old design?

Airbus decided it makes sense to average

the two, just as the old design worked.

Page 42: Ioannidis, anthony   salzburg 2015

What the **** can go wrong with that?

It makes sense, right?

No need to test it, right?

228 dead

It was never tested with pilots…

42

Page 43: Ioannidis, anthony   salzburg 2015

Does user testing work?

02:13:40 (Pilot) “Climb… climb… climb…”

02:13:40 (Co-pilot) “But I’ve had the stick back the whole time!”

02:13:42 (Engineer) “No, no, no… Don’t climb… no, no.”

02:13:43 (Pilot) “Descend… Give me the controls… Give me the controls!”

To gain speed, the plane should have been in a descend.

The pilot had no idea the co-pilot was pulling it up...

It was never tested with pilots, because ‘it made sense’.

43

Page 44: Ioannidis, anthony   salzburg 2015

Let’s lighten up a bit

What’s an easy way to make $200 million in 1 year?

Page 45: Ioannidis, anthony   salzburg 2015

Tools of the trade

Eye-tracking

User testing

A/B testing

Analytics

Page 46: Ioannidis, anthony   salzburg 2015

Google’s ‘50 shades of blue’

In Google UK's former MD Dan Cobley’s

own words:

46

Seven years ago, Google launched ads on Gmail. In our search we have ads

on the side, little blue links that go to other websites: we had the same thing

on gmail. But we recognised that the shades of blue in those two products

were slightly different.

We ran ‘1% experiments’, showing 1% of users one blue, and another

experiment showing 1% another blue. And actually, to make sure we covered

all our bases, we ran forty other experiments showing all the shades of blue

you could possibly imagine.

Page 47: Ioannidis, anthony   salzburg 2015

Google’s ‘50 shades of blue’

47

And we saw which shades of blue people liked the most, demonstrated by

how much they clicked on them. As a result we learned that a slightly purpler

shade of blue was more conducive to clicking than a slightly greener shade

of blue.

But the implications of that for us, given the scale of our business, was that

we made an extra $200m a year in ad revenue.

Page 48: Ioannidis, anthony   salzburg 2015

OKCupid’s ‘50 shades of gray’

48

OKCupid run some experiments on their users:

• removed text from users' profiles

• removed photos

• told some users they were an excellent

match, when they were only a 30% match

and vice-versa

Page 49: Ioannidis, anthony   salzburg 2015

OKCupid’s ‘50 shades of gray’

49

‘Conversation’ = Exchanging ≥ 4 messages

70% increase!

Page 50: Ioannidis, anthony   salzburg 2015

It all started years ago

50

First large-scale A/B experiments run by

direct marketing (junk mail) companies

A mailshot can go to 10 million people with

5 million different permutations

• Stuffing an envelope with gifts or physical items makes

people more likely to open it

• Thicker paper can make a letter feel more important, so

recipients are likely to give it better attention

• Delivering garden furniture mailshots during sunny spells

resulted in 34% higher sales of outdoor products (IKEA)

Page 51: Ioannidis, anthony   salzburg 2015

P.S: Did you know?

51

• Even the use of a “P.S.”, in which the recipient is prompted to

get in touch with the company and make a purchase has been

tested, and is now considered best practice

Page 52: Ioannidis, anthony   salzburg 2015

But,

What’s an easy way to make $200 million in 1 year?

I promised to show you

how to make $300 million…

Page 53: Ioannidis, anthony   salzburg 2015

Tools of the trade

Eye-tracking

User testing

A/B testing

Analytics

Page 54: Ioannidis, anthony   salzburg 2015

Analytics

Every time you visit a website, we know

• How much time you spend on it

• Where you came from

• Where you’re off to

• What you are interested in

• How old you are

• Where you’re from

• What you did last summer

54

Page 55: Ioannidis, anthony   salzburg 2015

Analytics

55

Page 56: Ioannidis, anthony   salzburg 2015

Analytics

The old form: massive drop-out rates

Page 57: Ioannidis, anthony   salzburg 2015

Analytics

45% of customers had multiple accounts in

the system, some as many as

160,000 forgotten password requests / day

75% of these never tried to complete

the purchase once requested

10

57

Page 58: Ioannidis, anthony   salzburg 2015

Analytics

The new form: no registration required58

Page 59: Ioannidis, anthony   salzburg 2015

Analytics

After the change:

45% increase in purchases

An extra $15 million in the first month

$300,000,000 in the first year

59

Page 60: Ioannidis, anthony   salzburg 2015

Three tips to improve user experience

(which you can actually use)

Page 61: Ioannidis, anthony   salzburg 2015

1. Less is more

61

Page 62: Ioannidis, anthony   salzburg 2015

62

Page 63: Ioannidis, anthony   salzburg 2015

63

Page 64: Ioannidis, anthony   salzburg 2015

By the way…

Did you notice this section had no text?

You still got it, right?

64

Page 65: Ioannidis, anthony   salzburg 2015

65

2. Signposting

Page 66: Ioannidis, anthony   salzburg 2015

Good.

66

Page 67: Ioannidis, anthony   salzburg 2015

Not.

so.

good.

Page 68: Ioannidis, anthony   salzburg 2015

Solving life’s little problems

A problem for busy airports:

- men’s aiming is very poor when using urinals

Schiphol Airport in Amsterdam found a solution:

- signposting

68

Page 69: Ioannidis, anthony   salzburg 2015

Erm…

Spillage reduced by 80%(true story!)

69

Page 70: Ioannidis, anthony   salzburg 2015

SEGA took it a step further

“Toylet”

A game that rewards… good aiming70

Page 71: Ioannidis, anthony   salzburg 2015

3. Progressive

disclosure

Page 72: Ioannidis, anthony   salzburg 2015

Progressive disclosure

72

“ An interaction design technique that sequences

information across several sections, in order to help

maintain the focus of a user's attention by reducing

clutter, confusion, and cognitive workload. By disclosing

information progressively, you reveal only the essentials

and help the user manage the complexity of feature-rich

sites or applications.

Wikipedia (again)

Page 73: Ioannidis, anthony   salzburg 2015

Progressive disclosure

A design technique that simplifies and

sequences information across several sections

Helps maintain a user's attention by reducing

• clutter

• confusion

• cognitive workload

Reveals only the essentials and helps manage

complexity

73

Page 74: Ioannidis, anthony   salzburg 2015

Manipulating behaviour

Manipulating opinions

Page 75: Ioannidis, anthony   salzburg 2015

Online vs offline

75

Interfaces that mimic the physical world

But what about human behaviour?

Page 76: Ioannidis, anthony   salzburg 2015

Do we behave the same way online?

76

vs

Real life

Page 77: Ioannidis, anthony   salzburg 2015

77

Why the different behaviour?

Social media has taught us that it’s OK to do

‘creepy’ things online (follow, like, poke...)

‘Everybody’ does it, so it’s socially acceptable

We are influenced by what our peers do

‘Social’ sites take advantage of this

Page 78: Ioannidis, anthony   salzburg 2015

A good example

Showing this message

on Facebook drove

about 60,000 extra

votes in the 2010 US

Congressional

elections

Showing this one

drove a further

282,000 votes

Page 79: Ioannidis, anthony   salzburg 2015

A bad example

Facebook unlawfully used the names and profile

pictures of its users, including minors, to

advertise without obtaining their consent

1 class action lawsuit settled ($20 million)

1 more in Canada, recently ruled to be

outside the local Courts’ jurisdiction

Page 80: Ioannidis, anthony   salzburg 2015

In a student survey…

1. How happy are you?

2. How often do you go on a date?

If asked in this order

- correlation is only 0.11

If asked in the reverse order

- correlation jumps to 0.62!

80

Page 81: Ioannidis, anthony   salzburg 2015

Choice Architecture

We have the power to frame questions in a

way that manipulates answers

We have the power to nudge people to the

right (or wrong) direction

A force for good or evil? (Discuss)

81

Page 82: Ioannidis, anthony   salzburg 2015

Choice Architecture

How much would you like to donate?

- $10

- $25

- $100

- $500

- Other amount

- $50

- $100

- $500

- $1,000

- Other amount

82

Page 83: Ioannidis, anthony   salzburg 2015

Choice Architecture

Simply by rearranging food in a school

cafeteria, researchers were able to increase or

decrease consumption of particular foods by

25%

(We do that every day, when you visit your favourite websites

– we just don’t tell you)

83

Page 84: Ioannidis, anthony   salzburg 2015

Can you

change the world?

Page 85: Ioannidis, anthony   salzburg 2015

Getting people to exercise

Page 86: Ioannidis, anthony   salzburg 2015

Getting people to donate

Research in South Korea showed that about

89% of people wanted to donate

In reality only 0.9% actually did

How can we get more people to donate?

86

Page 87: Ioannidis, anthony   salzburg 2015

Getting people to donate

Minewater has two barcodes:

• Scan one, and you'll be charged $1

• Scan the other, you'll donate 10 cents

87

Page 88: Ioannidis, anthony   salzburg 2015

Getting people to donate

51% of customers donated in the first 2 weeks

Sales rose by 244%

88

Page 89: Ioannidis, anthony   salzburg 2015

Last video for todayBut possibly the most important

How can you change the world?

Some ideas on how to tackle your projects(before you end up like this guy)

Page 90: Ioannidis, anthony   salzburg 2015

90

Page 91: Ioannidis, anthony   salzburg 2015

Over to you…

Please sign up for 2 x consultationsQuestions?

Page 92: Ioannidis, anthony   salzburg 2015

[email protected]

www.twitter.com/iasquare

Thank you

Page 93: Ioannidis, anthony   salzburg 2015

Sources• http://usableworld.com.au/2009/03/16/you-look-where-they-look/

• http://www.uie.com/articles/three_hund_million_button/

• http://onlinelibrary.wiley.com/doi/10.1002/ejsp.2420180505/abstract

• http://www.smartinsights.com/conversion-optimisation/checkout-optimisation

• http://www.fastcodesign.com/1669720/how-lousy-cockpit-design-crashed-an-airbus-killing-228-people

• http://www.bbc.co.uk/news/science-environment-19571053

• http://soulpancake.com/

• Design Thinking video (edited), by Brettnewman, Wikimedia Commons

• Ogilvy & Mather Dubai, for the RSF campaign

• Crisis Relief, for the ‘Liking isn’t helping’ campaign

• UNwomen.org, for the Google search posters

• Thaler & Sunstein, Nudge, Penguin books, 2009

• Animated gifs from giphy.com

All photos and videos © of their respective owners