1
00:00:01,440 --> 00:00:04,496
How can we get people to do more good,

2
00:00:04,520 --> 00:00:08,856
to go to the polls,
give to charity, conserve resources,

3
00:00:08,880 --> 00:00:12,216
or even to do something as simple
as washing their mugs at work

4
00:00:12,240 --> 00:00:14,621
so that the sink isn't
always full of dirty dishes?

5
00:00:14,645 --> 00:00:16,656
(Laughter)

6
00:00:16,680 --> 00:00:21,016
(Applause)

7
00:00:21,040 --> 00:00:23,176
When I first started
working on this problem,

8
00:00:23,200 --> 00:00:25,096
I collaborated with a power company

9
00:00:25,120 --> 00:00:28,656
to recruit customers for a program
that prevents blackouts

10
00:00:28,680 --> 00:00:30,680
by reducing energy demand during peaks.

11
00:00:31,600 --> 00:00:34,216
The program is based
on a tried-and-true technology.

12
00:00:34,240 --> 00:00:36,616
It's one the Obama
Administration even called

13
00:00:36,640 --> 00:00:39,760
"the cornerstone to modernizing
America's electrical grid."

14
00:00:40,880 --> 00:00:44,376
But, like so many
great technological solutions,

15
00:00:44,400 --> 00:00:45,840
it has a key weakness:

16
00:00:47,760 --> 00:00:48,960
people.

17
00:00:49,720 --> 00:00:50,960
People need to sign up.

18
00:00:52,000 --> 00:00:55,536
To try to get people to sign up,
the power company sent them a nice letter,

19
00:00:55,560 --> 00:00:57,576
told them about
all the program's benefits,

20
00:00:57,600 --> 00:01:00,656
and it asked them to call
into a hotline if they were interested.

21
00:01:00,680 --> 00:01:02,376
Those letters went out,

22
00:01:02,400 --> 00:01:05,160
but the phones, they were silent.

23
00:01:06,080 --> 00:01:08,840
So when we got involved,
we suggested one small change.

24
00:01:09,360 --> 00:01:11,256
Instead of that hotline,

25
00:01:11,280 --> 00:01:15,896
we suggested that they use sign-up sheets
that they'd post near the mailboxes

26
00:01:15,920 --> 00:01:17,160
in people's buildings.

27
00:01:18,120 --> 00:01:20,880
This tripled participation.

28
00:01:23,160 --> 00:01:24,360
Why?

29
00:01:25,680 --> 00:01:29,896
Well, we all know people care deeply
about what others think of them,

30
00:01:29,920 --> 00:01:32,576
that we try to be seen
as generous and kind

31
00:01:32,600 --> 00:01:35,320
and we try to avoid
being seen as selfish or a mooch.

32
00:01:36,120 --> 00:01:40,536
Whether we are aware of it or not,
this is a big part of why people do good,

33
00:01:40,560 --> 00:01:45,776
and so small changes that give people
more credit for doing good,

34
00:01:45,800 --> 00:01:48,136
those changes can make
a really big difference.

35
00:01:48,160 --> 00:01:51,056
Small changes like
switching from a hotline,

36
00:01:51,080 --> 00:01:54,096
where nobody will ever find out
about your good deed,

37
00:01:54,120 --> 00:01:55,856
to a sign-up sheet

38
00:01:55,880 --> 00:01:59,000
where anyone who walks by
can see your name.

39
00:02:00,520 --> 00:02:03,696
In our collaborations with governments,
nonprofits, companies,

40
00:02:03,720 --> 00:02:06,536
when we're trying to get people
to do more good,

41
00:02:06,560 --> 00:02:08,960
we harness the power of reputations

42
00:02:10,000 --> 00:02:12,376
and we have a simple checklist for this,

43
00:02:12,400 --> 00:02:15,640
and in fact, you already know
the first item on that checklist.

44
00:02:16,960 --> 00:02:19,256
It's to increase observability,

45
00:02:19,280 --> 00:02:22,000
to make sure people find out
about good deeds.

46
00:02:23,440 --> 00:02:26,336
Now, wait a minute, I know
some of you are probably thinking,

47
00:02:26,360 --> 00:02:28,056
there's no way people here thought,

48
00:02:28,080 --> 00:02:30,736
"Oh, well, now that
I'm getting credit for my good deed,

49
00:02:30,760 --> 00:02:32,096
now it's totally worth it."

50
00:02:32,120 --> 00:02:33,576
And you're right.

51
00:02:33,600 --> 00:02:34,840
Usually, people don't.

52
00:02:35,760 --> 00:02:38,816
Rather, when they're making
decisions in private,

53
00:02:38,840 --> 00:02:40,576
they worry about their own problems,

54
00:02:40,600 --> 00:02:44,736
about what to put on the table for dinner,
or how to pay their bills on time.

55
00:02:44,760 --> 00:02:47,856
But, when we make
their decision more observable,

56
00:02:47,880 --> 00:02:50,960
they start to attend more
to the opportunity to do good.

57
00:02:51,760 --> 00:02:54,576
In other words, what's
so powerful about our approach

58
00:02:54,600 --> 00:02:59,456
is that it could turn on
people's existing desire to do good,

59
00:02:59,480 --> 00:03:01,560
in this case, to help
to prevent a blackout.

60
00:03:03,160 --> 00:03:04,816
Back to observability.

61
00:03:04,840 --> 00:03:06,696
I want to give you another example.

62
00:03:06,720 --> 00:03:08,456
This one is from a collaboration

63
00:03:08,480 --> 00:03:10,856
with a nonprofit that gets out the vote,

64
00:03:10,880 --> 00:03:14,456
and it does this by sending hundreds
of thousands of letters every election

65
00:03:14,480 --> 00:03:17,800
in order to remind people and try
to motivate them to go to the polls.

66
00:03:18,760 --> 00:03:20,920
We suggested adding
the following sentence:

67
00:03:22,440 --> 00:03:25,840
"Someone may call you to find out
about your experience at the polls."

68
00:03:26,520 --> 00:03:29,920
This sentence makes it feel
more observable when you go to the polls,

69
00:03:31,040 --> 00:03:34,080
and it increased the effect
of the letter by 50 percent.

70
00:03:36,680 --> 00:03:40,416
Making the letter more effective reduced
the cost of getting an additional vote

71
00:03:40,440 --> 00:03:42,400
from 70 dollars down to about 40 dollars.

72
00:03:43,200 --> 00:03:45,136
Observability has been used to do things

73
00:03:45,160 --> 00:03:47,816
like get people
to donate blood more frequently

74
00:03:47,840 --> 00:03:50,816
by listing the names of donors
on local newsletters,

75
00:03:50,840 --> 00:03:52,480
or to pay their taxes on time

76
00:03:53,440 --> 00:03:56,256
by listing the names of delinquents
on a public website.

77
00:03:56,280 --> 00:03:58,120
(Laughter)

78
00:04:00,200 --> 00:04:01,400
What about this example?

79
00:04:02,560 --> 00:04:06,616
Toyota got hundreds of thousands of people
to buy a more fuel-efficient car

80
00:04:06,640 --> 00:04:10,256
by making the Prius so unique ...

81
00:04:10,280 --> 00:04:11,880
(Laughter)

82
00:04:12,800 --> 00:04:15,480
that their good deed
was observable from a mile away.

83
00:04:16,040 --> 00:04:18,856
(Laughter)

84
00:04:18,880 --> 00:04:21,255
Alright, so observability is great,

85
00:04:21,279 --> 00:04:24,816
but we all know, we've all seen

86
00:04:24,840 --> 00:04:27,000
people walk by an opportunity to do good.

87
00:04:28,080 --> 00:04:31,376
They'll see somebody
asking for money on the sidewalk

88
00:04:31,400 --> 00:04:34,016
and they'll pull out their phones
and look really busy,

89
00:04:34,040 --> 00:04:37,600
or they'll go to the museum and they'll
waltz right on by the donation box.

90
00:04:38,320 --> 00:04:40,416
Imagine it's the holiday season

91
00:04:40,440 --> 00:04:44,136
and you're going to the supermarket,
and there's the Salvation Army volunteer,

92
00:04:44,160 --> 00:04:45,416
and he's ringing his bell.

93
00:04:45,440 --> 00:04:47,416
A few years ago, researchers in San Diego

94
00:04:47,440 --> 00:04:50,896
teamed up with a local chapter
from the Salvation Army

95
00:04:50,920 --> 00:04:53,120
to try to find ways to increase donations.

96
00:04:54,240 --> 00:04:55,860
What they found was kind of funny.

97
00:04:56,680 --> 00:04:59,080
When the volunteer
stood in front of just one door,

98
00:05:00,040 --> 00:05:02,600
people would avoid giving
by going out the other door.

99
00:05:04,880 --> 00:05:06,080
Why?

100
00:05:07,080 --> 00:05:10,776
Well, because they can always claim,
"Oh, I didn't see the volunteer,"

101
00:05:10,800 --> 00:05:13,056
or, "I wanted to get
something from over there,"

102
00:05:13,080 --> 00:05:14,480
or, "That's where my car is."

103
00:05:15,560 --> 00:05:17,480
In other words, there's lots of excuses.

104
00:05:18,640 --> 00:05:21,616
And that brings us
to the second item on our checklist:

105
00:05:21,640 --> 00:05:23,360
to eliminate excuses.

106
00:05:24,800 --> 00:05:26,456
In the case of the Salvation Army,

107
00:05:26,480 --> 00:05:29,736
eliminating excuses just means
standing in front of both doors,

108
00:05:29,760 --> 00:05:31,816
and sure enough, when they did this,

109
00:05:31,840 --> 00:05:33,040
donations rose.

110
00:05:35,280 --> 00:05:37,456
But that's when things got kind of funny,

111
00:05:37,480 --> 00:05:38,680
even funnier.

112
00:05:39,960 --> 00:05:42,496
The researchers
were out in the parking lot,

113
00:05:42,520 --> 00:05:45,696
and they were counting people
as they came in and out of the store,

114
00:05:45,720 --> 00:05:49,536
and they noticed that when the volunteers
stood in front of both doors,

115
00:05:49,560 --> 00:05:51,736
people stopped coming
out of the store at all.

116
00:05:51,760 --> 00:05:54,816
(Laughter)

117
00:05:54,840 --> 00:05:58,976
Obviously, they were surprised by this,
so they decided to look into it further,

118
00:05:59,000 --> 00:06:04,096
and that's when they found that there
was actually a third, smaller utility door

119
00:06:04,120 --> 00:06:06,096
usually used to take out the recycling --

120
00:06:06,120 --> 00:06:07,656
(Laughter)

121
00:06:07,680 --> 00:06:11,176
and now people were going out that door
in order to avoid the volunteers.

122
00:06:11,200 --> 00:06:13,560
(Laughter)

123
00:06:14,320 --> 00:06:16,640
This teaches us
an important lesson though.

124
00:06:18,480 --> 00:06:22,016
When we're trying to eliminate excuses,
we need to be very thorough,

125
00:06:22,040 --> 00:06:24,416
because people are
really creative in making them.

126
00:06:24,440 --> 00:06:26,640
(Laughter)

127
00:06:29,760 --> 00:06:31,616
Alright, I want to switch this setting

128
00:06:31,640 --> 00:06:33,840
where excuses can have
deadly consequences.

129
00:06:36,160 --> 00:06:39,880
What if I told you that the world's
deadliest infectious disease has a cure,

130
00:06:40,960 --> 00:06:43,896
in fact, that it's had one for 70 years,

131
00:06:43,920 --> 00:06:46,120
a good one, one that works
almost every time?

132
00:06:47,920 --> 00:06:49,480
It's incredible, but it's true.

133
00:06:50,400 --> 00:06:52,376
The disease is tuberculosis.

134
00:06:52,400 --> 00:06:54,696
It infects some 10 million people a year,

135
00:06:54,720 --> 00:06:56,840
and it kills almost two million of them.

136
00:06:57,680 --> 00:07:01,736
Like the blackout prevention program,
we've got the solution.

137
00:07:01,760 --> 00:07:03,000
The problem is people.

138
00:07:04,000 --> 00:07:05,736
People need to take their medication

139
00:07:05,760 --> 00:07:07,656
so that they're cured,

140
00:07:07,680 --> 00:07:09,840
and so that they don't
get other people sick.

141
00:07:11,520 --> 00:07:13,656
For a few years now,
we've been collaborating

142
00:07:13,680 --> 00:07:15,976
with a mobile health startup
called Keheala

143
00:07:16,000 --> 00:07:18,680
to support TB patients
as they undergo treatment.

144
00:07:19,280 --> 00:07:22,376
Now, you have to understand,
TB treatment, it's really tough.

145
00:07:22,400 --> 00:07:24,976
We're talking about taking
a really strong antibiotic

146
00:07:25,000 --> 00:07:27,336
every single day for six months or more.

147
00:07:27,360 --> 00:07:30,256
That antibiotic is so strong
that it will make you feel sick.

148
00:07:30,280 --> 00:07:32,256
It will make you feel nauseous and dizzy.

149
00:07:32,280 --> 00:07:34,200
It will make your pee turn funny colors.

150
00:07:34,720 --> 00:07:37,616
It's also a problem because
you have to go back to the clinic

151
00:07:37,640 --> 00:07:40,056
about every week
in order to get more pills,

152
00:07:40,080 --> 00:07:43,536
and in sub-Saharan Africa
or other places where TB is common,

153
00:07:43,560 --> 00:07:46,016
now you're talking
about going someplace pretty far,

154
00:07:46,040 --> 00:07:49,056
taking tough and slow public transport,

155
00:07:49,080 --> 00:07:51,136
maybe the clinic is inefficient.

156
00:07:51,160 --> 00:07:54,376
So now you're talking about taking
a half day off of work every week

157
00:07:54,400 --> 00:07:57,280
from a job you desperately
can't afford to lose.

158
00:07:57,920 --> 00:08:01,416
It's even worse when you consider the fact
that there's a terrible stigma

159
00:08:01,440 --> 00:08:04,856
and you desperately don't want people
to find that you have the disease.

160
00:08:04,880 --> 00:08:07,736
Some of the toughest stories we hear
are actually from women

161
00:08:07,760 --> 00:08:11,296
who, in these places where
domestic violence can be kind of common,

162
00:08:11,320 --> 00:08:14,056
they tell us that they have to
hide it from their husbands

163
00:08:14,080 --> 00:08:15,720
that they're coming to the clinic.

164
00:08:17,800 --> 00:08:20,600
So it's no surprise
that people don't complete treatment.

165
00:08:21,840 --> 00:08:23,976
Can our approach really help them?

166
00:08:24,000 --> 00:08:25,880
Can we really get them to stick it out?

167
00:08:28,040 --> 00:08:29,240
Yeah.

168
00:08:30,080 --> 00:08:34,176
Every day, we text patients
to remind them to take their medication,

169
00:08:34,200 --> 00:08:35,736
but if we stopped there,

170
00:08:35,760 --> 00:08:37,816
there'd be lots of excuses.

171
00:08:37,840 --> 00:08:39,296
"Well, I didn't see the text."

172
00:08:39,320 --> 00:08:42,015
Or, "You know, I saw the text,
but then I totally forgot,

173
00:08:42,039 --> 00:08:44,256
put the phone down
and I just forgot about it."

174
00:08:44,280 --> 00:08:46,080
Or, "I lent the phone out to my mom."

175
00:08:47,600 --> 00:08:49,616
We have to eliminate these excuses

176
00:08:49,640 --> 00:08:51,656
and we do that by asking patients

177
00:08:51,680 --> 00:08:54,640
to log in and verify
that they've taken their medication.

178
00:08:55,680 --> 00:08:57,936
If they don't log in, we text them again.

179
00:08:57,960 --> 00:09:00,056
If they don't log in,
we text them yet again.

180
00:09:00,880 --> 00:09:04,096
If, after three times,
they still haven't verified,

181
00:09:04,120 --> 00:09:05,936
we notify a team of supporters

182
00:09:05,960 --> 00:09:08,216
and that team will call and text them

183
00:09:08,240 --> 00:09:10,040
to try to get them back on the wagon.

184
00:09:10,960 --> 00:09:12,160
No excuses.

185
00:09:13,520 --> 00:09:17,296
Our approach, which, admittedly,
uses all sorts of behavioral techniques,

186
00:09:17,320 --> 00:09:20,536
including, as you've
probably noticed, observability,

187
00:09:20,560 --> 00:09:21,800
it was very effective.

188
00:09:22,800 --> 00:09:24,976
Patients without access to our platform

189
00:09:25,000 --> 00:09:28,160
were three times more likely
not to complete treatment.

190
00:09:31,520 --> 00:09:32,896
Alright,

191
00:09:32,920 --> 00:09:34,456
you've increased observability,

192
00:09:34,480 --> 00:09:36,696
you've eliminated excuses,

193
00:09:36,720 --> 00:09:39,360
but there's still a third thing
you need to be aware of.

194
00:09:40,800 --> 00:09:44,056
If you've been to Washington, DC
or Japan or London,

195
00:09:44,080 --> 00:09:45,776
you know that metro riders there

196
00:09:45,800 --> 00:09:49,256
will be very careful to stand
on the right-hand side of the escalator

197
00:09:49,280 --> 00:09:51,200
so that people can go by on the left.

198
00:09:52,160 --> 00:09:54,536
But unfortunately,
not everywhere is that the norm

199
00:09:54,560 --> 00:09:57,736
and there's plenty of places
where you can just stand on both sides

200
00:09:57,760 --> 00:09:59,016
and block the escalator.

201
00:09:59,040 --> 00:10:00,656
Obviously, it's better for others

202
00:10:00,680 --> 00:10:02,976
when we stand on the right
and let them go by,

203
00:10:03,000 --> 00:10:05,720
but we're only expected
to do that some places.

204
00:10:06,720 --> 00:10:08,136
This is a general phenomenon.

205
00:10:08,160 --> 00:10:10,216
Sometimes we're expected to do good

206
00:10:10,240 --> 00:10:12,056
and sometimes not,

207
00:10:12,080 --> 00:10:15,456
and it means that people
are really sensitive to cues

208
00:10:15,480 --> 00:10:18,440
that they're expected to do good
in a particular situation,

209
00:10:19,600 --> 00:10:23,296
which brings us to the third
and final item on our checklist:

210
00:10:23,320 --> 00:10:25,096
to communicate expectations,

211
00:10:25,120 --> 00:10:26,536
to tell people,

212
00:10:26,560 --> 00:10:28,640
"Do the good deed right now."

213
00:10:30,680 --> 00:10:32,936
Here's a simple way
to communicate expectations;

214
00:10:32,960 --> 00:10:36,136
simply tell them, "Hey, everybody else
is doing the good deed."

215
00:10:36,160 --> 00:10:40,336
The company Opower
sends people in their electricity bill

216
00:10:40,360 --> 00:10:43,136
a small insert that compares
their energy consumption

217
00:10:43,160 --> 00:10:46,400
with that of people
with similarly sized homes.

218
00:10:47,000 --> 00:10:50,456
And when people find out that their
neighbors are using less electricity,

219
00:10:50,480 --> 00:10:52,376
they start to consume less.

220
00:10:52,400 --> 00:10:56,056
That same approach, it's been used
to get people to vote or give to charity

221
00:10:56,080 --> 00:10:58,200
or even reuse their towels in hotels.

222
00:11:00,240 --> 00:11:01,440
What about this one?

223
00:11:02,240 --> 00:11:04,736
Here's another way
to communicate expectations;

224
00:11:04,760 --> 00:11:08,840
simply do it by saying, "Do the good deed"
just at the right time.

225
00:11:11,840 --> 00:11:13,040
What about this one?

226
00:11:14,320 --> 00:11:16,936
This ticker reframes

227
00:11:16,960 --> 00:11:19,776
the kind of mundane task
of turning off the lights

228
00:11:19,800 --> 00:11:22,640
and turns it instead
into an environmental contribution.

229
00:11:24,320 --> 00:11:27,136
The bottom line is,
lots of different way to do this,

230
00:11:27,160 --> 00:11:29,216
lots of ways to communicate expectations.

231
00:11:29,240 --> 00:11:30,526
Just don't forget to do it.

232
00:11:31,320 --> 00:11:32,616
And that's it.

233
00:11:32,640 --> 00:11:33,840
That's our checklist.

234
00:11:36,440 --> 00:11:40,776
Many of you are working on problems
with important social consequences

235
00:11:40,800 --> 00:11:44,200
and sometimes you might need
to motivate people to do more good.

236
00:11:45,520 --> 00:11:48,536
The tools you learned today
can help you with this

237
00:11:48,560 --> 00:11:51,736
and these tools, they don't require
that you raise additional funds

238
00:11:51,760 --> 00:11:54,656
or that you develop
any more fancy technologies.

239
00:11:54,680 --> 00:11:57,456
They just require harnessing reputations

240
00:11:57,480 --> 00:12:00,896
by increasing observability,
eliminating excuses

241
00:12:00,920 --> 00:12:02,520
and communicating expectations.

242
00:12:04,120 --> 00:12:05,336
Thank you.

243
00:12:05,360 --> 00:12:09,800
(Applause)