TED Salon Brightline Initiative
Dao Nguyen: What makes something go viral?
Filmed:
Readability: 3.6
1,432,741 views
What's the secret to making content people love? Join BuzzFeed's Publisher Dao Nguyen for a glimpse at how her team creates their tempting quizzes, lists and videos -- and learn more about how they've developed a system to understand how people use content to connect and create culture.
Dao Nguyen - Media analytics expert
As Publisher of BuzzFeed, Dao Nguyen thinks about how media spreads online and the technology and data that publishers can use to understand why. Full bio
As Publisher of BuzzFeed, Dao Nguyen thinks about how media spreads online and the technology and data that publishers can use to understand why. Full bio
Double-click the English transcript below to play the video.
00:12
Last year, some BuzzFeed
employees were scheming
employees were scheming
0
841
3569
00:16
to prank their boss, Ze Frank,
1
4434
2330
00:19
on his birthday.
2
7415
1242
00:21
They decided to put a family
of baby goats in his office.
of baby goats in his office.
3
9319
4204
00:25
(Laughter)
4
13547
1405
00:26
Now, BuzzFeed had recently signed on
to the Facebook Live experiment,
to the Facebook Live experiment,
5
14976
4421
00:31
and so naturally,
6
19421
1371
00:32
we decided to livestream
the whole event on the internet
the whole event on the internet
7
20816
3882
00:36
to capture the moment
when Ze would walk in
when Ze would walk in
8
24722
3300
00:40
and discover livestock in his office.
9
28046
2962
00:44
We thought the whole thing
would last maybe 10 minutes,
would last maybe 10 minutes,
10
32019
3164
00:47
and a few hundred company employees
would log in for the inside joke.
would log in for the inside joke.
11
35207
4524
00:52
But what happened?
12
40132
1263
00:53
They kept on getting delayed:
13
41977
1718
00:55
he went to get a drink,
14
43719
1486
00:57
he was called to a meeting,
15
45229
1448
00:58
the meeting ran long,
16
46701
1743
01:00
he went to the bathroom.
17
48468
1165
01:01
More and more people
started logging in to watch the goats.
started logging in to watch the goats.
18
49657
4220
01:06
By the time Ze walked in
more than 30 minutes later,
more than 30 minutes later,
19
54493
4013
01:10
90,000 viewers were watching
the livestream.
the livestream.
20
58530
5106
01:16
Now, our team had a lot
of discussion about this video
of discussion about this video
21
64601
3957
01:20
and why it was so successful.
22
68582
1976
01:22
It wasn't the biggest live video
that we had done to date.
that we had done to date.
23
70582
2876
01:25
The biggest one that we had done
involved a fountain of cheese.
involved a fountain of cheese.
24
73482
3642
01:30
But it performed so much better
than we had expected.
than we had expected.
25
78063
3825
01:33
What was it about the goats in the office
that we didn't anticipate?
that we didn't anticipate?
26
81912
4263
01:38
Now, a reasonable person could have
any number of hypotheses.
any number of hypotheses.
27
86781
4058
01:42
Maybe people love baby animals.
28
90863
2418
01:45
Maybe people love office pranks.
29
93751
1724
01:47
Maybe people love stories
about their bosses
about their bosses
30
95499
2745
01:50
or birthday surprises.
31
98268
1533
01:52
But our team wasn't really thinking
about what the video was about.
about what the video was about.
32
100382
3700
01:56
We were thinking about
33
104106
1272
01:57
what the people watching the video
were thinking and feeling.
were thinking and feeling.
34
105402
4076
02:01
We read some of the 82,000 comments
that were made during the video,
that were made during the video,
35
109502
4783
02:06
and we hypothesized that they were excited
36
114309
3792
02:10
because they were participating
in the shared anticipation
in the shared anticipation
37
118125
3174
02:13
of something that was about to happen.
38
121323
2730
02:16
They were part of a community,
just for an instant,
just for an instant,
39
124077
3212
02:19
and it made them happy.
40
127313
1353
02:21
So we decided that we needed
to test this hypothesis.
to test this hypothesis.
41
129102
3344
02:24
What could we do to test
this very same thing?
this very same thing?
42
132793
3482
02:28
The following week,
43
136995
1391
02:30
armed with the additional knowledge
that food videos are very popular,
that food videos are very popular,
44
138410
4675
02:35
we dressed two people in hazmat suits
45
143109
3571
02:38
and wrapped rubber bands
around a watermelon until it exploded.
around a watermelon until it exploded.
46
146704
4261
02:42
(Laughter)
47
150989
1581
02:44
Eight hundred thousand people watched
48
152889
3372
02:48
the 690th rubber band
explode the watermelon,
explode the watermelon,
49
156285
4031
02:52
marking it as the biggest
Facebook Live event to date.
Facebook Live event to date.
50
160340
3453
02:56
The question I get most frequently is:
51
164483
2610
02:59
How do you make something go viral?
52
167117
1833
03:01
The question itself is misplaced;
53
169713
2206
03:03
it's not about the something.
54
171943
1693
03:05
It's about what the people
doing the something,
doing the something,
55
173660
3297
03:08
reading or watching --
56
176981
1203
03:10
what are they thinking?
57
178208
1280
03:11
Now, most media companies,
when they think about metadata,
when they think about metadata,
58
179841
3346
03:15
they think about subjects or formats.
59
183211
2383
03:18
It's about goats,
60
186412
1150
03:19
it's about office pranks,
61
187586
1776
03:21
it's about food,
62
189386
1297
03:22
it's a list or a video or a quiz,
63
190707
1854
03:24
it's 2,000 words long,
64
192585
1259
03:25
it's 15 minutes long,
65
193868
1153
03:27
it has 23 embedded tweets or 15 images.
66
195045
3122
03:30
Now, that kind of metadata
is mildly interesting,
is mildly interesting,
67
198191
3012
03:33
but it doesn't actually get at
what really matters.
what really matters.
68
201227
2966
03:36
What if, instead of tagging
what articles or videos are about,
what articles or videos are about,
69
204217
4158
03:40
what if we asked:
70
208399
1190
03:41
How is it helping our users
do a real job in their lives?
do a real job in their lives?
71
209613
3553
03:46
Last year, we started a project
72
214173
2820
03:49
to formally categorize
our content in this way.
our content in this way.
73
217017
2932
03:51
We called it, "cultural cartography."
74
219973
2747
03:55
It formalized an informal practice
that we've had for a really long time:
that we've had for a really long time:
75
223720
4002
03:59
don't just think about the subject matter;
76
227746
2226
04:01
think also about, and in fact,
primarily about,
primarily about,
77
229996
3168
04:05
the job that your content is doing
for the reader or the viewer.
for the reader or the viewer.
78
233188
3739
04:09
Let me show you the map
that we have today.
that we have today.
79
237450
2506
04:11
Each bubble is a specific job,
80
239980
2842
04:14
and each group of bubbles
in a specific color are related jobs.
in a specific color are related jobs.
81
242846
4770
04:19
First up: humor.
82
247640
1478
04:21
"Makes me laugh."
83
249918
1161
04:23
There are so many ways
to make somebody laugh.
to make somebody laugh.
84
251103
2788
04:25
You can be laughing at someone,
85
253915
1517
04:27
you could laugh
at specific internet humor,
at specific internet humor,
86
255456
2155
04:29
you could be laughing at some good,
clean, inoffensive dad jokes.
clean, inoffensive dad jokes.
87
257635
3392
04:33
"This is me." Identity.
88
261706
2300
04:36
People are increasingly using media
to explain, "This is who I am.
to explain, "This is who I am.
89
264030
3808
04:39
This is my upbringing, this is my culture,
90
267862
2199
04:42
this is my fandom,
this is my guilty pleasure,
this is my guilty pleasure,
91
270085
2338
04:44
and this is how I laugh about myself."
92
272447
2183
04:48
"Helps me connect with another person."
93
276600
1882
04:50
This is one of the greatest
gifts of the internet.
gifts of the internet.
94
278506
2411
04:52
It's amazing when you find
a piece of media
a piece of media
95
280941
2191
04:55
that precisely describes
your bond with someone.
your bond with someone.
96
283156
2532
04:59
This is the group of jobs
that helps me do something --
that helps me do something --
97
287020
2598
05:01
helps me settle an argument,
98
289642
1377
05:03
helps me learn something
about myself or another person,
about myself or another person,
99
291043
2673
05:05
or helps me explain my story.
100
293740
1856
05:07
This is the group of jobs
that makes me feel something --
that makes me feel something --
101
295620
2689
05:10
makes me curious or sad
or restores my faith in humanity.
or restores my faith in humanity.
102
298333
3157
05:13
Many media companies
and creators do put themselves
and creators do put themselves
103
301514
3719
05:17
in their audiences' shoes.
104
305257
1468
05:18
But in the age of social media,
we can go much farther.
we can go much farther.
105
306749
2704
05:22
People are connected to each other
on Facebook, on Twitter,
on Facebook, on Twitter,
106
310056
4062
05:26
and they're increasingly using media
to have a conversation
to have a conversation
107
314142
4263
05:30
and to talk to each other.
108
318429
1454
05:31
If we can be a part of establishing
a deeper connection between two people,
a deeper connection between two people,
109
319907
5836
05:37
then we will have done
a real job for these people.
a real job for these people.
110
325767
3189
05:41
Let me give you some examples
of how this plays out.
of how this plays out.
111
329394
2589
05:44
This is one of my favorite lists:
112
332578
1891
05:46
"32 Memes You Should
Send Your Sister Immediately" --
Send Your Sister Immediately" --
113
334493
3303
05:49
immediately.
114
337820
1302
05:51
For example, "When you're going
through your sister's stuff,
through your sister's stuff,
115
339583
3104
05:54
and you hear her coming up the stairs."
116
342711
1979
05:56
Absolutely, I've done that.
117
344714
1790
05:58
"Watching your sister get in trouble
for something that you did
for something that you did
118
346528
3155
06:01
and blamed on her."
119
349707
1192
06:02
Yes, I've done that as well.
120
350923
1608
06:04
This list got three million views.
121
352555
1799
06:06
Why is that?
122
354378
1165
06:07
Because it did, very well, several jobs:
123
355567
2976
06:10
"This is us."
124
358567
1520
06:12
"Connect with family."
125
360111
1265
06:13
"Makes me laugh."
126
361400
1181
06:14
Here are some of the thousands
and thousands of comments
and thousands of comments
127
362605
3134
06:17
that sisters sent to each other
using this list.
using this list.
128
365763
2849
06:21
Sometimes we discover
what jobs do after the fact.
what jobs do after the fact.
129
369508
3511
06:25
This quiz, "Pick an Outfit and We'll Guess
Your Exact Age and Height,"
Your Exact Age and Height,"
130
373592
5138
06:30
went very viral: 10 million views.
131
378754
2063
06:32
Ten million views.
132
380841
1176
06:34
I mean -- did we actually determine
the exact age and height
the exact age and height
133
382041
4526
06:38
of 10 million people?
134
386591
1584
06:40
That's incredible. It's incredible.
135
388199
2262
06:42
In fact, we didn't.
136
390485
1244
06:43
(Laughter)
137
391753
1129
06:44
Turns out that this quiz
went extremely viral
went extremely viral
138
392906
4093
06:49
among a group of 55-and-up women --
139
397023
3563
06:52
(Laughter)
140
400610
1172
06:53
who were surprised and delighted
141
401806
3649
06:57
that BuzzFeed determined
that they were 28 and 5'9".
that they were 28 and 5'9".
142
405479
5718
07:03
(Laughter)
143
411221
1610
07:04
"They put me at 34 years younger
and seven inches taller.
and seven inches taller.
144
412855
3208
07:08
I dress for comfort and do not give
a damn what anyone says.
a damn what anyone says.
145
416087
2948
07:11
Age is a state of mind."
146
419059
1779
07:12
This quiz was successful
not because it was accurate,
not because it was accurate,
147
420862
3050
07:15
but because it allowed these ladies
to do a very important job --
to do a very important job --
148
423936
4054
07:20
the humblebrag.
149
428014
1204
07:22
Now, we can even apply
this framework to recipes and food.
this framework to recipes and food.
150
430478
4983
07:27
A recipe's normal job is to tell you
what to make for dinner or for lunch.
what to make for dinner or for lunch.
151
435887
5573
07:34
And this is how you would normally
brainstorm for a recipe:
brainstorm for a recipe:
152
442205
3408
07:37
you figure out what ingredients
you want to use,
you want to use,
153
445637
2323
07:39
what recipe that makes,
154
447984
1168
07:41
and then maybe you slap a job on
at the end to sell it.
at the end to sell it.
155
449176
3245
07:44
But what if we flipped it around
and thought about the job first?
and thought about the job first?
156
452445
4367
07:49
One brainstorming session
involved the job of bonding.
involved the job of bonding.
157
457730
4665
07:54
So, could we make a recipe
that brought people together?
that brought people together?
158
462976
4300
07:59
This is not a normal brainstorming
process at a food publisher.
process at a food publisher.
159
467300
4421
08:05
So we know that people
like to bake together,
like to bake together,
160
473532
2715
08:08
and we know that people
like to do challenges together,
like to do challenges together,
161
476271
3285
08:11
so we decided to come up with a recipe
that involved those two things,
that involved those two things,
162
479580
4733
08:16
and we challenged ourselves:
163
484337
1911
08:18
Could we get people to say,
164
486272
1806
08:20
"Hey, BFF, let's see
if we can do this together"?
if we can do this together"?
165
488102
4086
08:24
The resulting video was
the "Fudgiest Brownies Ever" video.
the "Fudgiest Brownies Ever" video.
166
492807
3335
08:28
It was enormously successful
in every metric possible --
in every metric possible --
167
496166
2924
08:31
70 million views.
168
499114
1364
08:32
And people said the exact things
that we were going after:
that we were going after:
169
500935
3700
08:36
"Hey, Colette, we need to make these,
are you up for a challenge?"
are you up for a challenge?"
170
504659
3100
08:39
"Game on."
171
507783
1153
08:40
It did the job that it set out to do,
172
508960
2270
08:43
which was to bring people together
over baking and chocolate.
over baking and chocolate.
173
511254
4116
08:49
I'm really excited about
the potential for this project.
the potential for this project.
174
517294
3924
08:53
When we talk about this framework
with our content creators,
with our content creators,
175
521242
3142
08:56
they instantly get it,
176
524408
1165
08:57
no matter what beat they cover,
what country they’re in,
what country they’re in,
177
525597
2733
09:00
or what language they speak.
178
528354
1365
09:01
So cultural cartography has helped us
massively scale our workforce training.
massively scale our workforce training.
179
529743
4377
09:06
When we talk about this project
and this framework
and this framework
180
534621
3650
09:10
with advertisers and brands,
181
538295
1372
09:11
they also instantly get it,
182
539691
1578
09:13
because advertisers,
more often than media companies,
more often than media companies,
183
541293
5626
09:18
understand how important it is
to understand the job
to understand the job
184
546943
3522
09:22
that their products
are doing for customers.
are doing for customers.
185
550489
2446
09:26
But the reason I'm the most excited
about this project
about this project
186
554580
3743
09:30
is because it changes the relationship
between media and data.
between media and data.
187
558347
3871
09:35
Most media companies
think of media as "mine."
think of media as "mine."
188
563051
3434
09:39
How many fans do I have?
189
567547
1207
09:40
How many followers have I gained?
190
568778
1597
09:42
How many views have I gotten?
191
570399
1707
09:44
How many unique IDs do I have
in my data warehouse?
in my data warehouse?
192
572130
2777
09:47
But that misses the true value of data,
which is that it's yours.
which is that it's yours.
193
575372
3948
09:53
If we can capture in data
what really matters to you,
what really matters to you,
194
581166
6098
09:59
and if we can understand more
the role that our work plays
the role that our work plays
195
587288
4488
10:03
in your actual life,
196
591800
1389
10:05
the better content we can create for you,
197
593213
2863
10:08
and the better that we can reach you.
198
596100
1857
10:10
Who are you?
199
598760
1275
10:13
How did you get there?
200
601293
1580
10:14
Where are you going?
201
602897
1209
10:16
What do you care about?
202
604130
1269
10:17
What can you teach us?
203
605423
1379
10:19
That's cultural cartography.
204
607151
1460
10:21
Thank you.
205
609146
1156
10:22
(Applause)
206
610326
3380
ABOUT THE SPEAKER
Dao Nguyen - Media analytics expertAs Publisher of BuzzFeed, Dao Nguyen thinks about how media spreads online and the technology and data that publishers can use to understand why.
Why you should listen
Dao Nguyen is the Publisher of BuzzFeed, a reinvention of the traditional title in which she oversees the company’s tech, product, data and publishing platform, as well as ad product, pricing, and distribution. Nguyen joined BuzzFeed in 2012 and has been instrumental in its rapid growth as the largest independent digital media company in the world. Prior to joining BuzzFeed, Nguyen oversaw product for a financial careers venture within Dow Jones. She also previously served as Chief Executive Officer of Le Monde Interactif, publisher of the leading news site lemonde.fr. Before moving to France, she was Executive Producer at Concrete Media, a small web agency, and a consultant at Andersen Consulting (now Accenture). She has a degree in Applied Mathematics / Computer Science from Harvard and is based in New York City.
Dao Nguyen | Speaker | TED.com