Lecture 7 - How to Build Products Users Love (Kevin Hale)
-
0:01 - 0:01All right.
-
0:01 - 0:06So when I talk about making
products users love what I -
0:06 - 0:10mean specifically is like
how do we make things that -
0:10 - 0:13has a passionate user base
that our users -
0:13 - 0:17are unconditionally wanting
it to -
0:17 - 0:20be successful both on the
products that we build, but -
0:20 - 0:21also the companies behind
them. -
0:22 - 0:26We're gonna go over tons of
information. -
0:26 - 0:27Try not to take too many
notes, -
0:27 - 0:28mostly just try to listen.
-
0:28 - 0:31I'll post the link to the
slides on my Twitter account -
0:31 - 0:33and on that link, there will
be a way for you to -
0:33 - 0:36annotate the slide and you
can ask me questions. -
0:36 - 0:37And so if we don't get to
them, -
0:37 - 0:38I'll answer them after the
talk. -
0:39 - 0:41So, you guys have been.
-
0:41 - 0:41Listening to and
-
0:41 - 0:45hearing a lot about growth
over the last several weeks, -
0:45 - 0:48and to me I feel like growth
is usually fairly simple. -
0:48 - 0:50It's the interaction between
two sort of concepts or -
0:50 - 0:51variables.
-
0:51 - 0:54Conversion rate and churn.
-
0:54 - 0:55Right?
-
0:55 - 0:58And the gap between those
two things pretty -
0:58 - 1:01much indicate how fast
you're gonna grow. -
1:01 - 1:04And most people, especially
business type people, -
1:04 - 1:06tend to look at this
interaction in terms of -
1:06 - 1:09a very calculated in a
mathematical sort of way. -
1:09 - 1:11And today, I sort of want to
-
1:11 - 1:12talk about these things at a
more human scale. -
1:12 - 1:15All right, cuz at start up
when you're interacting with -
1:15 - 1:17your users, you have a
fairly intimate -
1:17 - 1:19interaction that you have in
the early stages. -
1:19 - 1:21And so, I think there's a
different way of looking at -
1:21 - 1:23this stuff in terms of how
we build our products. -
1:23 - 1:26And we'll look at a lot of
different examples of that. -
1:26 - 1:26And how it's executed well.
-
1:28 - 1:31My philosophy behind a lot
of things that I teach -
1:31 - 1:33startups is the best way to
sort of get to a billion -
1:33 - 1:36dollars is to focus on the
values that help you -
1:36 - 1:39get that first dollar, to
acquire that first user. -
1:39 - 1:42If you sort of get that
right everything else will -
1:42 - 1:43sort of take care of itself.
-
1:43 - 1:45It's a sort of faith thing.
-
1:45 - 1:50So I came to be a partner at
YC by way of being alumni. -
1:50 - 1:52I went to the program in
winter of 2006, so -
1:52 - 1:54it's the second ever program
and -
1:54 - 1:55I built a product called
Wufoo. -
1:55 - 1:58Wufoo's an online forum
builder. -
1:58 - 1:59It helps you create contact
forums and -
1:59 - 2:01online surveys and simple
payment forms. -
2:01 - 2:03It's basically a database
that looks like it's -
2:03 - 2:06designed by Fisher Price.
-
2:06 - 2:09Interesting, though, is that
because it was fairly easy -
2:06 - 2:06What's.
-
2:09 - 2:13to use, we're had customers
from every industry, market, -
2:13 - 2:16and vertical you can think
of, including a, -
2:16 - 2:18a majority of the Fortune
500 companies out there. -
2:18 - 2:21Ran the company for five
years, and -
2:21 - 2:25then we were acquired by
SurveyMonkey in 2013. -
2:25 - 2:26And at the time,
-
2:26 - 2:28we were a very interesting
acquisition. -
2:28 - 2:30We were only a team of ten
people at the time. -
2:30 - 2:33And, while we acquired
funding out here in -
2:33 - 2:36Silicon Valley through Y
Combinator we actually ran -
2:36 - 2:37the company from Florida.
-
2:37 - 2:39We had no office.
-
2:39 - 2:40Everyone worked from home.
-
2:42 - 2:44And we're an interesting
outlier. -
2:44 - 2:48So, each dot here represents
a start up that was, -
2:48 - 2:50that exited through IPO or
acquisition. -
2:50 - 2:52And we're this outlier to
the left. -
2:52 - 2:54The bottom is the funding
amount they took. -
2:54 - 2:56And the vertical axis is,
-
2:56 - 2:57the valuation of the company
at the time. -
2:57 - 2:58To sum it up,
-
2:58 - 3:02the average startup rates as
about $25 million. -
3:02 - 3:05And they return to their
investors about 676%. -
3:05 - 3:09Wufoo raised about $118
thousand total. -
3:09 - 3:14And our return to our
investors is about 29,000%. -
3:14 - 3:15>> So a lot of people were
very interested. -
3:15 - 3:16And are sort of like.
-
3:16 - 3:16What makes.
-
3:16 - 3:18Wufoo a little bit different
or -
3:18 - 3:20how do we run the company
very differently, and -
3:20 - 3:21a lot of it was focused on
product. -
3:23 - 3:25We weren't interested in
building software that, -
3:25 - 3:29I guess that just people
wanted to use. -
3:29 - 3:31Right.
That reminding you that you -
3:31 - 3:31worked in a cubicle,
-
3:31 - 3:34cuz it was database app at
its sort of core. -
3:34 - 3:37We wanted a product that
people wanted to love. -
3:37 - 3:39That people wanted to have a
relationship with, and we -
3:39 - 3:42were actually very fanatical
about how we approached. -
3:42 - 3:44This idea.
-
3:44 - 3:45At the point where it's
almost sort of -
3:47 - 3:49sort of sciency sort of way.
-
3:50 - 3:52So what we said was like
Okay. -
3:52 - 3:55What's interesting about
start-ups in terms of -
3:55 - 3:59us wanting to create things
that people love is that, -
3:59 - 4:01love and unconditional sort
of feelings are things that -
4:01 - 4:03are difficult for us to do
in sort of real life. -
4:04 - 4:07And at start-ups we have to
do it sort of at scale. -
4:07 - 4:08So we've decided to do is
-
4:08 - 4:10just start of by just
looking at, Okay, -
4:10 - 4:13how does real relationships
work in the real world, and -
4:13 - 4:14how can we apply them to
sort of how we -
4:14 - 4:18run our business and sort of
build our product that way. -
4:18 - 4:20So we'll go over basically
these two metaphors, -
4:20 - 4:22find new users as if we're
trying to date them and -
4:22 - 4:26existing users as if it were
a successful marriage. -
4:26 - 4:28So when it comes to dating,
lot of the stuff that -
4:28 - 4:32we uncovered had to do with
first impressions. -
4:33 - 4:36all, all of you often talk
about your relationships in -
4:36 - 4:37terms of the origin story.
-
4:37 - 4:39If I asked you to tell me
about the first kiss, -
4:39 - 4:41how you sort of met, how you
proposed, -
4:41 - 4:43these are the things that we
say over and -
4:43 - 4:43over and over again.
-
4:43 - 4:46They're basically the
word-of-mouth stories of our -
4:46 - 4:48relationships, and they're
the same kinds of -
4:48 - 4:50things that we do with
companies. -
4:50 - 4:51Human beings
-
4:51 - 4:52are relationship
manufacturing creatures. -
4:52 - 4:56We cannot help, but create
and anthropomorphize -
4:56 - 5:00the things we interact with
over and over again. -
5:00 - 5:01So, whether it's the cars we
drive, -
5:01 - 5:04the clothes we wear, the
tools and software's we use, -
5:04 - 5:06we eventually prescribe
characteristics to it. -
5:06 - 5:07A personality.
-
5:07 - 5:09And we expect it to behave a
certain way. -
5:09 - 5:11And that's how we sort of
interact with it. -
5:11 - 5:13Now first impressions are
important for -
5:13 - 5:15the starting of any
relationship, -
5:15 - 5:16because it's the one we tell
over and over again. -
5:16 - 5:17Right?
-
5:17 - 5:20And there's something
special about how we -
5:20 - 5:21regard that origin story.
-
5:21 - 5:22I'll give an example.
-
5:22 - 5:24If you're on a first date
with somebody, -
5:24 - 5:27and you're having a nice
dinner, and -
5:27 - 5:29you catch them picking their
nose. -
5:29 - 5:32You are probably not gonna
have another date with them. -
5:32 - 5:33Right, but if you're married
to someone for -
5:33 - 5:34about 20 to 30 years,
-
5:34 - 5:38and you catch on the Barca
Lounger digging for -
5:38 - 5:39gold right?
-
5:39 - 5:40You don't immediately like
call your lawyer, -
5:40 - 5:43right, and then say like we
have a problem here. -
5:43 - 5:45I need to start drawing up
papers for divorce. -
5:45 - 5:47You shrug your shoulders and
-
5:47 - 5:49say at least he has a heart
of gold. -
5:50 - 5:51So solving about.
-
5:51 - 5:55First time interactions
means that the threshold is -
5:55 - 5:58so much lower in terms of
pass fail. -
5:58 - 6:01So, in software, and for
-
6:01 - 6:03most products in internet
software that we use, like, -
6:03 - 6:05first impressions are pretty
obvious. -
6:05 - 6:06And they're the things that
you see a lot of -
6:06 - 6:08companies sort of pay
attention to in terms of -
6:08 - 6:10what they send their
marketing people to work on. -
6:10 - 6:14My argument for people who
are very good at product is -
6:14 - 6:16they discover so many other
first moments, and -
6:16 - 6:19they make those something
memorable. -
6:19 - 6:22Right, the first email you
ever get from a piece of -
6:21 - 6:22software.
-
6:22 - 6:25What happens when you first
log in? -
6:25 - 6:26The links, the advertising,
the very first time you -
6:26 - 6:28interact with customer
support. -
6:28 - 6:32All of those are
opportunities to seduce. -
6:32 - 6:34So how did we think about.
-
6:34 - 6:36Sort of like making first
moments on there, -
6:36 - 6:39and we actually took this
concept from the Japanese. -
6:39 - 6:41They actually have two words
for -
6:41 - 6:43how to describe things when
you're finished with them -
6:43 - 6:46in terms of saying like is
this a quality item. -
6:46 - 6:47And the two words of quality
are atarimae hinshitsu and -
6:47 - 6:48miryokuteki hinshitsu.
-
6:48 - 6:51And the first one means,
taken for -
6:51 - 6:53granted quality, basically
functionality. -
6:53 - 6:55And the last one sort of
means, -
6:55 - 6:57enchanting quality, right?
-
6:58 - 7:00Take for example a pen,
right? -
7:00 - 7:04Something has miryokuteki,
right, if the weight of -
7:04 - 7:06the pen, the way the ink
flows out of it, the way -
7:06 - 7:09it's viewed by the people
reading the handwriting from -
7:09 - 7:12the pen is pleasurable both
to the user of the pen and -
7:12 - 7:15the people who experience
the byproducts of it, -
7:15 - 7:17right, taking it to the sort
of next level. -
7:19 - 7:20Start with some examples.
-
7:20 - 7:23So this is Wufoo's Login
link, and -
7:23 - 7:26it has a dinosaur on it,
which I think is awesome. -
7:26 - 7:30But if you hover over it,
-
7:30 - 7:33the spec has the added
benefit of having a tool tip -
7:33 - 7:34that doesn't explain like
how to login or -
7:34 - 7:36what it does, but basically,
rawr. -
7:36 - 7:38And.
-
7:38 - 7:40What we noticed about this,
like in early usability -
7:40 - 7:42studies as like this put a
smile on people's faces. -
7:42 - 7:44Like, hands down.
-
7:44 - 7:44Right?
-
7:44 - 7:45Universally.
-
7:45 - 7:47And I think a lot of times
when we -
7:47 - 7:49are assessing products, we
never think about, -
7:49 - 7:50like, hey what is the
emotion on -
7:50 - 7:52the person's face when they
interact with this? -
7:54 - 7:56This is Vimeo's log in page
this is -
7:56 - 7:58actually a couple integr,
iterations ago. -
7:58 - 8:00It's one I find to be the
most beautiful, but. -
8:01 - 8:02It lets you know that when
you're starting out on -
8:02 - 8:03this journey with Vimeo,
-
8:03 - 8:05that this is gonna be
something different. -
8:06 - 8:07They do this all over the
app. -
8:07 - 8:10If you search for the word
fart, as you scroll up and -
8:10 - 8:12down it makes fart noises as
you do this. -
8:12 - 8:13Right?
-
8:13 - 8:14There's something different,
-
8:14 - 8:16like this site interacts
with you. -
8:16 - 8:16>> Mm-hm.
>> It's a little bit -
8:16 - 8:18magical, it's a little bit
different. -
8:18 - 8:20And it's something that you
wanna talk about. -
8:21 - 8:24You don't have to always do
it with design. -
8:24 - 8:25This is a sign up form for
-
8:25 - 8:26Cork which used to be a
social network for -
8:26 - 8:29people who loved to drink
wine. -
8:29 - 8:31On it, it says email
address, -
8:31 - 8:33it's also your sign-in and
has to be legit. -
8:33 - 8:36First name, what your mom
calls you, last name, -
8:36 - 8:37what your Army buddies call
you. -
8:37 - 8:40Password, something you'll
remember, but hard to guess, -
8:40 - 8:42password confirmation, think
it a, type it again. -
8:42 - 8:44Think of it as a test.
-
8:44 - 8:47It's literally a poem as you
fill out the form. -
8:47 - 8:48Right?
-
8:48 - 8:51And this is a kind of like
thing where you like, -
8:51 - 8:52oh, I like the people behind
this. -
8:52 - 8:54I, I, I'm gonna enjoy this
experience. -
8:54 - 8:56Now what does it say when
you fill out form like this? -
8:56 - 8:56Right.
-
8:56 - 8:59On Yahoo, about what the
personality. -
8:59 - 9:00Of the site it's gonna be.
-
9:00 - 9:02And what's disappointing to
me is like, -
9:02 - 9:05Yahoo forces every product
and -
9:05 - 9:08service under them to use
this exact same login form. -
9:08 - 9:09Flickr, I had thought,
-
9:09 - 9:11had one of the best sort of
call to actions. -
9:11 - 9:14It was, get in there.
-
9:14 - 9:14Right?
-
9:16 - 9:17This is Heroku's.
-
9:20 - 9:21I think this is an older
version. -
9:20 - 9:20Sign up page.
-
9:21 - 9:24But, what's remarkable about
it is that, what you -
9:24 - 9:28start getting a feel for is
like, oh, scaling up. -
9:28 - 9:29My sort of server, and
-
9:29 - 9:33back end services is as easy
as sort of dragging up and -
9:33 - 9:36down, different sort of nobs
and levers. -
9:36 - 9:38It's gonna be beautifully
used, and -
9:38 - 9:40it looks fairly easy to
scale. -
9:41 - 9:45Since we're in a room full
of computer science people, -
9:45 - 9:47I think you'll appreciate
this. -
9:47 - 9:48This is Chocolat.
-
9:48 - 9:50This is a code editor, and
-
9:50 - 9:51they only have one call to
action. -
9:51 - 9:55When the time limit is up,
they say, everything in -
9:55 - 9:58terms of all the pieces are
exactly the same except we -
9:58 - 10:00change the font.
-
10:00 - 10:01To comic sans.
-
10:04 - 10:07And what they're basically
saying is like hey -
10:07 - 10:10we know who our users are,
who our real customers are. -
10:10 - 10:12They're gonna be the people
who care about this. -
10:15 - 10:17This is Hurl, this is a
website for -
10:17 - 10:23checking htp requests, and
sometimes the places where -
10:23 - 10:25you get errors are
opportunities for -
10:25 - 10:27first moments,
-
10:27 - 10:29you had a four of four this
is what you get it, -
10:29 - 10:32when we need help oftentimes
what we do is where we -
10:32 - 10:40create like really beautiful
mark, marketing materials. -
10:40 - 10:43But when you actually need,
like, documentation we sort -
10:43 - 10:46of like skimp out on sort of
design features. -
10:46 - 10:49And this is a point that,
like, -
10:49 - 10:50you see happen over and over
again. -
10:50 - 10:52A company that gets this
right is MailChimp. -
10:52 - 10:53And what they did was they
redesigned all of their -
10:53 - 10:56help guides so that they
looked like magazine covers. -
10:56 - 10:58And overnight, basically,
-
10:58 - 11:00readership goes up on all
these features. -
11:00 - 11:03And customer support for
these things that sort of -
11:03 - 11:04help people optimize emails
goes down. -
11:06 - 11:08Speaking of documentation,
stripe. -
11:10 - 11:14What's interesting about an
API company is that there is -
11:14 - 11:14no UX.
-
11:14 - 11:17The UX is actually just
documentation, right? -
11:17 - 11:19And there's opportunities
even in documentation. -
11:19 - 11:21Sort of the enchant and
amaze. -
11:21 - 11:22So one of the things that I
-
11:22 - 11:24love about them is their
examples are wonderful. -
11:24 - 11:25But if you're actually,
like, -
11:25 - 11:27sort of logging into the
app, -
11:27 - 11:29one of the things that is a
super pain for most people, -
11:29 - 11:31when you're dealing with
most people's APIs, is, -
11:31 - 11:34like, grabbing your API
credentials and keys. -
11:34 - 11:36And what Stripe does is it
says oh, -
11:36 - 11:37if you're logged into the
app. -
11:37 - 11:37We automatically put
-
11:37 - 11:40your API credentials into
the examples, so you only -
11:40 - 11:44have to copy paste once when
trying to learn their API. -
11:48 - 11:50When Wufoo wanted to launch
the third version of our API -
11:50 - 11:52we realized like, Okay, that
finally this is good enough -
11:52 - 11:55that we want people to sort
of build on top of it. -
11:55 - 11:55We were trying to figure
out, -
11:55 - 11:57like, how do we launch this
out to the world, -
11:57 - 11:59that sort of has our
personality behind it. -
11:59 - 11:59Because a lot of people,
-
11:59 - 12:02they usually do things like
a programming API contest, -
12:02 - 12:04and they give out iPads and
iPhones. -
12:04 - 12:06And it makes you look like
everyone else. -
12:06 - 12:09And so, in our company one
weird value to -
12:09 - 12:11have it's a quirk of us,
-
12:11 - 12:14is that the co-founders are
big medieval nuts. -
12:14 - 12:17And we take everyone out to
Medieval Times every single -
12:17 - 12:19year, on the anniversary of
the founding of the company. -
12:19 - 12:22And so, we said we have to
do something in that flavor. -
12:22 - 12:24And so we contacted the guys
at armor.com, -
12:24 - 12:27we said can you forward us a
custom battle-ax. -
12:27 - 12:28And what we said was if
-
12:28 - 12:32you win our programming
contest you would win one. -
12:33 - 12:36The result is like people
wanted to talk about this. -
12:36 - 12:37It's something that people
wanted to work on cuz -
12:37 - 12:40they wanted to be able to
say it like I'm programming. -
12:40 - 12:40For a weapon.
-
12:43 - 12:47And what's cool is we had
over 25 different -
12:47 - 12:51applications created for us
of quality and quantity that -
12:51 - 12:53we could not have paid for
on the budget and -
12:53 - 12:55the sort of time that we had
for this. -
12:55 - 12:56We got things like an iPhone
app and -
12:56 - 12:59Android and WordPress
plugins, right? -
12:59 - 13:01And all because what we did
was we changed how -
13:01 - 13:03people wanna talk about the
origin story of -
13:03 - 13:05how they're interacting with
one of our services. -
13:06 - 13:09And go like all day long of
going over these examples. -
13:09 - 13:11But I'm gonna short cut this
by saying, -
13:11 - 13:13you should just subscribe a
little bit of details. -
13:13 - 13:16It's just basically tons of
screen shots of software -
13:16 - 13:17that's just doing it right.
-
13:17 - 13:18That shows that they're
being -
13:18 - 13:22conscientious of the user
and the customers. -
13:28 - 13:31When it comes to long term
relationships or -
13:31 - 13:34marriages, the only research
that we ended up having to -
13:34 - 13:37read was the stuff that was
done by John Gottman. -
13:37 - 13:40He's been featured in This
American Life, -
13:40 - 13:41in Malcolm Gladwell's books.
-
13:41 - 13:43He's a marriage researcher
up in Seattle. -
13:43 - 13:45And he has an interesting
parlor trick that he can do. -
13:45 - 13:49He can watch a video tape of
a couple fighting about some -
13:49 - 13:50issue for 15 minutes.
-
13:50 - 13:53And predict with an 85%
accuracy rate whether that -
13:53 - 13:54couple will be together or
-
13:54 - 13:56not, or divorced in four
years. -
13:57 - 13:59If he increases that video
up to an hour. -
13:59 - 14:01And asks them to also talk
about their hopes and -
14:01 - 14:02dreams.
-
14:02 - 14:03That prediction rating goes
up to 94%. -
14:03 - 14:06They showed these same video
tapes to marriage -
14:06 - 14:08counselors, successfully
married couples, -
14:08 - 14:12sociologists, psychiatrists,
priests, et cetera and -
14:12 - 14:15they can't predict with
random chance whether people -
14:15 - 14:16are gonna be together or
not. -
14:16 - 14:19So John Gauntman understands
something fundamental about. -
14:20 - 14:22How relationships work in
the long term. -
14:22 - 14:26And that basically how we
fight, even in a short term -
14:26 - 14:29period can indicate sort of
the whole system and -
14:29 - 14:30what it's gonna look like.
-
14:30 - 14:32And one of the surprising
things he discovered, -
14:32 - 14:32it's not that
-
14:32 - 14:34successfully married people
don't fight at all. -
14:34 - 14:36It turns out everybody
fights. -
14:36 - 14:40And we all fight about the
exact same things money, -
14:40 - 14:41kids, sex, time and others.
-
14:41 - 14:45And others are things like
jealousy and in-laws. -
14:45 - 14:48To bring this around, right,
you can actually attribute -
14:48 - 14:50every single one of these to
problems that you see in -
14:50 - 14:52customer support when you're
building out your products. -
14:53 - 14:57Right, so.
-
14:57 - 14:58This costs too much.
-
14:58 - 15:00I'm having problems with a
credit card. -
15:00 - 15:02If you're building a service
that helps people deal with -
15:02 - 15:03their clients they're very
-
15:03 - 15:05sensitive about anything
happening with that. -
15:05 - 15:07Performance.
-
15:07 - 15:09How long you're up and how
fast. -
15:11 - 15:14Others are I said jealousy
and in-laws. -
15:14 - 15:16Right, so that's competition
and partnerships. -
15:16 - 15:17So anything weird happening
there, -
15:17 - 15:20people are gonna write to
you about. -
15:20 - 15:21And the reason I like to
think about this in -
15:21 - 15:23terms of customer support is
that in every ones that -
15:23 - 15:26are processing of like a
conversion funnel, -
15:26 - 15:28customer support is the
thing that happens in -
15:28 - 15:30between every one of these
steps. -
15:29 - 15:31It's the reason why people
don't make it -
15:31 - 15:32further down there.
-
15:32 - 15:34It's the thing that prevents
conversion from happening. -
15:35 - 15:38Now, as we were thinking
through all these ideas and -
15:38 - 15:40as we were building up the
company, -
15:40 - 15:42we realized that there's a
big problem about how -
15:42 - 15:44everyone sort of starts
their company or -
15:44 - 15:46build up their sort of
engineering teams. -
15:46 - 15:49And that is that there's a
broken feedback loop there. -
15:49 - 15:51People are divorced from the
consequences of -
15:51 - 15:52their actions.
-
15:52 - 15:54And this is the result of
actually the natural -
15:54 - 15:56evolution of how most
companies get founded, -
15:56 - 15:58especially by technical
co-founders, right. -
15:58 - 16:02Before launch it is a time
of bliss, nirvana, and -
16:02 - 16:04opportunity, right?
-
16:04 - 16:05Nothing that you do is
wrong, right? -
16:07 - 16:10By your hand, which you feel
is like god, everything that -
16:10 - 16:13you write, every line of
code feels perfect, right, -
16:13 - 16:15and is ingenious to you.
-
16:15 - 16:18The thing that happens is
after launch reality sort of -
16:18 - 16:20sets in and then all these
other tasks sort of -
16:20 - 16:22come into place that we have
to deal with. -
16:23 - 16:26Now what technical
co-founders wanna do is get -
16:26 - 16:29back to that initial state
and so what we often do, -
16:29 - 16:31and what we often see is
that companies start -
16:31 - 16:34siloing off all these other
things that actually is -
16:34 - 16:37what makes a start up or a
company sort of real, right? -
16:37 - 16:40And have other people do
them. -
16:40 - 16:43To, in our minds, these
other tasks are inferior. -
16:43 - 16:44Right?
And we have other people in -
16:44 - 16:44the company do them.
-
16:46 - 16:48And so for us, what we're
trying to figure out -
16:48 - 16:50is how do we change software
development so that we -
16:50 - 16:54inject some values that we
don't talk about enough? -
16:54 - 16:56Responsibility,
accountability, -
16:56 - 16:58humility, modesty.
-
16:58 - 17:01Right, and we called this,
like a lot of other people, -
17:01 - 17:06we had an acronym, Support
Driven Development, and -
17:06 - 17:10it's very similar to TDD or
other agile practice. -
17:10 - 17:13It's a way of creating high
quality software, but -
17:13 - 17:14it's super simple.
-
17:14 - 17:15You don't need like, a
SCRUM, -
17:15 - 17:18you don't need a bunch of
post-it notes, all you -
17:18 - 17:20have to do is make everyone
do customer support. -
17:22 - 17:23And what you end up having
is -
17:23 - 17:24you fix the feedback loop,
right? -
17:24 - 17:25The people who build the
software are the ones -
17:25 - 17:26supporting it and
-
17:26 - 17:29you get all these sort of
nice benefits as a result. -
17:30 - 17:33So one of them is, support
responsible developers and -
17:33 - 17:35designers and people who
build the stuff, -
17:35 - 17:37they give the very best
support. -
17:37 - 17:39Now we're not the first
person to think of this. -
17:39 - 17:41Paul English was a big
proporter of this at Kayak, -
17:41 - 17:45and what he did was install
a red customer support phone -
17:45 - 17:47line in the middle of the
engineering floor. -
17:47 - 17:50And they were just regular
customer support calls. -
17:50 - 17:54And people would ask him
often times, -
17:54 - 17:56why would you pay engineers
$120,000 or -
17:56 - 17:59more to do something that
you can pay other people -
17:59 - 18:03a fraction of to handle in
like a call center? -
18:03 - 18:05And he says, well, after the
second or -
18:05 - 18:07third time that phone rings
and the engineer gets -
18:07 - 18:10the same problem they stop
what they're doing, they fix -
18:10 - 18:14the bug, and we stop getting
phone calls about it. -
18:14 - 18:17It, it's a way of having Q-A
in a sort of -
18:17 - 18:18nice elegant solution.
-
18:20 - 18:24Now, John Gotman talks about
the reason that we often -
18:24 - 18:27break up with one another as
due to four major causes and -
18:27 - 18:27their warning signs.
-
18:27 - 18:29He calls them the four
horsemen, right? -
18:29 - 18:33Criticism, contempt,
defensiveness, -
18:33 - 18:34and stonewalling.
-
18:34 - 18:39Now, criticism is basically
people starting to focus, -
18:39 - 18:40not just on the specific
issue at hand, but -
18:40 - 18:41on the overarching issues.
-
18:41 - 18:46Like, you never, right,
listen to your users or -
18:46 - 18:48you never think about us all
the time. -
18:48 - 18:49Right?
-
18:50 - 18:54Contempt is when someone is
purposely trying to -
18:54 - 18:55insult somebody.
-
18:55 - 19:00Defensiveness is not trying
to take accountability -
19:00 - 19:02trying to make excuses for
the actions. -
19:02 - 19:04And stonewalling is
basically shutting down. -
19:04 - 19:07Stonewalling, to John
Gotman, is, is one of -
19:07 - 19:09the worst things that we can
do in a relationship. -
19:09 - 19:11Hold up.
-
19:15 - 19:18And oftentimes, you know, we
don't worry much about this -
19:18 - 19:20in customer se, criticism
and contempt. -
19:20 - 19:21Right?
Defensiveness, -
19:21 - 19:23you see this all the time,
all the times in companies, -
19:23 - 19:24especially as they get
older. -
19:24 - 19:25But stonewalling, this is
something I -
19:25 - 19:27see happen with start-ups
all the time. -
19:29 - 19:30You get a bunch of customer
support sort of -
19:30 - 19:32coming in and you just think
I don't need to answer it. -
19:32 - 19:33I don't need to respond.
-
19:33 - 19:34Right?
And -
19:34 - 19:36that act of just not even
getting back to -
19:36 - 19:38them is one of the worst
things you can do. -
19:38 - 19:40And it's probably some of
the biggest causes of -
19:40 - 19:42churn in the early stages of
start-ups. -
19:44 - 19:45This is how support worked
out with Wufoo. -
19:45 - 19:48When we were acquired we had
about 500,000 users on -
19:48 - 19:51the system, 5 million people
used Wufoo forms and -
19:51 - 19:53reports, whether they knew
it or not, and -
19:53 - 19:56all those people got support
from the same ten people, -
19:56 - 19:57and usually it was
-
19:57 - 20:00only one person dedicated
support a day, or any shift. -
20:00 - 20:02Results in about 400 issues
a week. -
20:02 - 20:04That's about 800 emails.
-
20:04 - 20:06But a response time from 9
a.m. -
20:06 - 20:06to 9 p.m.
was -
20:06 - 20:08between seven to 12 minutes.
-
20:08 - 20:09Right?
And from 9 p.m. -
20:09 - 20:11to midnight it was an hour.
-
20:11 - 20:13And on the weekend it would
be no longer than 24 hours. -
20:13 - 20:16And we carried this up all
the way up to the scale. -
20:18 - 20:20What a lot of people forget
about, and -
20:20 - 20:23often talk about with
Airbnb, is how, like oh, -
20:23 - 20:25they did this interesting
thing where they had, -
20:25 - 20:27went up to New York and
offered, like, -
20:27 - 20:28professional photographers,
and the founders would go -
20:28 - 20:30out there and actually take
pictures of -
20:30 - 20:32the people's apartments to
help them sell more. -
20:32 - 20:34Focusing on the stories
around conversion. -
20:34 - 20:36What most people don't
realize is a lot of -
20:36 - 20:39times when I saw Joe in the
early days of Airbnb, -
20:39 - 20:41he had a phone, sort of head
set stuck to his head all -
20:41 - 20:45the time, because he was
doing phone support nonstop. -
20:46 - 20:48Churn is a story we don't
like to talk about often, -
20:48 - 20:49all the time.
-
20:49 - 20:53Airbnb's sort of growth
really started picking up -
20:53 - 20:53once they figured out
-
20:53 - 20:56how to match capacity to the
demand, or the phone -
20:56 - 20:59calls that they were getting
into their support system. -
21:01 - 21:03At Wufoo we actually
constantly did experiments -
21:03 - 21:06around support, because
we're so obsessed with it. -
21:06 - 21:09One experiment we did was,
we heard Kathy Sierra -
21:09 - 21:11do a talk about there's a
disconnect between the -
21:11 - 21:15motions that we have when we
need help, and sort of. -
21:15 - 21:17The content and the
reactions we get from -
21:17 - 21:18people when we get help from
them, especially online. -
21:18 - 21:23Because they just don't see
all those nonverbal cues. -
21:23 - 21:25So she said unless there's
face recognition on the web, -
21:25 - 21:26we're just always going to
be -
21:26 - 21:28disconnected from our users.
-
21:28 - 21:29Our feeling was,
-
21:29 - 21:31like, well we're not face
recognition experts but -
21:31 - 21:33I think there's another way
of getting empathy. -
21:33 - 21:36So, as form builders, we
added a drop-down and -
21:36 - 21:38what we said was like, hey,
what's your emotional state? -
21:40 - 21:41And our hypothesis was
-
21:41 - 21:42that no one was gonna fill
this out. -
21:42 - 21:46We basically thought, oh
okay, you know what the, -
21:46 - 21:49this is gonna be pretty a
lame experiment, but -
21:49 - 21:50we'll see how it sort of
goes. -
21:50 - 21:51And it turned out the
Emotional State -
21:51 - 21:54drop-down field was filled
out 75.8% of the time. -
21:55 - 21:57The browser type drop-down
filled just in -
21:57 - 22:00comparison was filled out
78.1% of the time. -
22:00 - 22:01All right?
-
22:01 - 22:03So people were basically
telling us, for -
22:03 - 22:07my technical support issue
how I feel about this -
22:07 - 22:10problem is just as important
as, like, all the technical -
22:10 - 22:13details you need to sorta
figure out how to debug it. -
22:14 - 22:17Now we didn't prioritize
things or -
22:17 - 22:18triage things by emotion,
right, and for -
22:18 - 22:20the most part, people didn't
game the system. -
22:20 - 22:23One of the interesting
byproducts of it was that we -
22:23 - 22:24noticed that people started
being nicer to -
22:24 - 22:26us in the customer support.
-
22:26 - 22:27It was something sort of
subconscious. -
22:27 - 22:29We just were thinking like,
wow, -
22:29 - 22:30users are so much better
now. -
22:30 - 22:31What's going on?
-
22:31 - 22:34And we went back and looked
at the data and -
22:34 - 22:37we did some text analysis
and we realized is that, oh, -
22:37 - 22:40when it comes to only
communicating with -
22:40 - 22:44people over written words,
like email, there's -
22:44 - 22:46only three ways that you
show strong emotions, right? -
22:46 - 22:50Exclamation marks curse
words, and all caps. -
22:50 - 22:52And sure enough on all three
of those metrics, -
22:52 - 22:53they've gone down.
-
22:53 - 22:56In sort of the way people
were talking to us in -
22:56 - 22:56the customer support.
-
22:56 - 22:59Once people had a simple
outlet for their emotions, -
22:59 - 23:00right, people would be a lot
more rational, -
23:00 - 23:03and it made our jobs a lot
more pleasant as a result. -
23:03 - 23:09The other byproduct that is
awesome is that -
23:09 - 23:11you actually build better
software when you do this. -
23:11 - 23:12Far better software.
-
23:12 - 23:15This is actually backed up
by tons of research. -
23:15 - 23:16Jared Spool, a user
interface engineer, which is -
23:16 - 23:19sort of the big players in
this space says like, -
23:19 - 23:21there's a direct correlation
to how much time we -
23:21 - 23:25spend directly exposed to
users and -
23:25 - 23:28how good our designs sort of
get. -
23:28 - 23:29He says it has to come in
this specific way. -
23:29 - 23:31It has to be direct
exposure, right? -
23:31 - 23:32It can't be something where
-
23:32 - 23:34someone generates a report
or through a graph, you -
23:34 - 23:38have to be interacting with
them somewhat real time. -
23:38 - 23:39It has to be a minimum of
every six weeks. -
23:39 - 23:41And it has to be for at
least two hours. -
23:41 - 23:44Otherwise your software will
get worse over time. -
23:46 - 23:49Our developers, our people
who are on, -
23:49 - 23:51on Wufoo were getting
exposed to our users four to -
23:51 - 23:53eight hours every single
week. -
23:54 - 23:57And what it does is it
changes the way you sort of -
23:57 - 23:57build software.
-
23:57 - 24:00Jared Spool has another way
of talking about how -
24:00 - 24:01we build products.
-
24:01 - 24:03Right?
-
24:03 - 24:07Let's imagine that this
represents all the knowledge -
24:07 - 24:10needed to sort of use your
app on a spectrum. -
24:10 - 24:12Right?
This is like no knowledge. -
24:12 - 24:12Right?
And this is -
24:12 - 24:14all the knowledge needed.
-
24:14 - 24:14Right?
And -
24:14 - 24:16these two lines are pretty
much your interactions with -
24:16 - 24:18users what you're trying to
get them to. -
24:18 - 24:20This is currently where
their knowledge point is and -
24:20 - 24:22this is the target knowledge
point that you're trying to -
24:22 - 24:25get them to, to understand
that to use your app. -
24:25 - 24:27The gap between those is
called the knowledge gap, -
24:27 - 24:28Jared Spool called it, Spool
calls. -
24:28 - 24:32And what's interesting about
this is there's only two -
24:32 - 24:34ways, right, to sort of fix
this. -
24:34 - 24:36That gap represents how
intuitive your -
24:36 - 24:38app is, right?
-
24:38 - 24:41You either get the user to
increase their -
24:41 - 24:45knowledge or you decrease
the amount of knowledge -
24:45 - 24:46that's needed to use your
application. -
24:46 - 24:48And oftentimes, as engineers
and people who build and -
24:48 - 24:51work on products, we think
let's add new features. -
24:52 - 24:53And new features only
-
24:53 - 24:55means let's increase the
knowledge gap. -
24:58 - 25:01So for us we actually
focused a lot on -
25:01 - 25:02the other sort of direction.
-
25:02 - 25:04And so what that meant we
spent a lot of time, -
25:04 - 25:0730% of our engineering time
was spent on internal tools -
25:07 - 25:09to help with our customer
support stuff. -
25:09 - 25:10But oftentimes it
-
25:10 - 25:13was spent helping people
help themselves. -
25:13 - 25:16Things like frequently asked
questions, tool tips. -
25:16 - 25:18Things like, if you just
click the help link, -
25:18 - 25:21right, instead of taking you
to the generic help sort of -
25:21 - 25:21documentation page,
-
25:21 - 25:24you go to the specific page
where you're looking at -
25:24 - 25:26is going to be most, sort
of, -
25:26 - 25:28appropriate for what you're
working on. -
25:28 - 25:30We redesigned our
documentation over and -
25:30 - 25:32over again, AB tested it
constantly. -
25:32 - 25:34One iteration of our
documentation reduced -
25:34 - 25:36customer support by 30% over
night. -
25:37 - 25:39It's one of those things
where, -
25:39 - 25:41like, overnight, all the
people that work on -
25:41 - 25:44the product immediately had
30% less work to do. -
25:47 - 25:48Now what happens if you have
everyone work on -
25:48 - 25:50customer support constantly,
and -
25:50 - 25:52thinking about it in terms
of a remarkable way? -
25:52 - 25:53Well, I talked a lot about,
-
25:53 - 25:56in the very beginning growth
is a function of -
25:56 - 25:58conversion and churn.
-
25:58 - 26:01This is Wufoo's growth curve
for the first five years. -
26:01 - 26:02Right?
-
26:02 - 26:03What's interesting is that
we -
26:03 - 26:08paid no money on
advertising, on marketing. -
26:08 - 26:11All of it was done by word
of mouth growth. -
26:11 - 26:12Right?
-
26:12 - 26:13And the interaction between,
like, -
26:13 - 26:16new users and downgrades are
this. -
26:19 - 26:21It's so slight what it
takes, that gap, -
26:21 - 26:22making that sort of work.
-
26:22 - 26:25And what a lot of people
keep forgetting is that -
26:25 - 26:27there's almost no difference
between an increase in -
26:27 - 26:30conversion rate, 1% increase
and a 1% decrease churn. -
26:30 - 26:33They do exactly the same
thing to your growth. -
26:33 - 26:36However, the ladder is
actually much easier to do. -
26:36 - 26:38It's much cheaper to do, in
your apps, and -
26:38 - 26:40a lot of the times we
neglect this, -
26:40 - 26:42neglect this to way far
along. -
26:42 - 26:43Right?
And we usually have our -
26:43 - 26:47B team works on these sort
of projects and services. -
26:47 - 26:48This is actually not the
graph that we track most of -
26:48 - 26:49the time at Wufoo, it's not
even the one I'm proud of. -
26:49 - 26:52This is the one I'm proud
of. -
26:52 - 26:55Cuz even though we have this
sort of nice, awesome curve -
26:55 - 26:58of growth, this is what, how
loud is this scale? -
26:58 - 27:00Keep the company small, have
an awesome culture. -
27:02 - 27:03And that required doing a
lot of -
27:03 - 27:09these things to help people
sort of do what they need. -
27:09 - 27:14So John Gotman noticed there
was a different type of -
27:14 - 27:19behavior for relationships
and why people divorced. -
27:19 - 27:22Basically there would be
some subset of people who -
27:22 - 27:24would stay together 10, 15
years and -
27:24 - 27:27then all of a sudden they
divorce and -
27:27 - 27:30there was none of the other
indicators which sort of -
27:30 - 27:32show that this is what was
gonna happen. -
27:32 - 27:35And I was looking through
the data and -
27:35 - 27:36I realized oh, there's no
passion, -
27:36 - 27:38there's no fire between
these people, right? -
27:38 - 27:40When it comes to
relationships they -
27:40 - 27:43kinda follow the second law
of thermodynamics, right? -
27:43 - 27:45In a closed energy system
things tend to run down so -
27:45 - 27:47you have to constantly be
putting energy and -
27:47 - 27:48effort back into it.
-
27:48 - 27:50Now the way a lot of people
sort of think about showing -
27:50 - 27:54people that I care about you
in products and in companies -
27:54 - 27:57is to do things like let's
have a blog, right? -
27:57 - 27:58Lets' have a newsletter?
-
27:59 - 28:02The thing is, we look at
these rates and basically it -
28:02 - 28:04was such a small percentage
of our active users that it -
28:04 - 28:06was like, most of our users
have no idea all -
28:06 - 28:09the awesome stuff that we're
doing for them. -
28:09 - 28:10So we built a new tool.
-
28:10 - 28:12We called it the Wufoo
system, and -
28:12 - 28:14what it allowed us to do was
just time stamp every new -
28:14 - 28:16feature that we're building
for users and -
28:16 - 28:18then every time they would
log in, we would look at -
28:18 - 28:21the difference between their
log in time or last log in -
28:21 - 28:23time and the new features
that were implemented. -
28:23 - 28:24And when you had this
message show up, -
28:24 - 28:25hey since you've been gone
-
28:25 - 28:28here's all the awesome stuff
that Wufoo did for you. -
28:28 - 28:30Hands down this was the most
talked about feature I've -
28:30 - 28:33ever had every time I went
out to talk to users. -
28:33 - 28:35Right they'd say like dude I
love that since you've be, -
28:35 - 28:37since you've gone thing even
though I pay the same amount -
28:37 - 28:40every single month you guys
are doing something for -
28:40 - 28:42me almost every week and
it's totally awesome and -
28:42 - 28:44makes me feel, I'm getting
maximum value. -
28:47 - 28:48The other thing that we did
in -
28:48 - 28:50addition to having everyone
support the people that -
28:50 - 28:52paid their pay check is have
them say thank you. -
28:52 - 28:57And this was a large part
due to us injecting sort of -
28:57 - 28:59humility and modesty in to
sort of the equation. -
29:01 - 29:03Every single Friday we would
get together and -
29:03 - 29:05we'd just write simple hand
written thank you cards to -
29:05 - 29:06our users.
-
29:07 - 29:11And I know there's tons of
people who would not be sort -
29:11 - 29:14of excited about doing this,
but it was a ritual that -
29:14 - 29:16made sort of all the
difference in terms of, -
29:16 - 29:20like, having a team that was
very tightly neat, -
29:20 - 29:21tightly knit, also.
-
29:21 - 29:23And working on stuff that
they really cared about. -
29:23 - 29:26They always constantly knew
what the mission was for, -
29:26 - 29:28and why we sort of did what
we did. -
29:28 - 29:30These aren't fancy thank you
cards, right? -
29:30 - 29:31They're just simple,
-
29:31 - 29:33like handwritten stuff on an
index card. -
29:33 - 29:34We threw in a sticker, and
-
29:34 - 29:37slapped on a dinosaur on the
front of it. -
29:37 - 29:42And, what's interesting is
we started this practice as -
29:42 - 29:46a result of the early days
of starting Wufoo. -
29:46 - 29:48Chris, Ron, and I were
talking, and we're trying to -
29:48 - 29:50figure out, what are we
gonna do to sort of show -
29:50 - 29:52users that we appreciate
them around Christmas. -
29:52 - 29:55And he, Chris came up with
this idea where he said -
29:55 - 29:59hey guys, so a couple years
ago my mom, like, made me -
29:59 - 30:02write thank you notes to all
of my relatives. -
30:02 - 30:04For my Christmas gifts.
-
30:04 - 30:05And I didn't really like to
do it but -
30:05 - 30:08the following year all my
presents were super good. -
30:08 - 30:12So, I think we should try
this for our business and -
30:12 - 30:14see how it goes.
-
30:14 - 30:17So, that first year we wrote
handwritten Christmas cards -
30:17 - 30:20to all of our users that
first year. -
30:20 - 30:21Second year rolls around,
and -
30:21 - 30:23we have too many customers,
like, and -
30:23 - 30:25it's still just the three
founders. -
30:25 - 30:27And we were going like we're
kinda screwed, -
30:27 - 30:29I don't know what we're
gonna do. -
30:29 - 30:31And we read a book called
the ultimate question, and -
30:31 - 30:33in it he talks about hey,
-
30:33 - 30:35just focus on your most
profitable users. -
30:35 - 30:37And just tend them and
-
30:37 - 30:38take care of them, then
it'll work out. -
30:38 - 30:39So we're like, all right,
-
30:39 - 30:41that, that makes sense,
that's scalable. -
30:41 - 30:42So that year we
-
30:42 - 30:43only write to our highest
paying customers. -
30:44 - 30:48And the January rolls around
that second year and -
30:48 - 30:51one of our long time loyal
users writes us and -
30:51 - 30:56he's basically like, hey
guys I, I really loved -
30:56 - 30:58that Christmas card you sent
me the first year and I just -
30:58 - 31:01wanted you to know I haven't
received my second card yet -
31:01 - 31:03and I'm just looking forward
to it. -
31:03 - 31:04I know you didn't forget
about me. -
31:06 - 31:07Thanks a lot.
-
31:07 - 31:12So we're like, fuck, because
the best way to sort of -
31:12 - 31:14exceed expectations is not
to send any to begin with, -
31:14 - 31:16so we were like sort of in
this conundrum. -
31:16 - 31:18And what we decided, after
thinking about it for -
31:18 - 31:20a while, is that we need to
stop doing it, -
31:20 - 31:22you know, just one time a
year. -
31:22 - 31:24It needs to be something
that's part of the culture. -
31:24 - 31:26Happens every, every, every
sort of week, even. -
31:26 - 31:28And even though we'll never
catch up to all of -
31:28 - 31:30our customers.
-
31:30 - 31:30Just the practice of
-
31:30 - 31:31doing it will make all the
difference. -
31:33 - 31:34I talked a lot
-
31:34 - 31:37about a bunch of like
lovey-dovey stuff and -
31:37 - 31:39sort of like touchy-feely
things that I think a lot of -
31:39 - 31:41engineers don't like to
think about too often. -
31:41 - 31:46And so I'll end on some sort
of hard business -
31:46 - 31:47data or research.
-
31:47 - 31:49There's an article that was
put out by -
31:49 - 31:51the Harvard Business Review
several years ago by -
31:51 - 31:54Michael Treacy and Fred
Wiersema and in it they -
31:54 - 31:56talk about the discipline of
market leaders. -
31:56 - 31:57They say there's only three
ways that you -
31:57 - 32:02achieve market dominance and
depending on how you want to -
32:02 - 32:03achieve that market
dominance you have -
32:03 - 32:06to organize your company in
a very specific way. -
32:06 - 32:09Best price, best product and
best overall solution. -
32:09 - 32:10If you want to be
-
32:10 - 32:12the best price out there you
focus on logistics. -
32:12 - 32:14A Wal-Mart, an Amazon.
-
32:14 - 32:15If you want to be the best
product out there you -
32:15 - 32:17focus on R&D.
-
32:17 - 32:17Apple's usually
-
32:17 - 32:19a quintessential example of
that. -
32:19 - 32:21Best overall solution is
about -
32:21 - 32:22being customer intimate.
-
32:22 - 32:23And this is the path that
you see followed by -
32:23 - 32:26luxury brands and
hospitality industry. -
32:27 - 32:30What I love about this path
towards market dominance is -
32:30 - 32:35that the third one is the
only one that -
32:35 - 32:38everyone can do at any stage
of their company. -
32:38 - 32:40Requires almost no money to
get started with. -
32:40 - 32:42Usually just co.
-
32:42 - 32:45Requires a little bit of
humility and some manners. -
32:45 - 32:49And as a result you can
achieve the success as any -
32:49 - 32:50other people in sort of your
market. -
32:52 - 32:54That's all I got.
-
32:54 - 32:54Thank you very much.
-
33:00 - 33:03Yeah, let's take some
questions if you guys -
33:03 - 33:04have any.
-
33:04 - 33:05Right in the back there.
-
33:05 - 33:07>> Building products that
users love? -
33:07 - 33:10You might have multiple
different types of users. -
33:10 - 33:14How do you build one product
that all users love? -
33:14 - 33:16Maybe there is a feature
that one really likes but -
33:16 - 33:18detracts value from one
that. -
33:18 - 33:19>> All right.
So what do you do -
33:19 - 33:21when you have a product with
lots of -
33:21 - 33:22different type of users,
right? -
33:22 - 33:23Some users will love one
thing and -
33:23 - 33:25another will, will another.
-
33:25 - 33:26And I agree, there's
-
33:26 - 33:27a interesting fine line for
that. -
33:27 - 33:29What I always, usually tell
people, -
33:29 - 33:32is focus on the people who
are the most passionate, -
33:32 - 33:32especially in the early
stages. -
33:32 - 33:34Right?
Whoever's, whatever niche -
33:34 - 33:37it's gonna be, that's who I
focus on completely. -
33:37 - 33:39Things that a lot of
different projects did. -
33:39 - 33:40I think Ben Silverman of
-
33:40 - 33:43Pinterest started off with a
designer bloggers, right? -
33:43 - 33:45Curtail your thing for
-
33:45 - 33:47them and eventually you'll
figure out sort of universal -
33:47 - 33:49values that will appeal to a
lot of other people. -
33:49 - 33:51So, just start one at a
time. -
33:51 - 33:52And.
-
33:52 - 33:54The, a lot of the examples
that you see up there, -
33:54 - 33:55a lot of people make the
mistake is like, oh, -
33:55 - 33:57I'll just make my app funny.
-
33:57 - 34:00But, humor is like really
difficult to do. -
34:00 - 34:01Right?
What you wanna shoot for -
34:01 - 34:02is something sort of witty.
-
34:02 - 34:04And, quite honestly,
-
34:04 - 34:06you have to get
functionality right. -
34:06 - 34:07So like the Japanese
quality. -
34:07 - 34:10If you don't have a on
there, right, don't try to -
34:10 - 34:13do anything witty, right,
cuz it will backfire on you. -
34:13 - 34:16So hands down, our number
one focus is make it as -
34:16 - 34:18easy to use as possible for
and -
34:18 - 34:20anything else on top was
polish. -
34:20 - 34:21Right here.
-
34:21 - 34:28>> So so everybody says that
to focus on your product. -
34:28 - 34:29I'm also good at that.
-
34:29 - 34:32I love to do a project and I
love to make it the best. -
34:32 - 34:35But we are to that certain
point that we are focused on -
34:35 - 34:38our product but we don't get
like constantly right? -
34:38 - 34:38Sorry.
-
34:38 - 34:42So second thing so how much
we should focus on product? -
34:42 - 34:44But because we should do now
marketing. -
34:44 - 34:45We should get somebody
customers and -
34:45 - 34:48like and like start talking
to customers but -
34:48 - 34:50when you are too focused on
your product. -
34:50 - 34:53Like users online have them.
-
34:53 - 34:55Right, so what exactly do
you guys mean when you -
34:55 - 34:58are saying like focus only
on your product and -
34:58 - 34:59give the best product?
-
34:59 - 35:02>> Okay.
-
35:02 - 35:05So the question sort of is
how do we balance this -
35:05 - 35:08sort of thing where we wanna
be obsessed with -
35:08 - 35:10working on product.
-
35:10 - 35:12Yet.
But all the other skills, -
35:12 - 35:15and sort of tasks that are
needed by a company, -
35:15 - 35:16like marketing and branding
and -
35:16 - 35:18all that stuff, and how, how
we sort of balance that. -
35:18 - 35:20And the thing is, like, it
starts off as you juggling, -
35:20 - 35:23like, tons of things
constantly in the air. -
35:23 - 35:24The thing is, if you're
working on products, like, -
35:24 - 35:26you should also always have
this flip side as -
35:26 - 35:28when you're talking to
users. -
35:28 - 35:29Right?
And for -
35:29 - 35:32us inside of Wufoo, the way
we got people to talk to -
35:31 - 35:32users is they just did
customer support. -
35:32 - 35:35And they got to see
firsthand, right away. -
35:35 - 35:36Whether that feature sucked
or not, -
35:36 - 35:38and also impacted everyone
else in the company, -
35:38 - 35:39because everyone had a
customer support shift. -
35:39 - 35:42So you have this sort of
social incentive to sort of -
35:42 - 35:43make everything work.
-
35:43 - 35:46And so, like I said, there
should be no point where -
35:46 - 35:47you're only focused on
product. -
35:47 - 35:49You should always have time
where you work on product, -
35:49 - 35:51and then you see sort of
what users say, say to you. -
35:51 - 35:53And you should always have
this virtual, like, -
35:53 - 35:54feedback loop on there.
-
35:54 - 35:57So be careful when you don't
have that. -
35:57 - 35:58Usually what ends up
happening, if you're lucky. -
35:59 - 36:01In terms of marketing and
-
36:01 - 36:04sales, like, usually my
feeling is like, you have to -
36:04 - 36:06spend money on marketing and
advertising, all this stuff. -
36:06 - 36:07It's usually a tax you pay
because you -
36:07 - 36:08haven't made your product
remarkable. -
36:08 - 36:10Right?
-
36:10 - 36:12Word of mouth growth is the
easiest kind of growth, and -
36:12 - 36:14it's how a lot of the great
companies sort of grow. -
36:14 - 36:16So figure out how to wait,
how to like, -
36:16 - 36:18have a story that people
want to tell. -
36:18 - 36:19About your product.
-
36:19 - 36:20Where they're the most
interesting person at -
36:20 - 36:22the dinner table, right?
-
36:22 - 36:24And then that person is your
sales person, right? -
36:24 - 36:25That person is your sales
force for you. -
36:25 - 36:28Right here.
-
36:28 - 36:32>> like, where do you find
crystal clear customer or -
36:32 - 36:36user need and the demand is
there is the right solution. -
36:36 - 36:38How do you communicate with
engineering and -
36:38 - 36:42designing team to make sure
that because sometimes -
36:42 - 36:44people in the team come up
with ideas, and but -
36:44 - 36:46still at the end of the day,
-
36:46 - 36:48how can you make a decision
with where to go? -
36:48 - 36:50>> Oh, so how do you make a
decision on product? -
36:52 - 36:53And communicate that with
your sort of -
36:53 - 36:54engineering team when
there's like lots of -
36:54 - 36:55different directions to go?
-
36:57 - 36:59My feeling is that.
-
37:01 - 37:03So for us we just looked at
support. -
37:03 - 37:04It was really easy cuz
-
37:04 - 37:06you often just saw what are
things that -
37:06 - 37:08people are having the most
amount of problems with? -
37:08 - 37:10Or people asking all the
time. -
37:10 - 37:12You cannot help but get
feature requests from -
37:12 - 37:14people no matter like
whatever opening or -
37:14 - 37:16orifice you have in your
product or app. -
37:16 - 37:18Like, people will like jam
feature requests in there. -
37:18 - 37:19So you're easily going to
-
37:19 - 37:20know sort of what they sort
of what. -
37:20 - 37:22Your job as a product person
and -
37:22 - 37:26engineer is to not just do
what they say, because that -
37:26 - 37:28way, you'll just be a slave,
is to figure out, -
37:28 - 37:31sort of deeply, what are the
reasons why underlying those -
37:31 - 37:33things and sort of solve
that deep underlying reason. -
37:33 - 37:35The thing is that everyone
wants to -
37:35 - 37:36have a different way of.
-
37:36 - 37:38To sort of go, then
ultimately it comes down to, -
37:38 - 37:40like, someone's gonna figure
something out. -
37:40 - 37:43But I also make the smallest
version of each little idea. -
37:43 - 37:47No longer than a week or two
weeks to build it out there. -
37:47 - 37:47And you can try them out and
-
37:47 - 37:49see sort of what works and
don't work. -
37:49 - 37:51I think it's dangerous to
-
37:51 - 37:52have multiple different
product directions that -
37:52 - 37:54requires lots of time to
sort of figure out. -
37:55 - 37:56Sam.
-
37:56 - 37:58>> Related to that can you
tell the story of -
37:58 - 38:01how the king for a day thing
>> Yeah. -
38:01 - 38:02Okay.
So -
38:04 - 38:07so I don't like hackathons.
-
38:07 - 38:10I think they sort of suck in
terms of ones done inside -
38:10 - 38:10of companies.
-
38:10 - 38:11Because.
-
38:11 - 38:14You spend like 48 hours
working on something really -
38:14 - 38:16hard that you're sort of
passionate about and -
38:16 - 38:2099% of them never make it to
production right, and -
38:20 - 38:22it's sort, sort of real like
super sad. -
38:22 - 38:24So for us we like flipped it
on it's head and -
38:24 - 38:26we came up with an idea that
we called king for -
38:26 - 38:29a day and it actually worked
over the weekend. -
38:29 - 38:32But how it worked is someone
randomly in the company got -
38:32 - 38:35drawn and they got to be the
king, and the king got to -
38:35 - 38:36tell everyone else what to
do on the product. -
38:36 - 38:40So everything that was
bothering them about Wufoo. -
38:40 - 38:41About the customer support
stuff, or -
38:41 - 38:43some feature they really
want to have built. -
38:43 - 38:44They've got the engineering
resources, the market -
38:44 - 38:47resources, the advertising
resources of everyone inside -
38:47 - 38:49of the company, to make it
sort of happen. -
38:49 - 38:50And of course, we'd work
with them to -
38:50 - 38:53figure out like what can be
actually done in 48 hours. -
38:53 - 38:57But we would do this one to
two times a year. -
38:57 - 38:58And it was like a huge hit
and -
38:58 - 39:00it was a boost to morale,
cuz what people most love. -
39:00 - 39:03It is like working on things
where it's like, oh, -
39:03 - 39:04I made a difference to the
app. -
39:05 - 39:06Right?
-
39:06 - 39:09And so, for us, that's one
way that we would like -
39:09 - 39:11sort of divide time for like
product direction. -
39:11 - 39:14It's like some times the
people that work for -
39:13 - 39:15you, they have a strong
opinion about where it, -
39:15 - 39:15where it should go.
-
39:15 - 39:18And it's a good way to sort
of democratize it -
39:18 - 39:19a little bit, by rotating it
around. -
39:21 - 39:22Yes.
-
39:22 - 39:24>> You said you guys all
work from home. -
39:24 - 39:25Usually seems like a
nightmare. -
39:25 - 39:28In that office, how do you
make that work? -
39:28 - 39:29>> Okay, so we all work from
home. -
39:29 - 39:30So, I will tell you this.
-
39:30 - 39:32We all still work within the
Tampa Bay area. -
39:32 - 39:33We would allow anybody to
-
39:33 - 39:35work from any where but
usually. -
39:35 - 39:36As we tried to recruit
them,they sort of -
39:36 - 39:38meet our team, and they just
decide, -
39:38 - 39:40okay we just want to come
and move here anyway. -
39:42 - 39:47Remote working, is
especially tricky, -
39:47 - 39:48a lot of people like to
romanticize it, -
39:48 - 39:50especially people, who are
like employees. -
39:50 - 39:54But the thing is, An office
gives you a lot of, sort of, -
39:54 - 39:57benefits, right, and
efficiencies that you now -
39:57 - 39:59have to compensate for when
you remote working. -
39:59 - 40:01But remote working also has
these other sort of -
40:01 - 40:03efficiencies in place, for
example, -
40:03 - 40:05I don't have to worry about
my employees losing two -
40:05 - 40:08hours of their day to
commuting, for instance. -
40:08 - 40:09So the biggest thing that we
had to do for -
40:09 - 40:12remote working is to respect
people's time. -
40:12 - 40:14And so the way we had it set
up is we actually had -
40:14 - 40:16a four and a half day work
week at Wufoo. -
40:16 - 40:19Half day on Friday was for
all the meetings and stuff. -
40:19 - 40:21We said like no business
deal, -
40:21 - 40:23meetings, no talking with
other outside parties. -
40:23 - 40:26They all have to be done on
Friday, on that half day. -
40:26 - 40:27It couldn't be done in the
middle of the week. -
40:27 - 40:29And then also one day of
everyone was -
40:29 - 40:31already dedicated customer
support. -
40:31 - 40:33So everyone in our company
effectively only had -
40:33 - 40:35three days each week to
actually build or -
40:35 - 40:37work on whatever they were
doing. -
40:37 - 40:38But I actually firmly
believe that if you -
40:38 - 40:41have three solid days,
right, eight to ten hours, -
40:41 - 40:43where you are only working
on what you need to build, -
40:43 - 40:45you can get a ton of shit
done. -
40:45 - 40:48And so what we said was,
-
40:48 - 40:49you have to respect
everyone's time during that -
40:49 - 40:51three day period, if they're
in that three day period. -
40:51 - 40:53And what we came up with is
a 15 minute rule and the way -
40:53 - 40:57it worked is you can have a
discussion like a chat or -
40:57 - 40:59a phone call or whatever
with someone but it could go -
40:59 - 41:02no longer than 15 minutes so
if you have some complicated -
41:02 - 41:04issue that you couldn't
figure out we'd say, -
41:04 - 41:07at 15 minutes you are to
immediately table that item. -
41:07 - 41:10Right and have us discuss it
on Friday and -
41:10 - 41:13you'd move on to the next
item on your list right, -
41:13 - 41:14the enhanced productivity.
-
41:16 - 41:18More often than not I, I
would say 90% of the time -
41:18 - 41:20the item never got brought
up on Friday because usually -
41:20 - 41:23what would happen is people
would sleep on it and then -
41:23 - 41:26they'd just magically say
like, hey I found a solution -
41:26 - 41:28or like hey that's not a big
problem whatsoever. -
41:29 - 41:31Because most problems inside
of companies, -
41:31 - 41:34they don't need to be solved
in real time, or right away. -
41:34 - 41:37The only things are like
when the site is down, or -
41:37 - 41:38payments aren't working.
-
41:38 - 41:39Right?
Everything outside of -
41:39 - 41:43that is basically kind of
luxury, so -
41:43 - 41:46focus on your primaries as
much as possible. -
41:46 - 41:46And, as a result,
-
41:46 - 41:48our ten person team did far
more than many, -
41:48 - 41:51many other companies, as a
result. -
41:51 - 41:55But, it takes extra work to
make remote working happen. -
41:55 - 41:58We are an extremely
disciplined sort of team. -
41:58 - 41:58And I would have to say
-
42:00 - 42:03there's almost not many YC
companies that actually have -
42:03 - 42:04been able to replicate sort
of what we do. -
42:04 - 42:06I think there's only two
other companies in YC that -
42:06 - 42:10sort of have the same sort
of discipline working style. -
42:10 - 42:12It takes more work in a very
different fashion. -
42:12 - 42:12Right?
-
42:12 - 42:15An office allows you to be a
little bit lazier, right, -
42:15 - 42:17in terms of all these things
around productivity. -
42:21 - 42:22Okay.
Over here. -
42:22 - 42:23>> Right.
-
42:23 - 42:26Just to go off that question
as the leader of the team -
42:26 - 42:30how did you manage to
instill a company culture -
42:30 - 42:31but and also the count,
-
42:31 - 42:34accountability the employees
especially cuz. -
42:34 - 42:36Working space.
-
42:36 - 42:39>> Okay, how do we, how do
we set up accountability for -
42:39 - 42:39employees?
-
42:39 - 42:40As, as a manager.
-
42:40 - 42:44Alright, so, at Wufoo we
were profitable nine months -
42:44 - 42:47after launch, so, we had
profit sharing, right? -
42:47 - 42:50And so, it makes pretty
simple and clear. -
42:50 - 42:53It, it would be a multiple
of whatever bonus pull that -
42:53 - 42:54we sort of had, and
-
42:54 - 42:57the performance measures
would be based on sort of, -
42:57 - 42:58how they did in customer
support. -
42:58 - 43:00All right, on their duties
there, and sort of -
43:00 - 43:03what they said they were
wanting to accomplish or do. -
43:03 - 43:05I don't like process, and I
don't like lots of -
43:05 - 43:07tools to help get people to
be productive. -
43:07 - 43:09So the only thing that we
had for -
43:09 - 43:10helping people manage, like,
-
43:10 - 43:12sort of their projects, is
to-do lists. -
43:12 - 43:14And that is, like, simple
text files that we -
43:14 - 43:15shared in our Dropbox
account. -
43:15 - 43:17Each person had their name
on it, and you got to -
43:17 - 43:20see every time somebody
updated their to-do list. -
43:20 - 43:22What we said is every single
night just -
43:22 - 43:23set everything that you did
that day. -
43:23 - 43:24Right?
And then on Friday, -
43:24 - 43:25we would just go over.
-
43:25 - 43:26Okay, this is what you
-
43:26 - 43:27said last week that you're
gonna do. -
43:27 - 43:28This is what you actually
got done, or -
43:28 - 43:29the sort of the problems at
hand. -
43:29 - 43:31And it's super simple,
right? -
43:31 - 43:34It creates this like nice
written trail for -
43:34 - 43:36how to sort of handle stuff,
right? -
43:36 - 43:38And I don't have to worry
about managing them, right? -
43:38 - 43:39They sorta set the tone for
-
43:39 - 43:42how they want to be sort of
assessed. -
43:42 - 43:43And it makes it really
simple. -
43:43 - 43:45And for people who are
excellent at what they do. -
43:45 - 43:47Right?
It works very, very well. -
43:48 - 43:49And then when you actually
have problems, -
43:49 - 43:51it's very easy to fire
people. -
43:51 - 43:52I was fortunate that I
-
43:52 - 43:54never had to fire anyone at
Wufoo, right? -
43:54 - 43:55But we were able to correct
a lot of -
43:55 - 43:58people's behavior very, very
quickly. -
43:58 - 43:59Cuz we just kinda look at
this, and -
43:59 - 44:00it's like look, this is your
pattern of behavior. -
44:00 - 44:03You finish a fraction of the
items on your list. -
44:03 - 44:04You do most of
-
44:04 - 44:06the items at the last second
right before Friday. -
44:06 - 44:07That's a problem, you've
gotta manage your time -
44:07 - 44:10better and this is evidence
that you've provided to us. -
44:10 - 44:12All we have to do is sort of
describe it back to you and -
44:12 - 44:15because everyone in the
company sort of sees it, -
44:15 - 44:16right, there's social
pressure that's put -
44:16 - 44:18into place that helps make
it all sort of happen. -
44:20 - 44:21Right here.
-
44:21 - 44:23>> How did you hire people
that, you know, -
44:23 - 44:26you felt would be able to
work with in this kind of -
44:26 - 44:28environment that's, that's
>> So, -
44:28 - 44:32how do you hire people that
can work, remotely? -
44:32 - 44:34And then sort of work in
this sort of fashion. -
44:34 - 44:36So, pretty easily,
-
44:36 - 44:38you have them work on a side
project for you. -
44:38 - 44:39So you contract them out,
and -
44:39 - 44:41they have to work remotely.
-
44:41 - 44:43As such, usually, the
projects I like to -
44:43 - 44:45have them work on is about a
month long, right? -
44:45 - 44:47I could do things much
faster for a week. -
44:47 - 44:48But usually get a good sense
of, -
44:48 - 44:50like, how well people sort
of manage themselves, and -
44:50 - 44:52work on things from a
project like that. -
44:52 - 44:54So that was always the first
assessment. -
44:54 - 44:56Like we never did anything
just by interviews. -
44:57 - 45:00The other thing we had to,
sort of, screen them for -
45:00 - 45:01is their ability to do
customer support. -
45:01 - 45:02Because not any,
-
45:02 - 45:06every engineer sort of has
the empathy skills to -
45:06 - 45:08handle that stress.
-
45:08 - 45:12So sometimes I would have
people write breakup letters -
45:12 - 45:14to me, right, in an
interview, just like hey, -
45:14 - 45:17pretend you have to break up
with me you have 15 minutes -
45:17 - 45:19to write in on there and you
can only write it by hand, -
45:19 - 45:22what are you gonna say?
-
45:22 - 45:24And so you get a good sense
of sort of their writing -
45:24 - 45:26skills, because like 90% of
what you do in customer -
45:26 - 45:28support is tell them bad
news, like, oh, we don't -
45:28 - 45:31support that feature, sorry
that's not gonna work, or. -
45:31 - 45:33It's not gonna be available.
-
45:33 - 45:34And so people have to have
sort of tacked at that. -
45:34 - 45:36>> How 'bout one more
question? -
45:36 - 45:39>> One question, right here
the glasses. -
45:39 - 45:40>> So, since Wiki has all
these like tricks and -
45:40 - 45:43experiments that have really
helped the company. -
45:43 - 45:45Do you INAUDIBLE] of ones
that didn't work out? -
45:45 - 45:46>> Have all these tricks and
-
45:46 - 45:47experiments to help the
company, -
45:47 - 45:49are there any ones that
didn't work out? -
45:50 - 45:52All right, I'll talk about
one. -
45:52 - 45:55So one of the things that we
did early on to try to -
45:55 - 45:57motivate ourselves was try
to get, -
45:59 - 46:01like we understood this idea
of like crunch mode and -
46:01 - 46:03that it's really bad for
people. -
46:03 - 46:04Like if you're doing the
subscription business you -
46:04 - 46:06need people to last for the
long term, and -
46:06 - 46:08video games a lot of times
they like crunch people. -
46:08 - 46:10All, all the time.
-
46:10 - 46:11For like a specific deadline
or -
46:11 - 46:13have multiple sprints every
two weeks and -
46:13 - 46:14you have to shoot up to this
deadline and -
46:14 - 46:15it's like exhausting.
-
46:15 - 46:18And so, because, when it's
happening it's like you -
46:18 - 46:19might get an increase in
productivity but -
46:19 - 46:20the recovery period that you
need for people is -
46:20 - 46:24always greater, right, than
the productivity you gain. -
46:24 - 46:25And at a company where you
-
46:25 - 46:26need everyone doing customer
support and -
46:26 - 46:27being on their game and
constantly put, pushed out -
46:27 - 46:30features you don't have time
for recovery. -
46:30 - 46:33So, we were thinking about,
okay, we want to build like, -
46:33 - 46:37a company vacation into how
Wufoo sort of works to -
46:37 - 46:38reward our users every
single year. -
46:38 - 46:39And we said okay,
-
46:39 - 46:40if the vacation is sort of
built in there for -
46:40 - 46:43the recovery, we could have
one crunch period, right, -
46:43 - 46:44before that vacation set up.
-
46:44 - 46:46And we'll just only do
customer support that will -
46:46 - 46:48just sort of scale with
people. -
46:48 - 46:49So.
-
46:49 - 46:56The way we did the very
first crunch mode -
46:56 - 46:57is that it was just between
the three founders. -
46:57 - 47:01And we had each of us draw a
ten-item to-do list. -
47:01 - 47:02That would be fairly
aggressive. -
47:02 - 47:03And the first person to
-
47:03 - 47:06get through seven of their
items would win. -
47:06 - 47:07And the the last person to
-
47:07 - 47:09get through seven of their
items. -
47:09 - 47:11Would become what we called
trippage. -
47:11 - 47:13And trippage meant that you
carried other -
47:13 - 47:16person's luggage and got
people drinks when you're on -
47:16 - 47:16the company vacation.
-
47:18 - 47:22So we did that and during
that period, -
47:22 - 47:23everyone was like pretty
excited about it and -
47:23 - 47:25motivated, and only the
winner got to -
47:25 - 47:28choose the next company
vacation the following year. -
47:28 - 47:31And then all the sudden Ryan
had basically poorly -
47:31 - 47:33estimated the items on his
list. -
47:34 - 47:36And he realized very quickly
I'm going to fucking lose. -
47:36 - 47:39And so he was just like I
give up. -
47:39 - 47:40And he just sort of stopped.
-
47:40 - 47:42So crunch mode turned out to
be blah mode for -
47:42 - 47:44him because he knew he was
gonna lose and -
47:44 - 47:46became pretty demoralized.
-
47:46 - 47:47So as a result of doing
that, we -
47:47 - 47:50decided not to do it in that
similar fashion anymore. -
47:50 - 47:53So, good idea that we like
to talk about. -
47:53 - 47:54But is one that we, we never
did again. -
47:55 - 47:56All right, guys.
-
47:56 - 47:57Thanks a lot.
-
47:57 - 47:59You can email me at kevin@
- Title:
- Lecture 7 - How to Build Products Users Love (Kevin Hale)
- Description:
-
Lecture Transcript: http://tech.genius.com/Kevin-hale-lecture-7-how-to-build-products-users-love-part-i-annotated
Kevin Hale, Founder of Wufoo and Partner at Y Combinator, explains how to build products that create a passionate user base invested in your startup's success.
See the slides and readings at startupclass.samaltman.com/courses/lec07/
Discuss this lecture: https://startupclass.co/courses/how-to-start-a-startup/lectures/64036
This video is under Creative Commons license: http://creativecommons.org/licenses/by-nc-nd/2.5/
- Video Language:
- English
- Team:
PACE
- Duration:
- 48:02