0:00
okay so this is how to evaluate startup
0:05
ideas and this is actually a new set of
0:08
content that we've developed based on a
0:11
lot of feedback that we saw from the
0:13
last startup school and what we noticed
0:15
is a lot of people's challenges so last
0:17
year's curriculum actually had a lot of
0:20
content that ended up being when we
0:21
looked at the data for who's
0:23
participating in startup school was like
0:24
oh this is much more advanced it's much
0:26
further along a lot of people for
0:29
instance like I had no idea or like I
0:31
have too many ideas that they don't know
0:33
which one to pursue it was a main reason
0:35
why a lot of people are only able to
0:37
work on their startup sometimes
0:38
part-time yes they might be stuck but
0:40
that resources but they didn't have
0:42
conviction they didn't know like oh what
0:44
would I have to believe in order to say
0:47
like I want to quit my job this is also
0:50
a really great sort of skill to sort of
0:52
have because if you are realizing you
0:55
need to pivot how do you evaluate if you
0:57
need to do that and then also if you're
0:59
pivoting to something else like how do
1:01
you evaluate whether something is worth
1:02
going to and if you already have a
1:04
launch company then you might have
1:06
problems with like why isn't this
1:07
growing or how do i improve it and
1:10
evaluating your startup especially in a
1:12
way that sort of adventures evaluate
1:15
startups ideas we find it's going to be
1:18
really really useful there's just like
1:21
myth about Y Combinator is that people
1:25
think YC only funds companies who have
1:27
tons and tons of traction right that
1:30
nowadays the only way to get into IC you
1:32
have to have lots of revenue or tons of
1:34
users already and part of that has to do
1:36
with like the press and then the
1:39
exposure of the companies that make it
1:40
the demo day those are the stories you
1:42
tend to hear but that being said there's
1:44
lots of great examples of companies who
1:47
actually got accepted just with an idea
1:49
and so zenefits is a really good classic
1:52
one Parker was a single non-technical
1:57
founder who pitched an idea to YC and he
1:59
got in that way we also had reddit so
2:03
technically they were forced to pivot
2:05
right away so they hadn't written a
2:06
single line of code and then my
2:08
experience in in YC is actually exactly
2:10
the same so when I founded wafu and we
2:12
entered that second
2:14
we also had not written a single line of
2:16
code PG had invested us at just the idea
2:19
stage and so I feel fully committed now
2:21
as a partner to always be trying to find
2:24
and dedicate time and energy to funding
2:26
companies we're just at the idea stage a
2:28
lot of our efforts here for working on
2:31
startup school is to help us help you
2:34
work on how to talk and think about your
2:37
startup and that if we can fix that it
2:39
can help you sort of inspire us to be
2:41
like oh yeah I can believe in what
2:44
you're doing and a lot of times founders
2:47
get in the way when they're telling
2:48
their narrative so how can I predict if
2:52
an investor will like my ideas that's
2:54
ultimately what we're trying to figure
2:56
out and the answer is really easy and so
3:00
for us at YC the definition of a
3:02
start-up is a company that is designed
3:04
or created to try to grow very quickly
3:08
so if you're not trying to build a
3:10
company that grows very very fast then
3:13
you're just building a normal company
3:15
it's a small business and there's
3:17
nothing wrong with that but these
3:20
companies are the ones that investors
3:22
are interested in so if you look hoping
3:23
to build something that will have tons
3:27
of users that will have huge valuations
3:29
that'll be able to attract venture
3:30
funding then the evidence that we want
3:33
is evidence that shows that your company
3:35
can grow quickly I have a confession to
3:38
make if you ever meet me or talk to me
3:42
about your startup and kind of
3:45
recruiting or at one of these events I
3:46
will never tell you that I do not like
3:50
your idea and there's a reason for that
3:54
is not to be nice to you is not to blow
3:57
smoke up your ass
3:58
I learned this way of thinking from Paul
4:02
Graham he says like look the average
4:04
investor or a lot of investors that you
4:05
see when you talk to them about your
4:08
idea it feels like they're trying to
4:10
poke holes in your idea they're trying
4:12
to figure out what's wrong with it and
4:13
then I like trying to show some not just
4:15
how smart they are and he explained that
4:18
like his job in the way he sees our work
4:21
at YC is that it's not to figure out
4:25
what's wrong with the company
4:26
but to figure out how it could possibly
4:29
win because our bets the ones that win
4:33
are the ones that are non-obvious right
4:35
and so for us to figure out the
4:38
non-obvious stuff it's good it's not
4:40
sound obvious when they first tell you
4:42
and so we have to like work on our
4:44
imagination we have to work on our
4:46
optimism to figure out what is the way
4:48
that whatever story that they're telling
4:50
me could become a billion dollar company
4:53
and then a great investor pitches that
4:56
back to the founder I figure out all the
5:00
ways that I think whatever you're doing
5:01
could possibly become big and then I'm
5:04
trying to convince you that this is what
5:06
needs to happen so that I would have the
5:09
evidence or that you will be on the
5:12
right path to having company that grows
5:13
quickly a start-up idea is basically a
5:18
hypothesis and this is the way you
5:19
should think about it it's a hypothesis
5:21
about why a company could grow quickly
5:24
and your job is to figure out how to
5:27
construct your hypothesis basically the
5:30
pitch to the investor so they understand
5:33
how it can go grow quickly a lot of
5:36
times people make the mistakes of trying
5:38
to just accurately describe or over
5:41
describe a lot of different parts so I'm
5:43
gonna break it this down so just like a
5:45
normal hypothesis has a pretty decent
5:47
structure for this this will hopefully
5:49
help you sort of workshop like
5:51
understanding oh this is exactly all the
5:53
reasons why this should succeed and so
5:55
even before we start even building
5:57
anything we can have an understanding of
5:59
like oh here's the potential path of
6:01
this company or here's the things I need
6:03
to prove the show that this company
6:05
could do well so the first is the
6:09
problem so startup idea is composed of
6:11
three parts the first part is a problem
6:13
and it's basically the initial
6:14
conditions you have to explain to me
6:17
like what is the setting for this
6:19
company that allows it to be able to
6:22
grow quickly the second is the solution
6:25
so this is basically what is the
6:27
experiment that you're basically running
6:29
within those conditions for it to grow
6:32
really quickly and the third is what's
6:34
your insight so what's your explanation
6:35
why the thing that you're going to try
6:38
your experiment is going to end up
6:40
successful those are the three
6:43
components that are always trying to
6:44
figure out when I'm listening to
6:45
someone's pitch here's a tip for talking
6:48
about the problem or to know whether
6:50
your problem your initial conditions are
6:52
correct the first is good problems
6:57
they're popular so a lot of people have
6:59
the problem you want to avoid problems
7:02
that there's a small number of people
7:03
that have it
7:06
we like problems that are growing so
7:09
therefore the market basically like is
7:11
it growing at a rate that's like more
7:13
and more people going to be having the
7:14
problem and it's it's growing faster
7:16
than other people's or other types of
7:18
problems we like problems that are
7:21
urgent ones that need to be solved very
7:23
very quickly we like problems that are
7:26
really expensive to solve because if
7:28
you're able to sort of solve it then you
7:30
can charge a lot of money potentially we
7:32
like problems that are mandatory right
7:35
so therefore it's like ah people have
7:37
this problem and they have to solve it
7:39
and then we like problems that are
7:41
frequent ones that people are gonna
7:43
encounter over and over and over again
7:44
and often in a frequent time interval so
7:48
what you want to have is like some
7:52
aspect of the problem that you're
7:53
working on at least one of them and
7:56
that's ideal if you have multiple of
7:58
them you don't have to have all of them
7:59
but it's one of those things where it's
8:01
like if your company isn't growing or if
8:04
someone's not as excited about the
8:05
problem it's probably missing some of
8:07
these characteristics the last one about
8:10
frequency is super important because I
8:14
like problems and you'll find a lot of
8:16
other YC partners like problems a lot
8:18
that gives people a lot of opportunities
8:21
to convert part of that has to do with
8:24
some theory so BJ Fogg he's a researcher
8:27
at Stanford and he tosses his formula up
8:30
all the time I mean he says basically if
8:32
you're trying to change someone's
8:34
behavior you have three things you need
8:38
to have in place you need to have the
8:40
motivation the ability and the trigger
8:42
they need to all happen at the same time
8:43
so the motivation is like I have this
8:45
problem I need to solve what is it
8:47
whatever it is the ability is your
8:48
startup and the last is the trigger
8:50
what's going to be the thing that gets
8:52
them to all of a sudden realize Oh
8:54
need to solve it with your thing and so
8:56
a lot of companies will have like oh I
8:58
built something but for some reason no
9:00
one's signing up or like they're not
9:01
using it they're not engaged I have no
9:03
retention and a lot of times is because
9:05
it's like you're hoping that they will
9:07
somehow just remember on their own that
9:11
they have the problem and to start using
9:13
you and oftentimes most companies don't
9:15
send enough for example email
9:17
notifications or triggers or reminders
9:20
or figure out ways to come back into the
9:22
app or figure out ways to be back in
9:24
front of the user at the right time and
9:26
if you can't figure out those
9:28
opportunities it's really hard to get
9:30
people to switch over to using your
9:32
solution service or product so our ideal
9:37
problems are millions of users right
9:41
millions of people have it that's why
9:43
people like to work on consumer
9:44
companies it's why some investors like
9:46
to focus on them we like markets that
9:48
are growing 20% a year the problem is
9:50
growing quickly we like problems where
9:53
people are trying to solve it right now
9:55
immediately we like problems that just
9:59
cost a ton of money so billions of
10:01
dollars right or at least they all add
10:03
up to some billion dollar total
10:05
addressable market we like problems
10:08
where the law has changed the law has
10:11
changed and regulation has put up there
10:13
and now people have to solve a bunch of
10:14
problems you saw a ton of healthcare
10:16
problem healthcare startups were born
10:19
after Affordable Care Act was passed and
10:21
a lot of that had to do with like there
10:23
was now all the sudden opportunity this
10:25
problem that all these hospitals and
10:28
clinics had to solve and then we like
10:31
problem is that people need to solve
10:34
multiple times a day or will use it
10:35
multiple times a day
10:37
Facebook's a good classic example but
10:40
people also really loves to slack right
10:42
because it's like oh I'm gonna be
10:43
engaged and using it multiple times a
10:46
day during the wait work day solution so
10:51
there's pretty much only one piece of
10:53
advice I really have for this solution
10:56
that's the best advice that you can ever
10:58
follow and that is don't start here so
11:01
what I mean by that is a YC we have an
11:03
acronym for a problem that we try to
11:06
avoid or basically an application
11:09
we have to go like oh man I wish they
11:11
had started with the problem first and
11:14
we call it sis P it means solution and
11:17
search of a problem and often what
11:20
happens is like you're an engineer
11:22
you're excited about technology some new
11:24
technology has come on the scene let's
11:26
say it's blockchain right let's say it's
11:30
like react native or whatever the new
11:33
thing is and you're like I want to build
11:34
something with this it's a large reason
11:36
why you start working on a side project
11:37
and then you go like okay what kind of
11:41
problem can I solve now I'm gonna use
11:43
this no matter what and then you try to
11:45
shoehorn a problem into the solution and
11:48
what ends up happening is that's a much
11:49
more difficult way to grow the company
11:51
it's not impossible for companies to
11:54
grow this way it's super inefficient
11:55
it's much better to be like let me see
11:58
what problems people have and then I
11:59
will use whatever is necessary to solve
12:03
them and therefore it's much more likely
12:05
that you will grow as a result because
12:07
the other way around is you might have
12:09
to go and try to drum up the problem or
12:11
you have to like brand the problem as
12:13
something that people have and it's so
12:15
much more difficult you end up growing
12:16
much more slowly as a result so look at
12:19
what you're building right now or look
12:22
at the reason why you're trying to do
12:23
this startup and is it because you only
12:25
care about the technology and building
12:26
something in that or have you started
12:28
with the problem you go like I'm gonna
12:30
do whatever it takes to solve people
12:32
users customers issues the last one's a
12:37
little tricky it's what is the inside
12:40
what's the reason why this solution is
12:43
going to work and this is where a lot of
12:45
companies sort of get tricked up because
12:47
it's really about like what is your
12:48
company's unfair advantage right why are
12:51
you gonna win versus everyone else why
12:54
are you going to be the fastest one to
12:55
sort of grow because that insight is
12:58
what's needed for the investor to choose
13:00
you over anyone else and it has to be
13:04
related to growth you have to have an
13:06
unfair advantage that explains why
13:09
you're going to grow quickly if it's not
13:10
related to that then it's not it's not
13:13
going to be something that an investor
13:14
is going to find valuable and the last
13:17
one is you need one you you can't just
13:19
be like I have a problem I'm solving it
13:21
and it had
13:22
no explanation why without that last
13:24
sort of explanation I can't use my
13:26
imagination
13:27
I can't evaluate just solely on how well
13:30
you've thought through this problem all
13:33
right and so let's go through the types
13:36
of unfair advantages that your company
13:39
have so there's five different types and
13:42
companies do not have all of them really
13:45
great ones surprising not surprising
13:47
we'll have all of them and we'll go
13:48
through two examples but you want at
13:52
least one and it's nice if you can have
13:54
two or three but for most of you it's
13:56
probably just one so the first one so
14:00
how do you know if you have a founder
14:02
unfair advantage and so all of these
14:04
will be connected to numbers actually
14:06
which will help this make this really
14:07
easy it's like are you one in ten of all
14:11
the people in the world who can solve
14:15
this problem are you super expert and
14:19
99% of the people we find out why see do
14:22
not fall into that category and so if
14:25
you think it's like well I'm a product
14:28
manager at Google there's a lot of
14:30
product managers at Google if you say
14:34
you're an engineer at Microsoft there's
14:36
a lot of engineers at Microsoft it's
14:37
like great but it's not one that will
14:40
make me think oh you have a grater on a
14:42
very understanding someone else if
14:44
you've done a PhD and let's say you've
14:47
done it on some kind of crazy biotech
14:49
research and you have like a special
14:51
patent to be able to cure some kind of
14:54
disease then you have a founder
14:56
advantage your market is it growing 20%
15:02
a year like by default if you just build
15:05
the solution in the space you should
15:07
just automatically grow because you're
15:08
just following a trend if this is your
15:12
only company advantage then it's one of
15:16
the weakest ones that you could have it
15:17
is great to be in that space but you
15:20
want to have something in addition to
15:21
this like you're gonna do like better
15:23
than average because you've picked the
15:25
right problem space and the right set of
15:27
customers that want your problem but
15:30
again if you're in a market that is
15:32
stagnating or shrinking
15:34
then you're gonna have investors worried
15:35
about the long-term viability every
15:38
company as a result product so super
15:42
simple is your product 10x better than
15:45
the competition
15:46
if it is then you potentially have an
15:49
unfair advantage and has to be very very
15:51
clear someone should be able to look at
15:53
your product and go like oh this is
15:57
so much better than everything else I've
15:58
ever seen it is 10x faster it is 10x
16:01
cheaper etc and if it's not an order of
16:03
magnitude let's say it's just like 2x or
16:06
3x again that's nice but it's not enough
16:10
for an investor to go like oh this is a
16:12
slam dunk in regards to that 10x product
16:15
and showing that you are able to have
16:17
that a later in startup school going to
16:20
do a lecture on pricing and we'll talk
16:21
things about cost and value and that'll
16:24
help you sort of better understand it's
16:25
like oh how to better prove out that 10x
16:28
multiple using sort of metrics and
16:30
numbers and pricing acquisition so a lot
16:34
of people think that if you go to
16:37
investor and you've done a bunch of
16:38
Facebook or Twitter or Google Ads and
16:40
you show your CAC and an LTV that you
16:44
were able to prove that you have a
16:45
sustainable sort of acquisition model
16:47
and I want you to know that if paid
16:48
acquisition is the only way that you
16:51
were able to grow your company then I'm
16:53
gonna discount that channel of growth
16:56
greatly that is because if you actually
16:58
get really popular you can actually
17:00
start being someone significant let's
17:03
say becoming a hundred million dollar
17:04
revenue company then you're gonna
17:06
attract a lot of competitors into the
17:08
space and that advantage is going to
17:10
quickly dwindling is a really good
17:14
example of this almost all their
17:15
acquisitions isn't paid and then once
17:17
they ate through that there's almost
17:19
nowhere else for them to sort of go you
17:21
want to find acquisition paths that cost
17:24
no money and my favorite companies the
17:27
ones that become really great are the
17:29
ones that can grow by word-of-mouth this
17:31
is a good percentage of the way they
17:33
grow and so in the early days of your
17:36
startup if you don't have any money
17:37
that's actually very great way of
17:40
exercising how do I grow this without
17:43
having to pay for it and so in the
17:45
beginning we tell you to do things that
17:47
don't scale
17:48
but this is what you sort of want to
17:50
accomplish it's like do I have an
17:52
advantage that is free and the last one
17:56
is do you do you have a monopoly and so
17:59
we don't mean this in the monocle
18:02
monopoly game sense so we made it as
18:05
like as your company grows is it more
18:08
difficult for you to be defeated by
18:10
competitors do you get stronger and so a
18:14
good examples of that are like companies
18:16
with Network effects in marketplaces
18:19
where marketplaces where it tends to be
18:22
a winner-takes-all a one company will
18:24
tend to win and network effects is just
18:27
basically as my network grows the
18:29
strength of my company and the value of
18:32
the product or service also grows with
18:34
it
18:34
not every company has it well when you
18:36
do have that it works out great there's
18:40
something to keep in mind also other
18:43
things I'm looking to believe about a
18:44
company and that is something that trips
18:47
up a lot of founders and so there's two
18:50
types of beliefs that I have about a
18:51
company and so there's the threshold
18:54
belief which is like what's the default
18:56
just for them to even succeed so
18:58
oftentimes for me it's like oh then
19:03
building it can they even built it
19:05
that's a threshold belief if they can't
19:07
even build it none of it even matters
19:09
and so to me that question is not the
19:11
most important what will determine
19:13
whether I'm gonna win the lotto
19:15
it's a miracle belief that like oh my
19:17
god if I believe that they can do this
19:20
that actually going to be able to take
19:22
off really well and sometimes are really
19:23
simple so if you are heavy engineering
19:27
team or doing a B DB or Enterprise
19:29
startup again the default is you have to
19:31
build it so if you can't even build it
19:33
then it's not even gonna work so I don't
19:35
spend actually a lot of time looking at
19:37
that for me I'm trying to figure out
19:39
success will be determined by how well
19:41
you can do sales how well you can tell
19:44
the story how well you can actually
19:46
convince customers and work through a
19:49
sales process I want evidence that shows
19:51
that you know how to work through that
19:52
and make that happen and so all of my
19:56
work with most of those companies is
19:57
like not working on product it's like
19:58
hey alright let's prove this other thing
20:01
if you have that that'll be the thing
20:03
that actually will help people go like
20:05
oh they have this super combo let's
20:10
go some some quick examples so YC is a
20:14
good one because we like to think of YC
20:15
as a startup so the problem the way we'd
20:19
word this is it's hard for founders to
20:22
raise money without knowing someone in
20:24
venture capital so at the time it was
20:26
started you basically had to be an
20:29
insider it was the only way you can sort
20:31
of get money and that's super sucked and
20:34
the solution that basically Paul Graham
20:37
came up with is like investing companies
20:40
through an open application you don't
20:41
need to know anyone you just tell us
20:43
your idea tell us a bit about yourselves
20:45
and that should be enough for you to get
20:48
funding now there's a bunch of unfair
20:52
advantages that YC had number one the
20:56
founders are pretty incredible so Paul
20:58
Graham had wrote a textbook on Lisp RTM
21:02
like wrote the very first worm he's like
21:03
amazing programmer and they had built
21:06
and sold the first SAS company that was
21:10
via web to Yahoo so they were kind of
21:13
experts at both evaluating technology
21:15
and also understanding kind of starts in
21:18
that whole process the market basically
21:23
believed that future billion-dollar
21:24
companies would be technology companies
21:27
they'd be powered by software and the
21:30
wonderful thing about tech companies
21:33
especially at that time was Moore's law
21:36
was making it cheaper and cheaper for
21:37
software companies to be started and
21:40
they furred they needed a whole lot less
21:42
money and he could make a lot more bets
21:44
as a result the product so basically you
21:48
know they pay the founders that come for
21:52
three months they get some advice they
21:54
work on their product for a relatively
21:56
small amount of money and then at the
21:59
end of the time instead of offering
22:01
co-working space they work from their
22:03
own home and then they pitch to a bunch
22:05
of different investors and the idea was
22:08
like that would be so valuable to a
22:10
potential founder who had no connections
22:12
that it would attract a lot of really
22:15
great
22:15
or a lot of people who were hungry to
22:17
get into the space acquisition so most
22:21
people don't realize this is like PG was
22:23
able to sort of build up YC and attract
22:26
the right talent because he had a huge
22:28
reach or audience when he got started he
22:30
had written that textbook yes but he
22:32
also brought all these popular online
22:33
essays and had a large audience of his
22:36
target users hackers to come and
22:40
evaluate his product and you can acquire
22:42
them relatively cheaply just by baking a
22:45
website and letting it be known and then
22:47
the last one is something he didn't even
22:49
realize when he started YC and that is
22:51
that as the YC alumni network grew it
22:55
got more powerful and more valuable over
22:57
time the results are funded 2,000
23:01
companies there's 4,000 founders they're
23:03
the CEOs of some of the biggest
23:05
companies in the world they are over 15
23:09
companies worth a billion dollar a
23:10
billion dollars
23:11
there's 93 companies with over 100
23:13
million dollars and our total market
23:15
capitalization is over a hundred billion
23:17
dollars I'm gonna do one more example
23:19
with you guys so we flew this was my
23:20
startup and it's an online forum and
23:23
survey builder and so for us it was
23:25
basically every website needs to collect
23:29
some kind of data at some point but you
23:31
need to know how to code or hire a
23:34
programmer to be able to do it and so
23:37
the solution was to build something that
23:39
what they see is what they got like
23:41
basically a drag-and-drop visual editor
23:43
and that any non-technical person sort
23:46
of create and then if we did that we
23:48
would solve that sort of specific
23:50
problem so very quickly the market it's
23:53
kind of ridiculous like we were asked in
23:55
the early days calculate our Tam but it
23:58
was like every website needs I don't
24:00
really understand like what website
24:01
doesn't need a forum eventually it
24:05
allowed us to grow really really quickly
24:07
our product was easily shown to be 10x
24:10
faster and it's against the direct
24:12
competition with other builders because
24:14
it was so much faster to do the drag and
24:16
dropping and visually see and then
24:18
usually a hundred X faster than like a
24:20
lot of traditional routes for very
24:22
custom forums for hiring a programmer
24:24
and very cheaper because we had this
24:26
freemium model
24:27
that also led to an acquisition unfair
24:31
advantage and so we actually had started
24:33
off with building a blog and building an
24:35
audience as well
24:36
started off with a hundred thousand
24:37
developers subscribe to our blog we
24:39
launched it out to them after building
24:41
up that audience for a year and that's
24:43
actually what we applied to YC with it's
24:45
like we had built up this audience we
24:47
had proved that all these other things
24:50
are like in place and then part of the
24:53
other acquisition model is that you
24:54
could embed these forms on people's
24:56
websites and then our users basically
24:57
spread our form and software for us as a
25:01
result so we never had that hire any
25:03
salespeople and the results are our
25:06
product was used by every industry
25:07
market and vertical you can imagine the
25:10
homes of super large companies and we
25:12
did this with a relatively small amount
25:13
of team when we were acquired we were
25:16
weird outlier compared to all the other
25:18
acquisitions the average company raises
25:21
only raises like twenty five million
25:23
dollars before an exit and this is their
25:25
average return and for Ruffo we only
25:28
raised 118 thousands for the whole life
25:29
of the company and our returns are over
25:31
thirty thousand percent okay so in the
25:37
end this is very simple exercise but
25:38
super enlightening once you actually go
25:41
through it and to try to figure out that
25:43
narrative and story so I turn it back to
25:46
you go through and try to answer those
25:49
questions about problems solution your
25:50
unfair advantage figure out where are my
25:53
holes do I have one and then the
25:56
question becomes Oh what do I need to
25:59
prove what do I need to work on to make
26:01
that happen your startup idea is a
26:04
hypothesis about why you're going to
26:06
grow quickly for our next lecture we're
26:09
gonna talk about the first ways we try
26:11
to prove it out so to basically test our
26:14
hunches and that is by talking to users
26:25
you