🔍
Billion dollar startups are better faster cheaper— Lessons from Amazon, Instacart, Uber & Shopify - YouTube
Channel: Garry Tan
[0]
- So, should you work on a
startup idea that already exists?
[4]
Well, Snapchat wasn't
the first messaging app,
[6]
Facebook wasn't the first social network
[8]
and Discord wasn't the first chat app.
[10]
You don't have to be first,
[12]
but you do need to be different.
[15]
Being better, faster and cheaper matters.
[18]
In fact, it matters a lot.
[19]
It matters so much that
if you don't have that,
[22]
you can't succeed,
[23]
your startup will die
[25]
and neither you or I want that to happen.
[28]
So let's talk about it.
[29]
Let's get started.
[30]
(soft instrumental music)
[35]
(upbeat music)
[40]
When we started my startup
Posterous back in 2008,
[42]
it was definitely not the
first blog or social network.
[46]
However, it was better
in one specific way:
[49]
we supported post by email,
[50]
in a way that was really easy to remember.
[53]
If you knew how to send an email,
[55]
well, you could post something online.
[58]
And that was enough.
[59]
That's the thing,
[60]
you don't have to be the first.
[62]
WordPress today runs some crazy percentage
[64]
of the world's websites,
[66]
but it was first released in 2003.
[68]
Posterous was late by five years,
[69]
but we still grew fast
[71]
and had as good a shot as
anyone to build something great.
[74]
It's often a disadvantage
to actually be first.
[76]
Before Instacart,
[77]
there was the flaming
disaster that was Webvan.
[80]
Before Facebook,
[81]
there was Friendster and Six Degrees.
[83]
If being first is
sometimes a disadvantage,
[86]
why do people still look
for things that are so new?
[90]
Here's the thing,
[91]
investors do love novel approaches.
[93]
If you are trying to start something,
[95]
it's an important bias to be aware of,
[98]
and it's going to be
hard to fight against it,
[100]
so don't even try.
[101]
You just have to take it into account,
[103]
people who get a lot of startup pitches,
[105]
will use this as a filter
right out of the gate.
[108]
Is it fair?
[109]
Not always.
[110]
There's a reason for this,
[111]
brand new business
models are usually driven
[113]
by just a few things.
[115]
New technologies,
[116]
new online behaviors,
[118]
new needs,
[119]
changes in society's demographics,
[121]
new capabilities
[122]
and new regulation.
[124]
But not everyone will be lucky enough
[126]
to find one of these brand
new blue ocean spaces.
[129]
Sometimes you have to
wade into a red ocean:
[132]
a space with lots of competitors.
[134]
And if that is you,
[136]
then there's one thing
I need you to have clear
[139]
in your head.
[140]
Is it better, faster or
cheaper than alternatives?
[145]
(water whooshing)
[148]
Let's take these three things in turn.
[151]
First off, being better.
[153]
Here's some great
examples of top startups,
[155]
that got there just by being that.
[157]
Google was just clearly
better than AltaVista
[159]
and all the other search engines.
[161]
PageRank was an incredible advantage,
[163]
it just gave you better results.
[165]
It was clear they had
better engineering talent
[168]
and thus that talent could
build a better product
[171]
that could give a better experience.
[173]
They got there through
technical excellence.
[175]
Dropbox, when it first came
out was just clearly better
[179]
than a USB drive.
[180]
Again, technical excellence.
[182]
It was a breakthrough
experience that just worked.
[184]
Discord, when it first came
out, focused on gamers.
[187]
They had low CPU requirements,
[190]
which meant their initial
users wouldn't get bogged down,
[193]
and they could still play
their high frame rate games.
[196]
That was enough to get
Discord their first users
[200]
and put them on the map.
[201]
Airbnb started with an audience
that had no alternatives.
[205]
They were focused on conference goers,
[207]
when there were no hotels available.
[209]
What else could you do?
[211]
Actually, nothing.
[212]
You basically wouldn't
go to the conference.
[214]
Having a place to stay
was just clearly better.
[217]
That was a great first set of customers.
[220]
Coinbase is a great example
of being better as well.
[223]
I was one of the first
investors in Coinbase.
[225]
And for me,
[225]
it was because it was
clearly better than Mt. Gox,
[228]
one of the first places
you could buy Bitcoin.
[231]
Back then, you had to wire
money using Western Union
[233]
to an overseas bank,
[235]
and it somehow appeared
[236]
on this kind of sketchy looking
website that sold Bitcoin,
[239]
that was originally built
[241]
to trade Magic The
Gathering trading cards.
[243]
That's the entire reason
[244]
why it was even called Mt. Gox.
[246]
Today, Coinbase is worth over $8 billion,
[247]
because they focused on
a fringe and new crowd,
[250]
early believers in cryptocurrency.
[253]
Remember, this was 2013.
[255]
So being better is pretty obviously great.
[258]
And we just ran through a
bunch of tried and true ways,
[261]
to be better.
[262]
You could get there through
pure technical excellence,
[264]
like Google or Dropbox.
[266]
Or you could pick a particular customer
[268]
or situation that's underserved
[270]
like gaming for Discord,
[272]
conference goers for Airbnb,
[274]
or fringe groups like
crypto in 2013 for Coinbase.
[277]
(gentle music)
[282]
Second, being faster.
[283]
At Initialized, I funded Flexport,
[286]
which speeds air freight by
as much as two to three days,
[289]
compared to other methods
of shipping cargo.
[292]
Because they replaced traditional
methods with software,
[294]
their freight insurance can
actually be six times faster
[297]
to resolve a claim.
[298]
So then those are both things
[300]
that shippers care a lot about.
[302]
It's just clearly faster for them.
[303]
Uber, when it came out,
[305]
was just significantly
faster to find a car.
[307]
They focused on matching
supply and demand,
[309]
and built teams to increase
supply very quickly,
[312]
city by city.
[314]
And finally,
[315]
Facebook was just significantly
faster than Friendster.
[318]
Friendster early on was sort of notorious
[321]
for using Java Server Pages
and unoptimized MySQL.
[324]
It was often impossible to even log in,
[327]
and Facebook in contrast,
remained fast throughout.
[331]
Speed matters a lot especially
for consumer scenarios.
[334]
Look, there's a whole
internet of other things
[336]
to occupy someone's time,
[338]
if that particular website is slow.
[340]
Speed matters.
[341]
So those are a few ways to think
[343]
about how you could make
your product or service,
[346]
faster for your customers.
[347]
First, if you're using software,
[350]
you're beating paper and pencil,
[352]
and that makes things a
lot faster by default.
[354]
That is Flexport story to be sure.
[356]
It's easy to be faster
than your competitors,
[358]
when your competitors are
literally using paper and pencil.
[361]
Second, delivering more
supply liquidity like Uber.
[365]
They managed to use software
[366]
to scale one of the world's
largest mobile workforces.
[370]
More drivers meant faster response times.
[372]
There you go, faster.
[374]
Finally, technical excellence
yet again comes out
[377]
when you're looking at what
Facebook managed to do,
[379]
to scale their website,
[381]
especially as the hockey stick
graph went up into the right.
[384]
Faster is a great way to go.
[386]
(gentle music)
[390]
Finally, don't forget about cheaper.
[393]
If you can get the same or
better product or service,
[396]
for much less money,
[398]
customers are smart, they're rational
[400]
and they will vote with their feet.
[401]
Shopify is a great
example of this early on.
[404]
They were significantly cheaper
[405]
than running your own Magento Server,
[407]
and all you had to do is pay
a flat fee of $29 a month.
[411]
Shopify enabled a new
generation of brand and store,
[415]
that just couldn't have existed
without them previously.
[418]
Lyft is another great example
of being much cheaper.
[422]
They were the first to come
citizen-based ride sharing.
[425]
Uber was new, but at the
time only black cars.
[429]
Lyft was originally called Zimride
[431]
and they were already doing car sharing
[433]
over much longer distances.
[435]
When they saw Uber they said,
[436]
"Why don't we, we try to do that,
[437]
"but instead of using taxis or black cars,
[440]
"why don't we use
[441]
"the community we already
built for Zimride?"
[444]
And that resulted in a
service that was far cheaper.
[447]
Uber in fact had to fast
follow them pretty quickly
[450]
because they realized
UberX needed to exist.
[453]
Otherwise, Lyft was just
going to run away with it.
[455]
So cost matters a lot.
[456]
Instacart was grocery
delivery but asset light.
[459]
People had tried to do traditional
grocery delivery before,
[462]
but they had to buy warehouses
[464]
and buy huge fleets of cars.
[466]
Being able to do this
almost entirely in software,
[469]
with a workforce that was
deployed by mobile phone,
[472]
meant that the cost to deploy
the service came down a lot.
[474]
And as a result,
Instacart is now available
[477]
to more than 85% of people
in the United States.
[480]
That's crazy.
[481]
Amazon of course is famous for saying,
[484]
their margin is our opportunity,
[486]
especially when it came to retailers.
[488]
And retailers had 10 to
a hundred percent margin.
[490]
Margin they needed to
cover physical stores.
[493]
Amazon said we don't have that,
[495]
so why don't we pass
it on to our consumers?
[498]
And that's a powerful
way to get customers.
[500]
Robinhood did something very similar
[502]
when they brought fees
to zero for trading.
[504]
And they decided to make
money in a different way,
[506]
they sell data about retail
investors to hedge funds.
[509]
Making your product or
service cheaper matters a lot.
[512]
You could take advantage
[513]
of major shifts in technology cost curves
[515]
like what was enabled by cloud computing
[517]
and SaaS businesses like Shopify,
[519]
or you could look for
emergent new behavior,
[521]
like applying the business model
[522]
of Uber's early Black Car Service,
[525]
powered by a mobile workforce,
[526]
deployed by software to your own space,
[529]
whether it is grocery
delivery with Instacart
[532]
or the first citizen
ride sharing with Lyft.
[534]
Finally, you can just lower prices
[536]
in a way that cuts the margin
for the whole business,
[539]
but gives you massive market share,
[541]
the way Amazon and Robinhood did it.
[543]
(gentle music)
[547]
So that's it,
[548]
I wanted to walk you
through what better, faster
[551]
and cheaper looks like when it works.
[554]
At the end of the day,
[555]
great startups just clearly solve a need.
[557]
It might be a brand new need,
[559]
that people don't even
know they had, like Google
[561]
or a replacement for an
old and antiquated industry
[564]
that should have run in
technology years ago,
[566]
like Uber or Flexport,
[568]
either way you have to be
better, faster or cheaper
[571]
than the alternatives,
[572]
no matter how old or new
those alternatives are.
[575]
And if you aren't,
[576]
then you should find a way to become that.
[579]
Or you need to find a new idea.
[581]
I want to see you succeed as a startup.
[584]
Better, faster, cheaper.
[586]
You can do this.
[586]
(gentle music)
[588]
If you like this video,
[589]
please click subscribe
[591]
click the bell icon
[592]
and leave a comment.
[593]
I'm still trying to reply
to pretty much every comment
[596]
and I read them all.
[596]
So I really appreciate you watching this
[599]
and thinking about starting a business,
[601]
and maybe a startup.
[603]
I'll catch you next time.
[605]
(soft instrumental music)
Most Recent Videos:
You can go back to the homepage right here: Homepage