How to Build a Product II, Aaron Levie – Box – CS183F

How's everybody doing? Is everybody really pumped up to
talk about enterprise software? Good, that's what you want to do today. So I'm Aaron Levie CEO, co-founder of Box. And what I wanted to do was share
a little bit about our story at Box, and how we started building our company. We actually didn't start as
an enterprise software company, we eventually changed our
business model and our path. But I want to talk a little bit about
how we got to that decision, and how we started building up our product and
our company.

And then, if, and I really hope this
is true, if many of you go out and build enterprise software companies. Maybe there's some lessons that we've
learned that it took us a long time to learn, that maybe you can
learn a little bit faster, because of some of the things
that we've gone through. So and I think that with some time,
we can open up for questions or what not. So is that all cool? Agree with that?
Okay, this is going to be highly interactive, so we've gotta talk to
each other, you know what I'm saying? >> Gotcha.
>> Good, okay.

>> [LAUGH]
>> I love it when the least interactive part is when I call out that
we need to be interactive. But okay, so
here's how we have been building for the modern enterprise and maybe some
lessons that you can take with you. So we started Box in 2005,
and we were actually sophomores and freshman in college at
the time when the idea got started. So this was about 12 years ago. And what was happening was everywhere
that we were working from, everywhere that we were
trying to access files from, anybody we were trying to share data with,
was just too hard to do this. So we had to have these USB thumb drives, hopefully you've never
had to use one before.

You have USB thumb drives, you're emailing
yourself files, you have FTP accounts. And it was just way too complicated for
just getting daily work done, just way too slow, too cumbersome. And at the time, I actually did an
internship with a company that also used enterprise software, for sharing and
collaborating and managing content, and that technology was also way
too complicated and complex. And so this is back in, really we
got the idea in kind of late 2004. And what was happening was we saw this
confluence of major technology trends, that we felt could go and
solve this problem of, why is it so hard just to share files and
get work done with other people? And so we saw that there would be
a handful of major technology trends that could ultimately solve this problem of,
working with our information, working with our data from anywhere. So the big three,
though that we focused in on, the first was mobile ubiquity, and
now obviously this is a standard trend.

No software company would start without
a mobile strategy, but back in 2004 and 2005, it wasn't the case. But we had this vision, and we had this
foresight where we thought back in 2004 and 2005 that over the next decade, the entire planet would start to
be using Blackberrys everywhere. So that was our vision, and so we were
wrong, fortunately, we got better devices. But we knew that mobile was certainly
going to be a really powerful opportunity in the future, and was going to change
the way people worked with information.

So we're glad that it ended up being
a little bit of a better platform. We also saw this trend of
the cost of computing and the cost of storage was
dropping pretty rapidly. So it wasn't called cloud
computing at the time. But you could see that there was
this ability to buy lots and lots of infrastructure at scale,
centralize it in one place, and then be able to build applications
on it over the Internet. And so we sort of figure why is
it that people are going out and building their own infrastructure and
their own data center, when a software provider
can do this all themselves? And get the economies of
scale to serve lots and lots of customers with that technology? So we said okay, at some point in time, computing is going to reach this point
where it's really, really cheap and inexpensive for anybody, and so
thus it should be in the cloud.

And then finally,
we figured that as more and more people were connected
to the Internet. As more and more people and businesses
were working on mobile devices, and collaborating more and more, that we
would become much more interconnected. And so businesses that once had to
have all of their own people, and infrastructure and competencies
within their own organization. Would eventually be able to be networked
with partners, and clients, and vendors and colleagues all around the world,
and be able to get value from everyone. Which would mean, that you'd need
software that would be better at letting us interact with one
another over the Internet. As opposed to just people with
our own corporate network, which is where our market had
traditionally kind of been focused on. And so we took these three big trends,
everybody on the Internet, everybody on mobile and really,
really cheap computing and storage.

And we said, what would we design if
those conditions were true in the future? What would we design to build a really,
really simple way for people to share files and
securely access their information? And so we launched this thing, it was
actually called Box.net at the time. And we launched this site, and the idea
was really simple, you could upload your files, and you could access them from
any web browser, any mobile device. And within a few months of launching, thousands of people were signing up for
it. And we had worked on some mechanics to
make sure that was as simple as possible, so we had way easier sign up flow. It was just focused on the end user,
so in traditional IT environments, you usually had to go to corporate IT, or
you had go to corporate security to be able to be able to get
access to technology.

We just let anybody sign up with their
own information, their own email address, which was actually pretty disruptive
at the time to traditional IT. So anybody could go online, they could get
a couple gigabytes of storage space, and be able to share and
access their information from anywhere. And that was the original idea. And what we sort of ultimately realized,
that we actually had no clue about, because we were just in
college at the time. Was that enterprises all around the world,
and companies of all sizes, were just used to really,
really bad technology. And the bar was just really,
really low for what you could do.

Our first application was
pretty horrendous, but the bar was so low that all you had
to do was solve some basic problems. And then end users within those IT
environments, within those enterprises, would all of a sudden want to
adopt your technology. And so what we discovered was that
people all around the world, and companies of all sizes. Whether it was a Fortune 500 company that
was spending hundreds of millions of dollars on technology,
or small businesses, everybody had this interest and
demand for really, really simple tools. So one of our competitors at the time,
the primary product you had to use for sharing information looked
a little something like this.

This is a screenshot from
one of the products, and it's pretty horribly designed technology. I don't know even what it says,
it's just a screen grab from the Internet. But really not elegant software for
working with other people and collaborating. And so what sort of started out for
us as a really simple way to share files, it was that dead simple of a proposition,
actually turned out that it was the sort of crux of how people
could be working in the future. And it was a much easier way for
people to be able to work. And so, we're getting a little security
prompt, of course that has to happen. Okay, so what started out as
a really simple way to share files, turned into this new way of working. And we weren't trying to solve this vast
problem of how can companies work in a modern way? All we were trying to do was build really
simple technology that makes it easy for people to share files and
access information from anywhere.

And then we eventually
evolved the strategy. And things were growing so
quickly that we basically decided, like any classic start-up story,
we decided to get in a minivan And drop out of college and move to
the Bay Area to go and build the company. And there's always this part in the story,
where if you kind of roll back in time, so this was us living and
working out of a renovated garage. So that thing behind us is actually,
technically, actually zoned as a garage, but we were living out of it, so
it was actually pretty illegal. But we were just the four founders,
just going out, and we had all dropped out of college and
building at this. What we didn't yet fully now was this
enterprise software company that eventually would become a whole
enterprise software platform. And it's funny because
when you're dropping out, what you imagine the glory of, okay I'm
leaving school, this is going to be so glorious, think about all the people
that dropped out before us.

You have Bill Gates, Bill Gates dropped
out and this is going to be so exciting. And then, you've got Steve Jobs and
he dropped out and this is super exciting, and Michael Dell dropped out and
that's super exciting, but nobody ever really remembers this guy. >> [LAUGH]
>> And so we always sort of think about the glorious stories of dropping out,
but actually, I don't even know this guy dropped out, but it looks like he
must have or had to do something.

>> [LAUGH]
>> But we were so pumped, we were like, we're dropping out of college,
**** this, and we're going to go start
a software company. And we were in this position where,
we were living and working, and all together, and we were iterating
on our business, over and over again. We were pivoting almost
every every single day. So if you've ever done a startup, what you
find is like maybe every 24 hours you're sort of reflecting on your
business model and your strategy.

And is it going to work,
do you think it's going to scale, are the economics actually
going to play out? And so, we got to this juncture where we
basically had to say: do we want to be a company that serves consumers or do we want to be a company
that serves businesses? And the reason that we got to that
juncture was we had consumers that were spending, I think at the time, I think the most you could spend
on our product was $9.99 a month. So pretty standard consumer
software payment structure, so about ten bucks a month. But what we saw was that companies like
Google, and Facebook, and Apple, and Microsoft eventually would be giving
everybody free storage on the internet, and that you really wouldn't
want to be spending money for storage on the internet. And let's maybe as just a quick survey,
who actually pays for more storage in one of the storage
solutions that you're using? Maybe you can get a, wow, okay, all right,
you guys just create a lot of data or something, what's going on here? Okay, so who doesn't pay? Let's just see the, okay, right,
so not a great business model, when you only have about maybe 35%, or so, of the market that's willing
to upgrade for your solution.

You want to be able to reach
a broader population than that. So what we said was at some point Google,
Facebook, and others were going to give away unlimited
space, and it was going to be really, really hard to upsell people over time,
so that was on the consumer side. On the enterprise side though,
what we realized was, actually, companies were spending millions and millions of
dollars on software that we thought we could do with at a smaller amount of
spend and for a fraction of the price. So companies on one hand
were spending millions or tens of millions of dollars, and we thought we could deliver it to them for
maybe hundreds of thousands of dollars. So you could have almost a ten or 20x improvement on sort of
the cost of the service. On the other hand, we felt that at most we could charge
consumers maybe $10 per month.

And that would actually have
downward pressure over time, just because of the economics of the space
where storage is getting cheaper. So, it actually took us a surprising
amount of time to come to the conclusion that we needed to go in the enterprise
because we still were like, well yeah, but what if we could serve advertising
on the files, or something? And it just turns out that, you really
don't want to see ads when you're looking at your tax documents,
or your photos. And so, we didn't think that the
advertising model was going to work for consumers.

But we did decide that, actually,
if you could build really, really great enterprise software that
could solve all of the enterprises problems around managing information and
collaborating around files. And you can build one platform for
being able to do that, that maybe you could actually compress
some of the economics in the market and be fairly disruptive to
the traditional players. But we sort of stepped back and we said,
okay, if we're going to do this, are we really going to build
an enterprise software company? We were 23 and 22 at the time,
the founding team, and we had a few other employees.

But, it's really like, there actually
like a long list of things that you could do that aren't very cool, like
somewhere near the bottom of that list, when you're 23 and 22, is starting
an enterprise software company. So literally, if you just had a quadrant,
a McKinsey quadrant, of the lamous thing possible to do, it's
starting an enterprise software company. So we were pretty sort of
distraught of this idea, we were really going to go and build
a company, like an Oracle or an SAP or something, is that really going to be fun? Because think about what we think about,
when I say the words enterprise and B2B, what you should be thinking of, if you
have some experience in the spaces, you should be thinking like, okay, it's slow
companies that don't innovate quickly.

The technologies way to expensive,
like you're just selling to the buyer, you're not really selling to the user. The technology itself is super complex, so there's really no incentive structure to
make really simple enterprise software. The reason for that is that because the
people who have the budget don't actually have to deal with the complexity of
the technology, so they don't actually prioritize simplicity and user experience
when they're making buying decisions. You have all this, you have to deal
with frustrating sales reps, so if you're on the other end of
buying enterprise software, what ends up happening is you're just
getting called all day long by sales reps, and for
software that you basically don't want. There's unfriendly sort of policy terms,
so it's always sort of terms of services that
favor the vendor, not favor the customer. So the very structure of the contracts in this industry are totally
not favorable to the client.

Software design is pretty horrible, because again, if you don't have a market
that is prioritizing user experience or simplicity, then you can't get the best
designers, because the best designers don't want to work at companies that
aren't building really simple software. And so, then what's the point
of having great design? And so, then the design ends up sucking. You get really infrequent updates, so
it turns out when you sell to enterprises, this actually makes sense,
enterprises want a lot of predictability. They don't want their technology changing
on them at the same rate that Snapchat or Facebook is able to change
their software for us. Because if I'm a CIO,
Chief Information Officer, of a 20 or 50,000 person company, and my software
vendor is changing all the time, that's going to make it so
I have to go retrain my employees. And so, what that then means is that
the vendors, the software vendors, don't want to update their software and don't
want to update their technology a lot, because the customers want
to be infrequently updated.

And so, that then slows down
all of your innovation. Imagine working at a company where you
ship product updates maybe once a year, or once every three years,
that's pretty annoying, so you're not going to get
the best innovation. You have all this downtime, just cause it's like really creaky
technology that always has like these wierd problems that you run into,
corrupted data, so these are the things. The list could be about five times longer,
if you were to make a list, of like what words describe
the enterprise software industry, this is just sort of getting started.

And it all sort of comes,
if you kind of took it all together, and you sort of rounded it all together. Basically like the picture that
people have in their head, when you say enterprise software
is basically like this, it's just your vendor
just being angry at you. Actually, it looks a little
bit more like this, it's a little bit more pirate-oriented. And this is basically what enterprises think of when they think
of enterprise software companies. When they think about
enterprise technology vendors, at least ten years ago. So we said, okay, if we're going to
pivot the company and we're going to forego all the fun, and all the exciting
times we would have as a consumer company. Then we have to find a way to build
an enterprise software company that operates and acts and innovates and
moves like a consumer company. If we're going to forgo the consumer side,
and serving just consumers
in their personal lives.

We still have to find a way to build a
company that feels like a consumer company that just happens to sell to enterprises. And so then we made a list of basically,
what are the things that are changing in the technology world that allow for you to
build a company in a very different way? And this is, I think what makes enterprise
software so exciting for all of us today. Which is why it might not still be
the number one coolest thing to do of all time, but it's at least
getting in the top 300 options. So, we said, okay, if we're going to build
a different enterprise software company, what is true today? And this is back in 2006 and
2007, what is true in 2006 and 7 that allows us to build a different
kind of enterprise software company? Well the first is,
let's look at the old way.

So the old way was,
you had really complex technology. It was all about the sales,
what was your distribution channel. So you had to have salespeople as your
primary way into getting into customers. Because customers couldn't access
your technology on their own, they had to install it. To install it, you had to have
a salesperson go out and try and pitch to the customer that
they should start using it. And so all of your distribution was
constrained by the number of sales reps that you had. Technology was only available
to the biggest companies. And so
this ends up being a pretty big flop, because if you're trying to go have
a massive impact on the world.

But only the Fortune 500 can actually
afford to use your technology. Because nobody else has
the infrastructure, or the apparatus to be able to serve large,
to be able to install the technology. Then really you're not having that
big of an impact on the world, you're just having a big impact on
the top enterprises that are out there. We saw a lot of flaws with how technology
sort of solved the security problem. So, when you buy enterprise software ten
years ago, you bought the infrastructure, or the hardware. Then you bought software, then you bought security software to
secure the software and the hardware. [SOUND] Bless you, and so you had all
this technology that you had to have, buy from lots of different vendors. And it was the customer's responsibility
for putting it all together. And so that meant that there was a lot
of security vulnerability inside of those platforms. There's also a thing where
the customer's bearing the risk.

Because what happens is the customer
is buying the licenses for the software up front,
regardless of what they use. And so you have this sort of asymmetry
where the vendor always wins because the customer is buying the software
in perpetuity essentially. And what's happening is even if
the software doesn't work for them, the customer's already paid. So you'd see all the time, you'd have
these software projects that enterprises would have where they'd spend $10 million
on software and they'd never install it. Because, it wasn't what they wanted and
there's no recourse. Once they bought it they
can't do anything about it. So pretty fundamental
broken business model. And then finally, the platforms
themselves were fairly proprietary. So they were closed systems,
they weren't integrated and open to other platforms, and so
it was very proprietary technology. And so we said well, what if there's
a new version of each of these things? What if you could make it so
the technology was simple and elegant? And we always focused on the end user, and we never lost sight of
the power of the end user? What if the product was actually pulled
into organizations by those end users, as opposed to you having to
only do tops down in sales? What if the software was so good and so freely available that anyone could bring
it into the workplace on their own? What if it was available to people
inside of organizations of all sizes.

What if it could be a small
business anywhere around the world? What if it could be a large enterprise? What if it could be
a three person startup? So how could you make that technology
available to anybody of all size? How could you make sure that security was
baked in, so the customer doesn't have to buy all this other technology and
wrap it around these tools.

What if the vendor could
bear more of the risk? So if you're a customer of ours and the technology is not working,
you're not going to renew. So what if more of that risk to making
sure you're successful is actually something that the vendor has to focus on? And then finally,
what if you had to open systems? What if you connected to all the different
enterprise software that people were using, and you didn't have a closed
platform that locked in data? So we kind of said, okay, there's this old way that the Oracles
of the world built their technology. And then there's a new way, and
what if we could describe each of these new things that was
different about technology today? And go build an enterprise software
company that really solved this core set of issues.

And that's essentially what we've
been working on for the past decade. So we said, here are the new
conditions of this landscape. Let's build a software company for
the enterprise that is built and operates like a consumer company, and
that's been our strategy since about 2007. And fortunately, what's cool is that
it has played out in favor of a lot of the things that we've been talking
about or that we identified. So we now have about 71,000 customers. We're in about 63% of the Fortune 500. And this was generated mostly because
people just at work wanted to bring in better technology. And then we eventually were able to
sell to the broader organization. But it was really about solving for
that user. Pulling the technology into the workplace,
and then ultimately solving a bunch
of problems from there. And the exciting thing about what has played out in the cloud is that you
have companies of all sizes and in all industries that are now transforming
the way they work, because of the cloud.

And so we can work with companies
as wide ranging as Pixar, where they use Box to be able to share and
collaborate around movies. To Nike, where they use Box for
product and marketing. All the way to Eli Lilly where
you have things like R&D and drug discovery and
various medical discoveries happening. And collaborating around
through Box as well. So what's cool is,
it doesn't matter the industry. Companies of all sizes have realized all
of this potential for moving to the cloud. And we are one of many companies that
are benefitting from this set of trends. And the cool thing, I think,
for all of us and for anybody starting a company
in this space is really, this is just the beginning of a whole
significant change in this industry. So going back ten years,
I thought it was the beginning but it was still pretty slow start to
kind of get to where we are today. But if I look at all of the trends
that are happening in the enterprise, I think the sort of best opportunities
are actually yet to come.

And so this means there's going to be more growth
and more opportunity than ever before. And the reason for this is
basically you have two big factors. First is, most of the technology
still used by enterprises is still that legacy technology. It's still old, it's slow,
it's complex, it's expensive, it's sold by the legacy vendors. And so on one hand you just have
this gigantic market of old stuff that is going to have to get retired,
shut down, and moved to the cloud. That's the first thing.

But the other trend is that every business
on the planet is recognizing that that legacy technology is not going to let
them transform their business models, the way that they serve their customers. The way that they operate and you're
probably familiar with the fact that every single company on the planet at
this point knows they need to modernize their organization in
some significant way. So if you're a media company You're
freaking out because of Netflix. If you're a life sciences company,
you're freaking out because of 23andMe. If you're a financial services company,
you're freaking out because of Stripe. If you're a hospitality company,
you're freaking out because of Airbnb. So every single company, in every single
industry recognizes that you have all of these disruptive forces coming
from the digital industries and Silicon Valley and
broader than Silicon Valley. And that every single company
is going to have to operate and upgrade both the technology that they use, as well as the sort of business
models that they have. And what's cool is if you're building for
enterprises, that means there's an unbelievable amount
of opportunity because all those companies All the Fortune 500 and all the other
companies that need to be transforming are going to need modern software to be
able to work and operate in this new way.

They're going to need new
technology that allows them to be able to actually compete
effectively with the digital economy. So that means for all of us that there's
more opportunity than ever before. Because if I go back ten years ago, there
wasn't actually a significant catalyst for enterprises to go upgrade their IT
systems and their technology systems. But today, that catalyst is just the fact
that every company is being disrupted in sort of every single way
that they look at the market.

And so that just means that you have this
cowless which is If I'm a General Electric or I'm a Walmart or I'm a Proctor and
Gamble, I need to be working with start ups and new suppliers that are going to
make me more effective and more competitive, which means that then there's
opportunity if you're building for B2B. So if I can convince you to
build an enterprise software company here are some
lessons that we have learned. And a lot of these actually
apply probably for most markets. The ones that we've specifically taken
from building from the B2B market but certainly a lot of these apply whether
you're building a consumer app or a medical device company or
any kind of entrepreneurial activity. But here are sort of I think eight lessons
that we've taken from our journey. And a few that I've learned from other
companies over the years as well that I'll try and weave in. So some of the things that
we've been thinking about.

The first is, and this one has just been
the foundational way that we got started, which is start with something
insanely simple and then focus on expanding it over time. When we got started,
if you were in enterprise IT or you worked at one of our competitors
like Microsoft or SAP or at Oracle. And you looked at our software you
literally would have died laughing. You would have looked
at our technology and you would be like that is
the stupidest thing I've ever seen, like all it does is you upload
the file and the you access the file. And it was that simplicity that was
actually why it was so disruptive, because all of those vendors thought
that the problem was we just need to jam in as many features as humanly possible
into one piece of software, and that was correlated with
our value proposition.

That was sort of how enterprise
software companies operated. Mostly because we weren't
focused on the enterprise. We just took the opposite approach, we said actually the fewer features,
the simpler set of functionality in one place is actually going to
be our value proposition. And If you were in classic enterprise
software, you would have looked at our product and you would have
thought it was a toy essentially, you would have been like, that is never
going to serve enterprises, it's not going to serve regulated businesses, it's
not going to serve complex work flows. And what they didn't recognize
was that was just the start, that was just our day one use case. That was the day one thing
we were trying to solve. So, the instinct and the tendency,
and this still happens today like at our company 12 years later when
we've learned this, I think so directly.

You still have this tendency,
anybody in this room that goes out and brainstorms, how can we build
a big enterprise software company? You're going to brainstorm by listing
out lots and lots and lots of things. You're going to list out lots and
lots and lots of features, like this wish list of what if you
could just build all this cool ****? But that is not actually how
you are going to be successful. You are going to be successful
by nailing one use case that just happens to be a use case that
everybody has a massive problem with. And you can nail that one use case,
that becomes essentially the wedge to then use to expand out to all the other
use cases that you want to go and solve.

And so you have to sort of resist
the natural instinct, which is that more functionality, more powerful software is
better, and really make sure that you're focused on one thing that you're
going to do better than everybody else. A great example that I've been
re-reminded of this over and over from is a company called Gusto
that was previously called Zen Payroll. And what they did was,
they just said that payroll software and payroll services sucks in small
businesses, as an example. And so, they built a software
application that was just about making sure your employees get
paid as efficiently as possible.

And then eventually that was so
successful, and they got embedded into so many businesses that they could eventually
expand to more and more HR services, and more and more things that an HR team
has to do with their employees. But they started with that one use case
which is, let's just get payroll right. Make it so successful, so simple, so effective that then you
can build up over time.

So again, when you're starting out, you have to resist the urge to want
to build a lot of functionality. And really focus on one use case
that people are going to be able to easily describe to other people and
then be able to go from there. So that's the first lesson,
and again 12 years in, we're still having to relearn
this over and over again, the amount of product meetings that I end
up in, where we have to be like, we have to remove that feature, we have to hide
that thing, we have to make that simpler.

It never goes away. Even when you're bigger,
you still have to make sure you're really, really focused on keeping things
dead simple and almost deceptively simple in what you're actually
going to go and do for the customer. Because it just turns out that in these
cases, having less functionality gives people way more opportunity than go use
your service, and it's counterintuitive. But the more functionality you have,
the more you're going to narrow the set of customers that ultimately are going to be
able to make sense of your product and want to use your product. That's the first thing. The other major factor is,
and I think that as you're starting your company you
really need to think hard about this one, which is make sure that you're benefiting
from a major technology tailwind.

There are so many companies that
get started where there's no particular macro technology trend
that they're riding on top of. And in some cases even worse, there could
be technology trends that you're sort of resisting, or
that are points of friction for you. And so you really have to think hard
about what is going on in the technology landscape that I am going to be able
to ride that wave with my innovation. Where the growth of whatever
that technology trend is, is something that I can
then grow alongside. In our case, we benefitted from three or
four megatrends. We benefitted from the growth of mobile.

We benefited from the lower
cost of computing. We benefitted from the fact that
everybody was coming onto the Internet. And all three of those trends
were disruptive to our incumbents that were selling traditional
enterprise software. And so we were able to take these
megatrends that we had nothing to do with, take these mega trends, ride on top of
them, and they were just an onslaught of basically disaster for
the traditional competitors in the space.

What are the technology trends that you
think incumbents are not going to be able to respond well to, because they're
sort of economically, strategically, or from a talent standpoint, not able to leverage those technology
trends in the same way that you can? PlanGrade is a pretty
cool example of this. They basically do software For
construction companies and for architecture firms. And builders to be able to
let you do blueprints and other kind of construction projects
on the iPad and the iPhone. And and it's a pretty dead simple idea. But it's going after just the fact that
there's billions of dollars spent every single year on either software
that is stuck on your computer or paper based processes. And they could basically ride
the growth of mobile and the growth of mobile in the sort of field worker
scenarios to be able to grow with that. So they nailed the timing perfectly of,
okay, it was a couple years after the iPad. Now's the time where these devices
are going to start to show up in physical environments.

What is the software that we need to do
to be able to let construction teams and building companies be able
to use that technology? So really make sure that you're
going to benefit significantly from a major tailwind, and don't create
unnecessary headwinds for yourself. Your product, in a perfect world,
should really literally sell itself. And this is, again, when you're
going out and you're building for the enterprise, instinctively you're not
thinking about how can I make the product be that sort of point of viral adoption? Because we're so
used to sort of building up sales teams.

And building up all the sort of customer
relationships, which are still important, because you're still going to have to use
probably sales to ultimately close large deals with clients. But make sure that the product itself
is something that can be adopted, spread, and utilized without
any friction by the end user. Sales really should be used to close big
deals, not to drive product adoption. And that's a big distinction in
the traditional enterprise software world, where sales was, sort of the only way that you could push
out product adoption into companies.

And so in our case,
what we did was we put a huge emphasis on all of our
sharing functionality in Box. So when you had a document that you wanted
to share, we wanted to make it really, really, really simple to be able
to share those files around. So then the software would spread itself, as opposed to us having to just
keep selling it into companies like the traditional enterprise
software vendors were doing. And then you use sales
to close bigger deals, essentially to help sort of navigate
in that customer environment.

And so, Slack today is sort of the perfect
quintessential example of a viral enterprise software product where the very
usage of the service is to then spread to other people because by definition,
you're communicating with other people. And they've really done this
extremely well, which is make sure that you're building a product that is as
easy to adopt and as viral as possible. Every single thing about
that product is tuned for the end-user to be able to be
empowered to spread it on their own. So they can build their own sort
of viral distribution channel. Sales then comes in later. And then you can sell additional services
or capabilities to that enterprise.

And so you really want to have
a bifurcation between product adoption and sales, making sure you're
doing the bigger deals, okay. Next one, spend time, a lot of time,
with your sort of futuristic customers, but don't exactly just build
what they're asking for. And so what happens when your building for
the enterprise, you will always have, inevitably, you will always
have a bell curve of customers. Of sort of how innovative, or
how leaning into the future they are, versus how conservative they are. It's the customers that are sort of
at the first, sort of 5 to 10% of that bell curve of, they're at
the bleeding edge of everything. These are the people that wait in line and
they get the AirPods. And they have an Amazon Alexa
at their desk in their office. These are the customers that you
want to spend a lot of time with. Because they're going to see
things even before you will about how your technology could
be used in their environment. They're going to spot trends of what's
happening in their employee base, and how they want to push your technology
farther than you even would have imagined would have been possible.

The amount of sort of innovation that we've been able to drive because of
what we've heard from those customers, is probably a significant percentage
of what we've ended up building. But the difference is from legacy
enterprise software companies to today. Is back in the day what you would do, is you would just ask them what do you
want us to build, and then you'd go build those things because that would
essentially lead to a sale or an upsell. What you need to do is listen, to what
are the issues that they're seeing? What are the things they
wish your product could do? And then, either work with them,
or come back with solutions, that maybe aren't exactly
what they asked for. Maybe aren't the specific
features that they asked for, but are going to eventually get them
down the path that, really, they're trying to push your product in.

So we spend a lot of time
listening to customers. But not building exactly
what they ask for. Because when you just build
what a customer asks for, essentially you become this sort of
amalgamation of feature requests from every single one of your customers. And there's no way that you're going to
then be able to keep a consistent north star with your product, you're
going to be sort of pulled in lots of different directions by lots
of different customers. But the software vendors that do it
really wrong is by not listening at all ro those early adoptors. They start focusing on the most
conservative companies. They start focusing on the people
that will only adopt if they built all of the features
that they ask for. And that's really how you go wrong in
enterprise software, because then you really lose sight of your vision and
ultimately what you're trying to build. And so we have a bunch of different
groups that we work with to try to learn all of the things that
the customer would want.

So we have lots of advisory boards and
we're getting lists of hundreds and hundreds of requests. But we don't just go and
build whatever those requests are. We spend a lot of time thinking about, how can we connect the dots between
multiple requests or multiple ideas to build a solution that people didn't
really even think about previously? Also another area where software companies
go wrong is, you're going to get a lot of requests from customers that
are sort of contradictory in nature.

Somebody in financial
services is going to say, I want this functionality for my bank. And then somebody in life
science is going to say, I want this functionality for
my pharmaceutical company. And you're not going to be able to kind
of actually synchronize those requests. They're going to fundamentally
require different user experiences, different product functionality. And where a lot of software
companies go wrong is, you start to sort of build all
these features in isolation. And then you sort of zoom
out five years later and you've just built this massive technology. That has, okay we've got this
vertical offering for that industry. We've got a separate vertical offering for
that industry. You can learn a lot from companies like
Salesforce which essentially said, okay, we're going to build these modular
components which can be mixed and matched to create user experience for
different industries, for different lines of business,
for different job functions.

We're not going to bake in all of the
functionality directly into the product, we're going to give customers the tools,
or even our own team to do that
customization themselves. So a lot of companies get into a trap
where you just are mindlessly building lots of features. You sort of zoom out a few years later,
and what you have is this mess of serving a whole bunch of different industries with
a really, really bad user experience. The thing that we still
stay true to today. And this is where it gets harder and
harder over time, especially as the deals get bigger and bigger,
is making sure that at any juncture, you're not ever trading off
the user experience for some Enterprise feature or some feature
that's going to get you a bigger deal.

There's going to be hundreds of times
where you're at this moment where the customer is asking for
something that is antithetical Or going to go against the simplicity
of that user experience. Maybe its a security feature,
maybe its a compliance feature, maybe it's something that they're just
trying to put in the way of the end user. And at every single one
of those decision points, we have made sure that we always favor
the end user over the enterprise.

And in some cases,
that means that we will lose deals. And so, we'll have to say no to
a customer, because by building that feature, we don't think we're going to
be able to deliver the delight, and the user experience that we're looking for
over time. And so, that gets really really painful, because you lose then deals
all throughout the years. So if I look back over the past ten years,
we've lost tens of millions of dollars of business,
because we said no to certain customers.

What we benefited from though,
was a product that many, many years later could serve a much
wider population of customers. That ended up making our existing
customers much more successful with our product. And funny enough, the companies
that once didn't say yes to us, and took their business somewhere else, ended
up being unsuccessful with those products, and had to eventually come back to us. And so, if you're right, and the key
point here is you have to be right.

If you're right and that user trade off
truly was going to be disastrous for that user experience, and you remain true to that north star
of always focusing on the user. Eventually, over the long run and sort of long run can be defined
whatever market you're in. You will end up winning in
that market over the long run. But you're going to lose
deals along the way and that's going to be really stressful for
you, for your team, for investors. But it's really important that you're
crystal clear on the principles of how you're building your product and
what you're going to say no to.

Finally, Find ways, and this kind of goes
back to that technology trend thing. But try and find ways where you can have asymmetric advantages
over the competitors. And what you'll find is,
that your competitors the big incumbents, tend to be locked in pretty significantly
to particular business models. Because they in some cases have
tens of thousands of employees that are built up to build and to deliver
products in a very specific way.

And so if you can find points of
differentiation and disruption, that they can't respond to, because either it
is too expensive for them to respond to. Or strategically, it doesn't make as
much sense for them to respond, or they don't have maybe
the talent to respond to it. That is how you ultimately remain
disruptive to these bigger companies, that's the only way to survive,
as a small start up. Amazon Web Services, we take it for
granted now, how easily you can build a company and just by writing
some code and deploy it on AWS. 10 years ago, 15 years ago, what you had
to do is you had to buy all these servers, you had to put them in data centers,
you had to rack them yourselves. You had to connect them all together, and
that would take weeks or months and cost in some cases hundreds of thousands of
dollars just to get your start-up going.

But, that's the benefit to us
as building software companies. Think about now the difference if you're
one of the vendors that built that hardware. And you sold those servers, and
you sold that data center space. Think about how disruptive
Amazon Web Services was, coming out of nowhere, and
having a completely different model. Where now I don't have to buy servers
anymore because I can just go to Amazon. Amazon did something that was so
disruptive to the core business model of traditional server makers and
software providers. That literally those companies to
this day have not even responded. 12 years later they still have not
responded by being able to deliver these capabilities over the cloud, because of how disrupted it is
from a business mobile standpoint. So, really trying to identify are there
things from a product standpoint, that can be highly disruptive to how your
product can be delivered to clients versus how incumbent solutions were delivered? And I kind of wrapped all of
this up with a Steve Jobs quote.

It's a requirement that you have a Steve
Jobs quote in any product presentation, so I need to meet that quota,
but this is really true, and it's especially true in enterprise. It's really true in consumer, but it's
even more true in enterprise software. Because what happens is
enterprises are just asking for solutions to all their problems. But, in many cases, they don't know what
the right solution is, until you come out of nowhere with something they never
thought would never be possible. In our case,
just in our product, literally, if you took 100 IT people from 100
different companies 12 years ago. And you said draw out the best way for
your company to manage documents, manage files, secure their information. Not in 100 attempts would our
solution have been designed.

Because everybody was so used to
the technology has to really be complex, it has to have all these
interface buttons. It has to have all these features for
it to make sense. We came in with a product that nobody
would have anticipated because of how dead simple it was. And so you can't really go
ask your customers just like, hey what do you want us to do? And then, somehow you're going to
end up with a really great solution. You have to really think hard about
what is possible in the future? What are the things that maybe your
customers aren't anticipating? And you have to build that, and then, essentially have that land
inside of the customer base. So really make sure that you're not just
building what customers are asking for, because no great technology
has been built that way.

Last thing that I'll leave you with. It's my, I think they're my four
favorite books on building B2B stuff, but don't hold me to it. But here are four good books that
I would highly recommend reading, if you're going to build
a B2B software company. It's all about, it sort of gets you
into the mindset, and have the frame of reference of how to think about growth,
disruption, building a new market. And then ultimately going out and building
a company can scale in this space. So if I had read these
four books 12 years ago, I think we would've not had to
learn these lessons so slowly. So I would highly recommend
checking these out. So with that I would just say good luck,
please do not compete with me. >> [LAUGH]
>> I don't need more competition. Do not use my advice
against me in any way.

And hope to see you in the market,
so take care. >> [APPLAUSE] >> [APPLAUSE]
>> Okay, and I don't know what happens next. But, what? >> [INAUDIBLE]
>> I have time. I have like five minutes. But I don't know how it works with people. Okay, I can answer questions or
you can leave. Either way, okay, yes. >> So what are some ways that you would
actually suggest [INAUDIBLE] lot company. because similar to you when you
were actually back in school that may not have had [INAUDIBLE] companies. Big on that. >> Yeah it's hard. because I haven't had to
do what you are asking. Which is the question was how do you
discover issues that enterprises are having without necessarily
working inside the enterprise? You know I think one is you spend a lot
of time with people that everyday they are going into the office. And they are running into
all of these problems.

And I would just go
interview like 100 people. Just like walk me through your day,
what are all the places where you are less effective, less efficient,
having to do repetitive redundant work? And don't ask for technology problems,
just listen to where they're spending their time and what processes do they
keep repeating over and over again. And then you kind of start to imagine
well, how could software go and attack that problem? How can software attack
the problem of meeting scheduling, not a big market but one space. How could software attack
the problem of payroll? How could software attack
the problem of expense reporting? All these kind of things. Are the drudgery of working
inside of a big company. But if you haven't worked in one of those
big companies, you at least have to get as much information as you can from people
that do work in those big companies. And then you can start to see patterns
of this seems like a really big problem. And in our case, if you had
interviewed clients or people that worked in big enterprises 15 years ago and
you said, walk me through your day.

I think you would have eventually spotted
how much time is wasted when people are working with their documents, and
they're trying to share their documents. And they're trying to access their
documents from different devices. If you really peeled it apart, and
you asked all the right questions, you would have actually
discovered this problem. We discovered it because of what we were
doing in our personal lives in college. But you would have also discovered it
by just understanding where people are spending time in the work place. And then saying hey wait a second that
doesn't make sense, because technology today could solve that problem way better
than what they are getting as solutions. And so you can just, as soon as you find that gap that's
where you know you can build a company. >> I was wondering like you
mentioned about writing the pathway, [INAUDIBLE] what do you think
is the biggest high point now? >> Well it's not BitCoin, so
sorry to all the BitCoin startups. BitCoin could be up in the top five or
something, but you know, I don't want to offend all the BitCoin people, so
I'm going to get like hate mail.

Can we edit that part out
about my attack on BitCoin? So, I'm going to say
a bunch of the trends, I don't know how to rank order them but
obviously AR/VR is a big one. I think AR/VR is It's in a challenging
spot because we don't have ambiguous devices yet. And so the question is,
is ARVR like mobile was 12 years ago or is it still another five years away. The last thing you want to do is
be the mobile start up of 2001, because those start ups died.

So, and it's actually really hard to
predict these things, so I don't have any advice on like how do you predict if
it's actually going to happen or not. But there was a bunch of mobile companies,
and funny enough Sam would have good stories around this
back in 2004, building a mobile company. And how hard it was because
you didn't have the iPhone. And so before the iPhone,
it really sucked being in mobile. And that's kind of the state of things
right now if you're building ARVR, which is like,
we know this is probably the future. We don't know if we're looking
at the right era of it yet, so that would be the one caveat I'd just say. It's going to be the future,
we don't know exactly when. AI is a massive tech trend.

It doesn't necessarily benefit
though startups versus incumbent. It's going to be more
pervasive than that so I don't know how to necessarily use it as
an advantage over an incumbent because the incumbents are actually almost
superior in this technology right now. Let's see, you can always expect the cost
of computing is just going to continue to drop. So we thought that was
a trend 12 years ago. Well, It hasn't gone away and it's
actually more powerful than ever before. But I think there's some pretty cool
things, like as an example, and I might have to change my
own even definition of this. Of thinking about
the compounding of two trends. So for instance, AI and computing together are actually
the really powerful trend right now.

So Google just launched this new
service that does image recognition and video recognition, object
recognition inside images and video. So just via an API, you can send
Google an image or a video and it will tell you what's in that image or
video. So that would have taken like 100
engineers five years ago to go build that. And now you get it as
a simple API from Google. So thinking about, wait a second. What happens when you have a bunch of
these new services that are being created on our behalf from Google,
Amazon, Microsoft, etc. What's the new set of problems that
we could use those solutions and build software for.

So, maybe you build a company
that does security cameras. That plugs into Google
computer vision API. Maybe you build a new way to navigate or
search for images and data because of these API's. So, there's a lot of the new problems,
that you wouldn't have even been able to see were a problem before you knew
that this technology was possible. And so now you can start to look at the
world in a different way if you have kind of complete understanding of what all
those API's are that are out there now.

Maybe the last question, yeah. >> I was wonder,
what's your opinion, as a startup, using other enterprise softwares on your
platform, or just building up your own? >> What's an example? >> So, I mean, so
I guess servers are pretty obvious to use, but for instances money transactions,
using something for that. >> Yeah.
Do you have something specific that maybe could be built versus. >> Yeah, I think that the question
is when do you decide what to build versus what to use other
companies for and I think for us, I generally believe that anything
that is not your core competency. You need to make sure that you're using
the best in class outside solution for. And so, the moment that you realized. I'm not going to be the best company at
payments, you better be using Stripe or something else for that problem. You shouldn't be building that yourself. So, the last thing you want to do
is spend any time as a company building out things that the market is
going to solve better than you will and make sure that you are extremely
clear on your core competency.

And I think this is
actually where start-ups go sort of the wrong way a lot of times. Is not getting crystal clear on
what is the thing that we can do that the market is not going to go and
build and commoditize? And what's interesting, and kind of we've learned this over the
years is that's not a static definition. The market is always changing, and so your core competency might
actually be changing you know, every couple of years, because you have to
adapt with the realities of the market. If I got back ten years ago or 12 years ago,
we had that picture of the four of us. We would go into data centers and we were building out our
own storage infrastructure. And we knew that it sucked that we
were building out our own storage infrastructure. Because we knew somebody else could
be doing this better than us, but there was no other solution. Fast forward to today, we have no
problem using Amazon for all of that, we have no illusion that that is somehow a
point of differentiation or value for us.

And so the moment that the rest
of the market catches up and commoditizes something that
you used to be good at. You need to get away from that part
of the market as fast as possible. Cool, all right, I gotta. Good luck guys. >> [APPLAUSE]
>> Take care..

You May Also Like