Garden City Ruby 2014 - Closing Keynote by Prateek Dayal

So hi. Alright, so welcome to my keynote.

All right. Apologies for that.

I've been a Rails programmer for too long,

and I used Scaffold to generate this presentation,

and I kind of, I kind of kept this towards
the end.

You know, I'll fix it in the end,

and I forgot.

So yeah I'm really sorry about that. But,
anyway.

So my name is Prateek Dayal.

Unlike Chad, you know, I do need an introduction,

so I'll introduce myself as I go along.

So yeah, I'm currently the CEO of SupportBee,

and I started working with Ruby in 2007,

when I co-founded Muziboo.

And Muziboo is a music sharing website

and we had this idea,

how cool it would be if you know people had

a place like Flickr for music,

where they would go and upload music and discuss
it,

not just share with their friends.

And we launched Muziboo,

we worked on it for a few years,

we grew it to about,

you know about half a million users. 200,000
uploads.

And as I said, I started working on this in
2007,

and it was a great time.

It just felt like heaven.

So I did electronics engineering in my college
days,

and this was my first real programming project,

and unlike the unfortunate people like you,

you know I never did any PHP or Java.

So I came to the programming world straight
into Ruby,

and it was so awesome,

and I could sit and code all day.

That was the part that appealed to me the
most about doing a start up.

Actually we didn't even call it a start up
back then.

So we just did something,

and I was sitting and coding all day.

And the community was still there actually,

surprisingly even in 2007 in India.

We didn't have a Ruby meet up but we had Open
Coffee clubs,

and we had some start up meet ups

and actually all the people talk about

is programming pretty much, all the bar caps??.

So I remember meeting Sidoo??

and I remember meeting a few other people,

later I met Hemant. And we would sit

and chat about Ruby and you know,

they would give me all these cool ideas,

I would be like, I would be like hey dude,
you know,

how do I deploy?

And they would tell me there's this cool thing

called Mongrel, you should look at it.

And I would go back and check out Mongrel

and I would deploy. So that was a really fun
time.

I was having a great time.

But as you would guess it didn't last.

It wasn't as bright and sunny anymore eventually,

because I had to start worrying about making
money.

And I, you know I would just a couple of years
out of college,

not much cash, so I had to really start worrying
about,

OK, how am I gonna make money?

This is all fun, but you know let's get serious

about this start up thing now.

And I had never done a business

and so I did what you know any sensible programmer
would do, right.

I went to a business guy,

I went to a few business people to get their
advice.

And all the advice that I got was about SEO,

about partnerships, people said oh you're
a music company

you should go partner with the music labels,

you should hire a few sales people, sell music
listings.

Like if, I don't know how long, if you guys
remember,

but in 2007, India was sort of moving from

the offline world into the online world,

and the models were just sort of being adapted
to the online models.

And everybody was sort of like,

everything was a listing site essentially,

and you could just put ads on it.

So I, you know I didn't like this obviously.

I hated doing this work.

Because I loved doing programming.

But I still persisted with it,

I still persisted with it for a few years.

So I did some SEO, you know I tried doing
some sales -

it didn't work. And, you know, I really felt
like this guy.

I felt like I had all these super powers,

like metaprogramming and you know all the
cool stuff

that Ruby had to offer, but I would just sit
and, you know,

tweak page titles all day long.

To get myself ranking. You know I felt really
bad.

And I really felt disillusioned.

I felt really disillusioned with start ups
because,

I said this is not what start ups should be
about, right.

I should be able to do what I enjoy doing.

And that's when I kind of started asking myself,

can I leverage my strengths as a developer
to build a great business?

Can I do what I enjoy doing already

and somehow build a great business

and build a company that I'm really proud
of?

And I started sort of looking around to get
ideas,

and that's what my talk is about.

So my talk is about coding your business.

And I just want to share some experiences
I've had,

some learnings I've had.

Some ideas we are even now working on.

And just get your thoughts, maybe, you know,

it would be nice if you guys could tell me,
like,

what do you think, what if you tried something
similar.

Or even if later you try something

and you know just tweet to me. That'd be cool.

So when I was doing Muziboo, about a year
into Muziboo,

SoundCloud launched.

And you know they launched in a public data
or something.

And they are a nice product.

And they started getting some traction.

But other than working on the core product,

one thing I noticed was that they had this
thing called the API.

And somehow people were slowly starting to
use that API

and people started writing apps on top of
the API.

And if you go to their app listing now they've
got dozens of apps.

And you know they have apps which help them
get new musicians on board.

These are apps which will allow people to
upload directly

from let's say a music recording program.

All these are apps that get some new listeners.

So people have used API to build charts

that are hip-hop charts, specific charts.

So I think the API really, really helped them
grow their user base.

And that was a great learning for me.

In some ways it was you know it was the hard
way that

I was learning, because they were taking off,

but I mean there was definitely something
to learn from them.

And then I sort of started looking around

and I realized that API is definitely

the new business development.

You know today if you want to grow a product,

APIs are a great way to do that.

And it's like, if you look at a lot of other
products,

so GitHub, Twitter, Facebook, Google -

there are so many examples that wouldn't be
the same product

if they didn't have the API ecosystem.

And especially most in the last few years,

a lot of products are actually just APIs.

A lot of products around are simply APIs.

So for example, Firebase - I should use a
clicker probably, sorry.

So Firebase is a tool that gives you a back

and forth mobile of web apps.

Swifttype for Surge, mailgun for delivering
emails.

Plivo from the streets of Bangalore is computing
with

?? (00:06:31:06) head on.

And they both provide APIs for telephoning.

So there are a lot of products which are simply
APIs.

And I think as developers we are really,

really well-suited to do these kind of start
ups.

We already understand our target audience.

We are the target audience.

We, all of us have been to hack-a-thons,

we have worked with different APIs,

we know what it feels like when the documentation
is shitty,

what makes good code examples,

what kind of libraries do you need to kick
start the environment.

We really understand this stuff really well
already,

and I think if you were to start a company
like this,

you're definitely at an advantage.

And so you can build a product

and you know you can get maybe some users

but what about marketing?

That's typically the question that I get asked
most often from programmers.

Like how do you market your product?

How do you get customers?

And there are several ways to market a product.

Most of them don't work.

And are frustrating actually to execute on.

And what I - personally at least for us

and our company I've realized that writing

has turned out to be the best form of marketing
actually.

And what I mean by that is if you look at
developer docs

or marketing copy site or newsletters,

or even delegating great customer support,

it all boils down to being good at writing.

And, I mean that's just how it is.

For example, this is Mailgun's documentation.

If you go through it, it talks about setting
up SPF records,

you know how to get to deployability,

why do you need dedicated IPs,

it's stuff that all of us already understand
and can write.

We just need to be really good at communicating
this to our customers.

So we already understand this stuff.

We just need to get better

at being able to communicate it in writing.

The same way if you look at Firebase's landing
page,

any of us could sit down and write this

and we could build this and we could write
this

and you just need to be good at writing again.

So fortunately there is good news, right.

I mean it's, as developers,

I think it's very easy to improve at writing.

And I'm just going to show you a few examples
of how to do that.

So GitHub. I think one of the coolest things

that GitHub does is that every time you create
a new depository,

it says add a readme file.

And that's a great starting point to start
learning

how to write or communicate our ideas.

How many of you guys have opensource projects.

Even, or just some bit of code online?

And have you written a readme for it?

Oh, quite a few of you have, and quite a few
have not. Good.

One hand stood up very long actually.

So, there you are, yeah.

So if you haven't,

just go back and write a readme for that,

and if you already have, maybe try to improve
it.

Get some, maybe more feedback.

It's a great starting point to start writing

and learning how to write.

The other cool thing, at least in today's
world,

is that product development involves a lot
of writing,

and this obviously depends on company to company

but you can become a company like this

so you can choose to work for a company like
this.

So for example, I'll give you an example from
SupportBee.

So most of our feature requests are enhancement
requests

typically originate in our subcode desk ?? (00:09:41:07),

so some customer, for example in this case
somebody emailed saying,

can we have search word for word trash tickets?

And my colleague I asked who is here,

I asked him you know, do we support this?

So it turns out we support this in the API

but we don't support it in the front-end.

And so what we do is we write this ticket
into -

we have a GitHub app - so the code,

we would write that into GitHub,

and we create an issue.

And other than create,

other than using the issue to actually

just review code or attach code,

what we do is we try to flesh out the feature
completely in issues.

So we are encourage, not only encourage,

sort of we enforce everybody to just discuss

ompletely in issues. We don't even encourage

face to face or voice calls early on.

If you're not able to communicate something

in issues then you should probably go to a
call,

and then also sum it up and put it back in
the issue.

What this does it just forces you to

try to explain to your peers what you're trying
to do.

Sometimes in terms of code,

sometimes in terms of the marketing copy

that you think should be there,

or the subject line of the email.

It forces you to communicate your ideas.

And acquiring that skill also helps you

in communicating those ideas to your customer.

So it's really the same sort of skills.

And once we know what we want to work on,

what we do, we obviously attach code

and then we go through maybe a sprint or two
with that code.

And then finally once we are almost getting
done,

we post screen shots in the same issue of
the finished feature.

And what that does is, it -

one thing is that it really helps us review
faster,

so we can, you know, know that this feature
is done

and can be deployed.

But more than that,

it's also a starting point for writing a blog
post about the feature.

And we have taken the same GitHub issues process

and turned it into user for our blog as well,

you know we power our blogs by JQL.

We create issues for each new feature that
we have,

or if you want to put out a def post about
how we did the feature.

And we use these assets and we write the block
post.

And because we are already comfortable

with this GitHub issues process,

very easy to get feedback.

So just like code we can comment in-line,

we can help people if they're feeling a writer's
block

or if they want some feedback.

It's very easy to give out feedback

and push out this block post.

So we really tried to make this process work
for us,

and it has ended up, it has ended in a lot
of our developers,

or people who've worked with us getting better
at writing,

and in fact for the first time even realizing
that they enjoy writing.

And it's worked really well for us

and I encourage you guys to go and try it
out.

Because the cool thing about this process

if that you can start tomorrow.

You don't have to wait to start a company

to become good at writing.

You know you can do it in your current job.

So next time you have a feature release post,

go out and propose to do it yourself.

Maybe you have a job opening

and you're putting up a job on Haskee

(00:12:23:02 ??). You know, tell your founders

or tell your CEO that you want to write that.

And just practice.

So there's so many opportunities around us

to practice writing in development.

If you build a company like this,

one of the other cool things that happens
is,

it's very easy to recruit a developer.

Because developers love hanging out with other
developers, right.

That's why we are all here today

and will be here tomorrow.

And so it is a very natural fit there.

And I've been talking to a few developers

in the last couple of months

and I realize that most developers hate these
block,

you know, job posts written with the words
ninja,

or, like how many of you are ninjas here?
No one?

What about programming rock stars? No one?

So yeah, I mean, people kind of hate that

and if you are a company like this that is
developer friendly,

it will show through.

You won't have to put an effort to look like
that.

And, so it really does help if you hire better
people.

So what I'm saying is that,

we figured out that in the end it just falls
under two things.

Doing great work on the product,

on the ecosystem, and then talking about it.

I think especially in India,

where we are really good at doing great work,

I think we are very, incredibly talented set
of developers here.

But we are not so savvy about talking about
it,

and it's not as much of a skill thing

I think as much as it's a mindset thing.

So it's about doing great work and talking
about it.

But obviously, I mean, this is not easy.

You know building a company like this

is not going to be easy,

because it's all about good teams.

It's all about hiring the best people

that you can and after you hire them,

spending a lot of time helping them get even
better.

And empowering them, so they can make decisions
on your behalf.

And obviously this is a,

this is a very hard thing to achieve,

and it's very easy to get it wrong.

And you have to just keep trying.

And it takes a bit longer.

So if you are definitely trying to flip a
start up soon,

don't try this. Don't do this. It'll take
awhile,

but I think you'll get a much nicer company
in the end.

And one of the side effects of working like
this

or creating a company like this is that

you can actually work from the beach,

because now everything happens on the, online.

Everything happens in tools.

Everything is written, everything is documented.

And you become really comfortable.

It's like testing, right.

Before we started testing,

how many of you felt testing is gonna slow
you down?

And how many of you now feel that,

oh I can't write something serious without
tests.

Yeah, well same number of people.

Yeah. Good. Yeah, so it's the same thing.

It feels like writing is gonna slow your product
development down.

It's gonna take, make decisions longer,

but it actually helps you get way more structure.

And you'll definitely see that in a few weeks
or maybe a few months.

And - by the way that is a real picture

of us sitting and working from the beach,

so. That's one not-stock image.

And what's incredible to me personally

is that I was working on Musiboo for about
three years,

about two of those were miserable.

Right, there was, I was just banging my head
against a wall,

and I could have figured this out

and I could have changed my direction,

but I didn't. And I literally felt like I
was stuck

in a Groundhog Day.

Right, it's very easy to get stuck in a Groundhog
day.

And what you have to do is obviously

you have to step back,

and then you have to get a better perspective.

And I think the thing that prevented me from
doing that

was that I had no semblance of a work-life
balance.

Nothing. I was just working all the time.

I was under this assumption that if I am failing

it's only because I'm not working hard enough.

How many of you believe that actually?

That if you are not succeeding at something

then you just got to try harder? Yeah.

Quite a few people, right.

And I think this is kind of true

but I think it's not about the number of hours.

So definitely having some work-life balance

is very important in getting this perspective.

And the other thing I give thought a lot about

as a company is that there is simply no way
we can outwork our competitors.

So for example our competitors are companies
like FreshDesk,

which have thirty million dollars, two hundred
people team.

Even if like all of us work two days in a
day,

we would still not out-hour them.

So it's simply not going to work like that

and technology companies are not built like
that.

They're built one team member

and one great decision taken at a time.

And that happens when you have a fresh perspective

and a clear vision.

And it's amazing to me that being in the knowledge
industry

or industry driven so much by knowledge and
by creative energy,

how little at least I did, and when I talked
to other people,

people understand about how our brain itself
works.

And that's I think probably where a lot

of these misconceptions come from.

So I highly recognize that you guys

take a look at this book called Your Brain
at Work.

It's a very quick read. It's gonna take you
a few hours to read.

How many of you have read this book? OK, only
two people. Cool.

So. And both of them have (00:17:26:08) ??, so
yeah.

So yeah, please go ahead and read this book.

What this book shows you is how to get more
out of your work life.

How sometimes just trying harder is not the
solution.

And it's a very enjoyable read

and it gives you a very nice perspective.

Apart from this, I've also lately started
using RescueTime.

So I did a quick survey before the RubyConf

and I asked people how much they worked,

and so about twenty people I asked,

actually about twenty people filled out that
survey,

and I think seventy percent of them said

that they work about sixty hours a week.

And I encourage everyone to go back

and set up this application.

It's free for about a week.

And you'll be surprised by the results.

You'll be surprised by how much time you just
lose

in a YouTube video that somebody sent you

and you had no idea that you opened

and watched for thirty minutes.

So your time is definitely finite.

So there's no denying that.

But it's amazing that with that finite time

how much of it we lose without even knowing.

So try to use this application

and free up more of your time

and enjoy your life as well while you're building
a start up.

I think that's really the key.

So Bangalore. When Prakash invited me to do
the keynote,

he pushed out block which it said

I'm gonna bring a unique Indian perspective.

And I wasn't sure what that was gonna be entirely,

but one thing I do feel very passionate about

is being in this city actually.

So in the last couple of years I've traveled
a lot.

I've been to Chile,

I've spent a lot of time in Vietnam

and some other southeast Asian countries.

And I've traveled a lot

and I really enjoy living in some of these
places.

But I always come back to Bangalore,

and I really like spending time here simply
for one reason.

I think this place is full of incredible developers.

It's just so amazing.

I haven't been to U.S. Yeah, thanks.

Thank you.

I haven't been to U.S.

But I've been to some of the other hubs.

I've been to Syngapore.

But seriously the development activity here
is just amazing.

People are so passionate, they really like
the code,

they like trying out new stuff.

And in general if you see our English ability,

everything is really good.

And I really think that we,

and especially today if you look around you,

there's enough money, there's enough resources,

enough accelerators,

some may even say too many accelerators

or incubators around us, to help you launch
companies.

And I really hope that in 2014 a lot more

such companies get launched from Bangalore.

I'm totally rooting for you guys.

If I can help you guys in any way,

or anybody else, I'm totally up for it.

Let's build more such amazing companies.

So yeah. Happy New Year, and thank you so
much.

So I wanted to keep the talk a bit short,

so you can either use the time for Q&A

or reclaim your fifteen minutes of life

and go back home early. It's your choice.

V.O.: Questions, anybody.

QUESTION: Regarding the writing of,

I mean encouraging your team to start writing.

When you started up, how big was your team?

P.D.: Sorry?

QUESTION: When you started off with Muziboo,

how big was your team when you

started actually writing in the code and solving
issues?

P.D.: No, no, not really.

So none of this happened in Muziboo actually.

And that's why I kind of moved away from Muziboo.

I realized I'd gone too long, far too long,

to like, totally change

and the product had taken like a lot of fundamental
decisions.

QUESTION: OK, one more connecting question.

So when was the point that,

how many team members were there

when you actually started encouraging your

team members to start doing this?

P.D.: Right. So that's a cool question actually,

and you have to start doing this

when you are just one person actually.

And you have to start from day one.

So it's funny but even when I started doing
SupportBee,

I would actually, you know I was the sole
developer

but I used pivotal tracker

and I would try to document everything,

and not as much but still.

Because I think that's the biggest challenge

in setting any kind of culture,

that you have to start really early on.

And when you say early,

you have to basically start with yourself.

So the best time to start is

when you're starting the company.

And some companies have done a good job of
explaining this,

so recently you know we've been reading the
walf??

(00:21:54:23) handbook.

I don't know how many of you guys know about
it.

But there's a company called walf,

which makes a lot of cool games like Counterstrike
I believe, right.

And, sorry, I'm not a gamer, so, yeah.

And, so Walf has a good handbook,

and that's what they said.

To build a company like this,

you need a very strong commitment from day
one.

And one of the things I didn't mention

is that one reason we can do this

in a lot of companies can do this

and take longer to build better companies

is the fact that they don't take any investment.

So you know you might have to worry about
that as well, sometimes.

QUESTION: Hey, um, Prateek.

Something that always fascinates me

with entrepreneurs who have started product-

moved to a different product is,

at what time did you decide that, OK,

this product sucks.

Or it's not gonna work,

I need to move on.

When do you take that decision?

It could be a lot of factors -

finance, team, motivation.

P.D.: Right, right.

QUESTION: What was your move?

P.D.: Right. So I definitely had a moment

where I clearly knew I had to move on.

And so for awhile obviously I thought

that you know I should keep trying,

and I talked to a few more entrepreneurs,

and given. So we had a lot of traction,

we had about a million visits at one point.

And we did have a good amount of traction,

but the thing that frustrated me the most

was that over the three years that I did Muziboo,

I learned a lot of new stuff. And not just
in terms of programming,

in terms of how to go about product design
and things like that.

And in the case of Muziboo it had already
become

too big to sort of start over or to completely
change,

especially with, and we didn't have a big
team as well,

so it was just me and my co-founder

and I was pretty much the only programmer
anyway.

So it just felt like, with the new experiences

that we had and the new knowledge that we
had,

it just makes more sense to start something
fresh.

Because one of the things that also happens
is

when you are just starting out, sometimes
you

pick an idea that's close to your heart, or
looks cool,

but may not necessarily make business sense.

So with all those factors put in it

just felt that it's better to start over again

in a domain that I can relate to more.

One of the other problems in Muziboo

was I'm not a musician myself.

So after awhile I couldn't relate to their
problems actually.

And it's, in Support it's been incredible

because we use SupportBee ourselves.

And I wanted to get into a domain where

at least I could relate to the target customers
a lot more.

So even though I wasn't from customer support,

I could relate to business owners and small

and medium companies,

and so it made sense to just switch over.

QUESTION: OK, so your story is very close
to,

you might like this, ??, he's a
??.

P.D.: OK.

QUESTION: But he started with his own service
company.

No, he- yeah. First service company,

and then he expanded his product (00:24:27:19)

and then he learned this concept of ??, so.
??

P.D.: Right. Yeah I think a lot of entrepreneurs
do that.

I think visual website optimizer are another
one.

Paris ?? tried a couple of ideas.

In fact he had a music service as well.

I don't remember the name but it was there
for awhile.

And, so I think even India there are a lot
of entrepreneurs

who have done that actually.

And I think they've had pretty good success
that way.

QUESTION: Do you think it's important

for someone who wants to start a start up

to first work in a desk job, you know,

with an MNC and then start the start up?

I mean do you need the experience or can you
just,

you know, finish college and then go straight
into a start up?

This is because I'm a college grad,

almost, so I want to know.

P.D.: Right, so obviously this answer

will come with a disclaimer that, yeah,

you know, there are always exceptions to the
rule,

and so there are, like for example,

I think Fractor?? is one company

that came right out of college and did really
well in India.

I don't know if you know about them.

But typically what I've seen is,

it does help to have a few years of experience.

But at least today you don't need to get that
experience in an MNC,

necessarily.

You can actually get it in a well-run start
up.

A start up with a certain amount of validation.

So probably you don't want to join a start
up

that has no validation,

and you're not sure where the start up is
headed.

But as soon as you think there's enough validation

or you like the product story or you like
the founders,

you can definitely work with start ups.

And they can teach you a lot more.

Similarly because in starts ups you work,

you have to do what work,

and your work is always going to be used.

Big companies have the luxury of just sitting

on your work or just wasting your time.

But in a start up that luxury isn't there,

thankfully. So you're always going to gain
a lot more exposure.

So yes, to answer your question,

I think it's good to get some experience.

QUESTION: So hi Prateek,

I'm going to add a comment,

and then you can probably add more to it.

To start up, there's no right time.

But what I believe is the research that goes
into starting up.

So whether you're out of college or you work
for MNC,

whatever that means, but all your work for
any company,

as long as you do good work and you find a
problem you can solve,

and you validate it, by going, asking, talking
to people.

Then you can start up.

And then the rest of the start up stuff

like finances and team and stuff comes through.

So Prateek it'd be pretty interesting

to see how you started about the idea about
SupportBee.

Why you thought there was a gap.

When did you realize that this could actually
potentially take off?

And finally, you know, competition.

P.D.: Right. Can these lights be a little
bit dimmer?

It's, if it's OK.

So sorry, so your question was when did I,

when did I start and why did I think it was
a good idea?

QUESTION: So you, there must have been a time

when you found there's a pain point,

and was like hey, man, this is a cool idea,

then you did your research.

Then you actually said,

all right I think there is some app to be
found.

And finally competition.

P.D.: Right, right.

So when I was coming out of Muziboo-

actually in some ways that's the other lesson

I have learned. It's always easy to learn
in hindsight,

right. You can look back and it's much easier
to come up with lessons,

life lessons. So take it with a pinch of salt.

But one thing I realized was that after Muziboo,

I was pretty, like, not really desperate,

but I was like, in a rush. I was like,

oh my god man, you know, I've been sitting

for a whole twenty days without doing anything.

You know, three weeks. And, so I was like,

actively looking for ideas, and we've had,

we had this idea about having a customer support
tool before.

Because we tried to find form to use

where we ended up using something,

but it wasn't that great.

So it was around that time that I came up
with the idea,

all right. And so, one of the first things
I did was,

and we had a few other ideas before that,

so we tried validating those, and those didn't
work with customers.

So we would just put about,

put up a landing page and then try talking
to people.

And the one of the first things I did about,

when I came up with the SupportBee idea was,

I talked to Valerie from Balsamiq.

So Balsamiq is a very popular company

I think most of you guys would know.

And they were very, very well-known for their
customer support.

So I reached out to Valerie

and she was kind enough to come on a Skype
call

and talk to me about the idea,

and we did this with a few other companies
like BooFoo ??

and Discuss, and actually a lot of those interviews
are now online,

on our blog. And we have also written more
about this process,

it's still on our blog if you guys are interested
in checking it out.

And so with, in a month,

a month and a half even then we thought,

OK, we could do this. I think what Coby said
in the morning that,

you basically found it at least as hard

as you thought it was going to be,

in my case it was definitely true.

So I thought OK, so I just need to

pass an email and then I can get started.

And that's it. Passing the email too me a
night,

and then ?? (00:29:12:21) a little bit later,

it took us a whole year or something, pretty
much.

And so we started coding,

but getting the product out it took us

about a good eight months or something.

And in private ... ?? (00:29:27)

V.O.: Anymore questions? All right,

thank you Prateek.

P.D.: All right, thank you so much.