How does a dropout go on to build two hit startups?
Tikhon Bernstam created Scribd, the document sharing site that has over 100 million unique visitors per month. His most recent startup is Parse, which provides cloud services for mobile developers. It makes it dead simple to add a backend to your iOS or Android app.
In this program you’ll finally see how charging users can turn into a viable business model and how headline writing could make simple documents super-viral.
Tikhon Bernstam, Parse
Tikhon Bernstam is the cofounder at Parse, which provides cloud services for mobile developers. It makes it dead simple to add a backend to your iOS or Android app.
Andrew: In this program you’re going to see, finally, how charging users
can turn into a viable business model, maybe even for you. You’re going to
learn how headline writing can turn a simple document super viral and how
it can send the company behind it into the stratosphere of internet
success. And you’ll discover how today’s guest got hundreds of partners
within weeks of launching his latest start-up. All that and so much more.
I’ve got a great guest here, stay tuned.
First three messages: Who’s the lawyer that founders in the Mixergy
audience trust? Scott Edward Walker of Walker Corporate Law. Have you seen
what Chris Pritchard [SP] posted on my Facebook page? His new company’s
incorporations pages, that Scott Edward Walker helped him get. Scott Edward
Walker is the lawyer that publications like Forbes trust. Go to
Next, when my friend had to close her company’s office but still wanted to
give callers the impression that all her employees work well under one roof
together, what service did she use? Grasshopper. With Grasshopper everyone
who works for you can have an extension. They can pick up calls on their
extensions no matter where they are or what phones they use, and they can
transfer calls to each other, back and forth with ease. Get those features
and tons more at Grasshopper.com.
Finally, when Dave Jackson and Dave Petrillo invented a product that keeps
coffee at the perfect temperature, what platform did they use to create
their online store? Shopify.com. Look at how beautiful their store looks.
It’s because it’s built on Shopify. They did hundreds of thousands of
dollars in sales. Shopify stores are designed to help you sell. Patrick
Buckley invented an iPad case and used Shopify as his online store. Within
months, he sold over a million dollars in cases. Get your beautiful online
store at Shopify.com. Here’s your program.
Hey everyone, my name is Andrew Warner, I’m the founder of Mixergy.com. How
does a dropout go on to build two hit start-ups? Tikhon Bernstam created
Scribd, the document sharing site that has over a 100 million unique
visitors every month. His most recent startup is Parse, which provides
cloud services for mobile developers. It makes it dead simple to add a
backend to you IOS or Android app. Tikhon, welcome.
Tikhon: Thank you, Andrew. It’s a pleasure to be here.
Andrew: Thanks. You had an interesting experience where you went to the New
York Times and you saw something that every entrepreneur dreams of seeing.
Can you tell people about that?
Tikhon: Yeah. I think that was really the first time that I realized that
Scribd had really made it, and had really made a dent in the internet
universe. I was reading a story about Goldman Sachs and I noticed that they
had actually embedded the legal filings about Goldman Sachs at the end of
the article using Scribd. It was a really cool moment for me and I
immediately emailed it to my mom, my dad, and probably everyone else that I
knew. Since then, we have embeds now all over the internet, including the
New York Times, all sorts of folks, there’s one in TechCrunch today, and
it’s really rewarding to see all that work actually being used by people to
make the sharing of documents easier and faster.
Andrew: And there’s so many other ways that the New York Times could have
gone to share a document. They could have frankly just put it in a PDF and
embedded the PDF using one of Google’s iFrame systems, they could have used
a competitor of yours. I want to know why they specifically used you and
why 100 million other people every month use you. I’ll push you a little
bit about your competition to see why they don’t use the competition. I’m
saving that for this interview, I want to go in time and go step by step to
see how you got here. But let’s do one other thing and tell people about
Parse. You had a similar experience with Parse with Hipmunk. Can you tell
people about that?
Tikhon: Yeah, I’m a really loyal user of Hipmunk, I use it to book all of
my flights, it’s a really great company. I was using their IOS app when I
actually discovered that Hipmunk was actually using Parse to support it.
The same thing happened with Band of The Day, which is a great app for
discovering indie musicians, it was runner up for app of the year in the
Apple App Store, and I was user of that app for, actually, a long time
before I even realized that it was an app on Parse and it’s built entirely
on Parse. All the data is actually living on Parse.
Andrew: How long ago did you launch it?
Tikhon: Parse? Our official out-of-Beta launch was in early April, but our
Beta was launched in August.
Andrew: In 1500 apps, hundreds of partners…
Tikhon: We have over 15,000 apps on the platform.
Andrew: 15,000? I got this wrong. 15,000 apps built on the platform and
it’s just weeks old.
Tikhon: Yes. I was (?) by 10 so I thought I’d surprise you.
Andrew: You know what? I wrote it down, and, wow, impressive.
Tikhon: Off by a factor of ten. There are over 15,000 apps on the platform.
We’re powering apps that are installed on tens of millions of mobile
tablets and phones, and growing 40% month by month.
Andrew: I know that there are people in the audience who are thinking to
themselves, ‘You know what? I’ would like to get a hundred partners. This
guy’s got 15,000 people and the company’s fairly young. He’s got millions
of people with the other business.’ Do me this one favor. There’s a guy who
I know who has listened to past Mixergy interviewees. Don’t hold back.
Don’t be modest. Don’t be the guy who goes, ‘Aw, shucks. Things just worked
out. I partnered with these guys. Everything then worked out.’ Really dig
in and be useful based on your story. Fair?
Tikhon: It was 100% all me.
Andrew: There you go. Now take it another step back and go, ‘Here’s the one
thing that I did that really helped me get a customer.’ Just give me
specifics as we go through. Let’s go back in time and see how you did this.
Tikhon: In the case of Scribd, we launched on TechCrunch. We weren’t even
sure if it was going to work or not, or if people wanted the thing or not.
We launched on TechCrunch and the story got on Digg, which was a really big
deal, because back then there weren’t a lot of TechCrunch stories on Digg.
This was back when Digg was huge, and it was what everyone read. It was
incredibly influential. After that, Scribd just grew like crazy. In our
first month after launching, we were one of the top ten most Digg sites
that month, I believe. That really helped our growth because all those
inbound links we got from Digg and all the other places that read the
article on Digg, that would read stories about Digg and then actually link
to us, that ended up driving a ton of search traffic.
To this day, Scribd still does a ton of search traffic. Even more
interesting is actually what I didn’t expect, which was that imbeds would
be so popular. Now we get a ton of traffic from all kinds of blogs and
newspapers like the New York Times who are imbedding everything from legal
filings to manuals about how to fix your toaster from the 1930s.
Andrew: I can imagine someone sitting in the background or sitting in the
audience thinking, ‘Yes, that is brilliant. Of course! Create something
that people want to imbed on their site, then of course they’re going to
give you traffic that they have on their site and it starts to snowball.’
But let’s go back in time and see where this idea came from. Where did the
idea for Scribd come from?
Tikhon: It came from the difficulty of sharing documents with our parents,
actually. Trip was trying to share a document with his dad, Dr. (?), who
runs Neurosurgery at Stanford, and I tried to give my mom a postscript
file. It took me about an hour to (?) file, which she had never done
before. [Audio file cuts in and out] And also Dr. Adler, Trip’s dad, he
really wanted to publish his articles and papers online easily without
waiting for the journal process, which was taking months and months to get
anything done because they move at a glacial pace. So, he just wanted a
nice way for people to view his documents inside the browser. It was a
combination of us talking about those experiences that was the genesis of
the idea. We did that in Y Combinator in the summer of 2006 and launched it
Andrew: Did you get into Y Combinator with the idea?
Tikhon: No, I was in Y Combinator for a different idea and so were Trip and
Jared, and then we ended up teaming up and building Scribd.
Andrew: What was the idea that you had?
Tikhon: The original idea was a job site for kids coming out of college. I
always felt confused leaving school. Every job I found said they wanted two
to four years of experience with XYZ, and so I didn’t feel that there was a
good, dedicated site for students leaving school. I still don’t feel like
there is, actually, for finding jobs. All my friends just did corporate
recruiting because those were the jobs they thought were available. A lot
of folks who couldn’t get jobs in corporate recruiting felt like they were
lost, like their lives were over and they would never get a job. While
trying to explain to them that there were other jobs out there, I kind of
had the idea for that site. But I ended up abandoning that to work on
Scribd which I felt was more promising.
Andrew: Why? If this is an idea that you still think is needed today, why
did you give it up back then? Especially if you got into Y Combinator and
they funded you based on that idea. Why’d you give it up?
Tikhon: I was just really enthralled, Scribd to me was just a compelling
thing that I couldn’t not work on. Having had that problem my entire life,
sharing documents had been so difficult, it was something I instantly
Andrew: OK. What about the business model behind it? Did you think through
and think, hey, if we can get people to share documents then we can add an
ad to it? Tell me what you thought about the business model.
Tikhon: The honest answer is that we had no idea. We were going with the
polygram school of thought of make something people want and then we could
figure out the business model later. We actually were very bearish on ads,
we didn’t think ads would be a good model, but we were totally wrong about
that. We had heard a lot of stories about various websites having 5 cent,
or 10 cent CPMs and we ran the numbers and said, that’s kind of boring.
That’s not really an interesting model, maybe we’ll sell document
management systems to enterprises or something else down the road.
Actually, when we finally turned on ads, much too late, we discovered that
they were incredibly effective and the company actually got profitable just
off the ads, which I never would have guessed going in. I also never would
have guessed how important search would have been to the site, which in
hindsight seems dumb, but we had all of these inbound links, and we had
something on the order of 10 to 100 times the number of words that
Wikipedia has now, so with all that text and all those inbound links, we
get a ton of search traffic from sites like, Google, Yahoo, Bing, etc., and
that’s been really important to the website’s growth as well. There’s this
sort of positive feedback loop where some percentage of people upload a
document, that document gets indexed, people link to it, and then that
drives even more traffic.
So, it created this positive SEO viral loop, so that for us was a great
mechanism for growth and distribution. It’s always cool still, I remember
the first time I was just doing a random Google search, actually about [??]
physics and I ran into my own E&M homework solutions that I posted on
Scribd randomly as I was testing out our conversion system. I thought no
one would ever read these graduate level Jackson homework solutions.
They’re not interesting to anyone, but now, looking back on them for fun,
every year around when that intro E&M class is taught in graduate schools
I’d see a huge spike in my Scribd analytics, where all of these students
are looking for help with their homework.
Tikhon: So, it’s kind of rewarding. I remember how Trip and Jared, my co-
founders at Scribd, who are actually the other half of the reason I wanted
to work on Scribd, those two guys, they’re both brilliant, really great
guys. They mocked me for putting up all of my old homework solutions, they
said, who’s going to want to read that? Now they drive a really interesting
amount of profit, so, I still mock them about that.
Andrew: There are a lot of people who are brilliant in Silicon Valley,
especially at Y Combinator and surrounding Y Combinator, what was it about
these guys that made you say, not only are they brilliant, but I’m ready to
marry them, I’m ready to be in this long-term relationship with these two?
Tikhon: Yeah. So, I actually went to the same high school as Trip. [??]
Andrew: I went to the same high school with thousands of people, I wouldn’t
want to be in business with them. What was it about him specifically that
made you say, hey, this guy’s got what I need.
Tikhon: They moved really fast and got a lot of work done really quickly.
They were the fastest executing team.
Tikhon: Or one of the best executing teams I’ve ever seen. Still to this
day, they are. They got so much done so fast that looked so good, it was
really a no brainer. Jared is one of the smartest people I’ve ever met and
one of the best hackers also.
Andrew: Do you have an example of something they built really quickly back
in the Y Combinator incubation days?
Tikhon: Yes. So, they were actually admitted for a different startup in the
Y Combinator. They were building Craigslist for colleges so this was back
when Facebook was using the whole for colleges angle, X for colleges was
like a popular idea. It was actually a great site, it might still be up
somewhere. It was called Hulist, Harvard University List because it was
starting at Harvard. It was a great site and they built it really quickly,
just in a couple weeks, and Jared did a fantastic job on the technical side
and it was really a no brainer to work with those guys.
Andrew: So, what about you then? I can see you’re saying in a short period
of time they took one idea, scrapped it and built another idea, or set the
first idea aside…
Tikhon: They actually built like twelve ideas, they had built like twelve
Andrew: All in this three month period?
Tikhon: Yeah, this is how good they are. So, it was a no brainer. I
actually toyed around with a few ideas during Y Combinator. That’s
something a lot of folks don’t appreciate, how many ideas actually change
once folks are already in Y.C. Even in my last batch for Parse there were a
lot of companies that changed ideas midway through.
Andrew: How much of the idea do you develop if you’re building twelve of
Tikhon: So, some of them are landing pages, some of them they tested out
amongst friends, some of them they launched on various forums. They didn’t
all get TechCrunch coverage which at the time was the best way to launch a
consumer facing start-up but it was a lot of testing things out and just
kind of seeing what would garner interest from our friends and families and
random people on the internet. Scribd was one of those things that was
always growing. Even before TechCrunch when we had just a basic prototype
going, we noticed it growing and the people who came stayed engaged. A lot
of folks wanted a medium for distributing document-like content regularly.
Actually, I remember our very first big user was a guy named Bill Allen. We
were so excited because he got put on the Adam Carolla radio show. They
invited him on because of a document they found on Scribd.
Tikhon: That was actually the very first press hit Scibd ever got. The
first time Scribd was ever mentioned in the news was due to that.
Andrew: And this was before the official launch?
Tikhon: I believe it was right before the official launch, yeah.
Andrew: Wow. All right, so what about you? I understand why you’d want to
partner with them. This is where you have to let go of the modesty and be
open, what was it about you do you think that made them say, yeah, we want
to get this guy on board.
Tikhon: Yeah, I mean, I had built a lot of stuff and I had a lot to show
for it. I think they liked that I was a physics drop-out. I dropped out of
physics graduate school where I was studying string theory to join Y
Combinator. But also I studied computer science in undergrad and for what
it’s worth, I was early 5A, top 2 percent, and all that stuff.
Andrew: Do you have an example of what you built that might have excited
Tikhon: The stuff I had built inside of Y.C. during those few weeks with my
friend Eric, there was a lot of impressive technology there, we just
decided not to keep going with it But we had built some really, really
interesting technology there and I think they were impressed with that, how
fast we were also working as well.
Andrew: What happened with Eric? Did I just lose you?
Tikhon: Eric actually never moved to [??], he stayed in North Carolina so
he was kind of part time. He had a full-time job and he was working for me
outside of that. He’s a great guy, he was top of his class in the Stanford
Masters program in computer science. Brilliant, brilliant guy and one of my
best friends out of college. He just didn’t really want to go down the
startup route. We were working Scribd seven days a week, 15 and 16 hour
days, I didn’t see anyone from family to friends for months. Not everyone
can afford to do that, or wants to do that. As Mark Pincus once said, you
kind of have to have a screw loose to do a start-up because it can be so
difficult and taxing. I think he’s probably right about that.
Andrew: So, Eric didn’t have a screw loose.
Tikhon: Eric is too sane of a person and well-balanced.
Andrew: You mentioned the basic prototype for Scribd; even that got a lot
of attention. What did that look like? What did it do?
Tikhon: All it was was a simple way to put up a document and have your
picture next to it. It was a document inside of a Flash player so you could
view the document inside your browser, much like you would watch a YouTube
video, you know, using Flash, but at the time, there really had been
nothing like that before for .PDF, Excel or postscript.
Andrew: Fully embedded, right?
Tikhon: Yes. It was inside of a webpage so you could, just like the New
York Times, now instead of attaching a link to a .PDF file, they could just
show you the .PDF inside of the story which, at the time, was actually
revolutionary. Now there several types of that do it, but (?) still the
largest for all these document formats.
Andrew: So, I want to understand what the basic elements were that went
into this first version just so I understand how you made a decision of
what to include and what not to. Imbedding was important to you.
Tikhon: Yes. It only did three things. It uploaded documents, it would
convert it into flash and you could imbed in elsewhere. There were a few
minor features around analytics, so you could see how many visitors you
had, almost a poor man’s Google Analytics for your documents. Really, that
was the core of the product. There was a top ten list for our most popular
content only because Digg and Reddit were enormous at the time, and so we
said we’ll throw in a top documents list for the day.
Andrew: Why was analytics such an important part of the vision that you
decided to launch with it even back when you were in the basic prototype
Andrew: Analytics. Right. Why is it so important? I wrote a note here
because you brought it up earlier and I wanted to follow up. It seems like
it was really important to the business.
Tikhon: That’s a good question. I actually didn’t want to include it, but
folks like Bill Allen really wanted analytics. They wanted to know: How
effective are my essays? Am I getting more readership every day or less
readership? It was almost, for him, an alternative to a blogging platform
where he could write up these essays once a week. Actually (?) was every
day. It was one essay a day. That was his policy, and I think he’s still
doing it to this day. He wanted to know was his readership growing or
decreasing, and he had no good metrics for that. People like him kept
asking if we had features like, and we begrudgingly said, ‘OK. We’ll add
analytics to the platform.’ So we spent a couple days doing that, and
luckily with folks like Jerrod and Trip, we could pump stuff like that out
in a matter of hours that would normally take years to get done.
Andrew: Was it also about gamification, encouraging people to come up with
better contents and more people do it. You’re nodding.
Tikhon: The honest answer is that we didn’t think about that that way, but
it worked out that way. There is a sort of positive reinforcement where
people see that, “Oh my gosh, people are actually reading my document! It’s
amazing! I have ten links inbound.” That absolutely helped. I can’t admit
that I knew it would at that time. I think we begrudgingly added analytics
in the first place, mostly to quiet our user list.
Andrew: What didn’t you put in there that you might have or that someone
else who was going to go way too big and spend way more time than you did
would have? What would they have included? What did you decide not to?
Tikhon: One example of things people really wanted that we didn’t include
for a long time was the ability to follow other users so you could have a
feed of, for example, Bill Allen’s content and other people’s content on
the site. Now the way Tumblr or Twitter works. We didn’t have that for a
long time just because we were trying to keep the product as lean as
possible and launch as quickly as possible. That’s part of the whole
polygram (?) philosophy. Launch early and if you’re not embarrassed by what
you’ve launched you’ve waited too long. We were embarrassed by what we
launched, but we launched just enough that the core was there so we could
Andrew: All right. This was 2006, right?
Tikhon: 2006, yes.
Andrew: All right. Then TechCrunch writes about you and the launch.
Tikhon: Tech Crunch, yes. TechCrunch wrote about us. That was scary that
night, because at the time we were sure if Tech Crunch was going to write
the story or not. So, Nick Gonzalez, who’s now a good friend of mine who’s
not at TechCrunch anymore, we’d sent him a link. It was a secret link and
it had an extra pram attached so we could know if he clicked on the link or
not. The story was suppose to be published at 9:00 a.m. on I think it was a
Tuesday, and it was 3:00 in the morning, six hours before, and no one had
clicked the link. We thought, “Oh my God, they’re not going to cover us.
This whole thing is going to go down in flames.” At 3:00 in the morning,
we’re sitting there, still up, waiting. What’s going on? Is TechCrunch
actually going to cover us? And then he clicks the link. We thought, “Oh my
God. Finally.” So, he put the story together between 3:00 a.m. and 9:00
a.m. It got out there at 9:00 a.m. and the story got on Digg, which at the
time TechCrunch posts were rarely on Digg.
I think we were one of the firsts and so TechCrunch was actually thrilled
with us because Digg had thrown a ton of traffic to them via that post. I
think they told us we were one biggest post or most viewed post that they
had at the time. Now they’re huge obviously but back then TechCrunch was a
little bit smaller and the story got on TechCrunch and it just took off.
Then all of a sudden we had other people linking to Scribd documents from
Digg and then those started getting up voted and so that first month we
were one of the top ten like I said one of the most dug sites. I think that
very first month we had so many documents that were on Digg because just
previously it was such a pain. It’s like how would you share some
interesting, or funny, or creative document with someone else? We’d have to
give them a link to the PDF and then we’d need a w-reader and then you’d
have to wait for 30 minutes while w-reader opens up on your PC to actually
read the damn thing.
Andrew: You know what? You’re right this was a big pain in the but, anyone
who’s listening to us now, was on the internet at the time knows that
clicking a link and going to a PDF doc was a big pain in the but. That
referring to one from your site was a pain in the but so here are all these
people who were internet savvy, that were facing the same problem, who
didn’t catch this opportunity that you caught, that we’ll see in a moment
just took off, why? What weren’t we doing that you guys were doing? I mean
if it’s so obvious than why do we miss it?
Tikhon: Yeah, back then and even still today whenever people do link to
PDFs, like Hacker News or Reddit, they’ll put PDF in the title to warn you.
Tikhon: Be careful, it’s a terrible link, you download this thing and then
open it, you need a client on your PC or a Mac or whatever to actually open
the damn thing and so it’s always been this really painful experience
unlike the rest of the web that was kind of so easy to use from the
browser. So we didn’t really think about that at the time but now, to this
day next to every PDF link on Hacker News there’s still a link to the
Scribd version because it’s so damn painful to view these documents.
Andrew: So how did you notice it? I mean, when I get emails from bigger
companies what they care about when they listen to my interviews isn’t so
much the funding, it’s not the period of sorrow that we all go through as
entrepreneurs, they can’t relate to it, those aren’t issues for them but
what they’re fascinated by when it comes to start-ups is, ideation. Where
did these great ideas come from? How did they know that this is a hot idea
that needs to built and the rest of the world is ignoring it. So as a guy
who’s caught that what can you add to that understanding?
Tikhon: Well the honest answer, we tried a lot of ideas before Scribd that
didn’t work. So it’s not that we had this one genius idea, we just tried a
lot of things. We tried solving a lot of problems that we ourselves had
like, sharing the post Scribd file with my mom or Tripp’s dad having this
problem putting up his articles on line. So, we were really solving our own
problems much like the jobs site I was working onY Combinator and that was
a problem that I had seen first hand.
Andrew: So this solving your problems, solving problem that other people
had but also just trying a lot of things and seeing what stuck.
Tikhon: Yeah, just moving really fast and trying a lot of ideas.
Andrew: And getting feedback quickly so you’d show this thing, who’d you
show Scribd, the early version to who maybe said this is good. You’re
smiling, you tell me.
Tikhon: A couple of my good friends in this dialect Justin Kahn and Steve
Huffman looked at scribe and said no don’t do this. That’s not a good idea,
Scribd we actually preserved on and now they kind of laugh about it looking
back on it.
Tikhon: They were folks like Bill Allen who were so passionate about the
product that we knew we hit something , we realized we had hit on
something. They found us via links from some forum and that was just a
really rough prototype before we had publicly launched. Once you can make a
few, it’s like Paul Buffet said, if you can make a few people really,
really happy, that’s much better than making a lot of people like kind of
Happy. So when we have this like kind of rabid band of loyal users. We knew
it was OK that not everyone loved it or needed it. ‘Cuz we hit a real niche
that is clearly a huge pain point for a lot of people.
Andrew: So, also then the feedback process that I expect you need to go
through as a person who’s coming out with a new idea is, create it and have
a formal conversation with someone and say do you want it? You’re nodding.
I was going to say that’s not how you did it but it is right?
Tikhon: We talked all the early Scribd users and said how the hell did you
find this thing?
Andrew: So, first you got them to find it and then you sat down and asked
them how they found it and they what else did you ask them to make the
conversation real useful?
Tikhon: Why do you use it? What are you doing this? Why Scribd and not
something else? Why not link to the PDF file from your blog? Why are you
using embeds? We were constantly looking for feedback which is how we
settled on Scribd in the first place and rejected other ideas that we had
been working on.
Andrew: OK. All right, let’s talk about how you grew from there. So, now
you know you’re onto something, it’s starting to get some traction, what do
you do to push it along so that it goes way bigger?
Tikhon: Yeah. So, I didn’t initially anticipate this, but like I said,
there was this positive feedback loop where Scribd was really growing on
its on for a long time, even if we hadn’t been helping it. We were
constantly adding features to help that people kept clamoring for, like
better analytics, eventually HTML 5 embeds and HTML 5 support and stuff
like that instead of flash, but what was great about it was that it had
this interesting new viral loop where a small percentage of the users would
upload a document and then link to it from around the web saying, check out
my document, and then that would get us more page rank from, you know,
Google and Yahoo. That would drive more visitors who uploaded their own
We also threw contests, who has the most interesting hard drive was one
contest we threw and we gave away a Mac Book Pro to the person who uploaded
the most interesting documents based on the number of views they got. So,
there were a lot of gimmicks like that. We really encouraged our users to
put content on social sharing sites, at the time these were like Digg and
Reddit, and now it’s Facebook and Twitter. We always encouraged our users,
like, hey, are you sure you don’t want to put this up on Digg and Reddit
and try to promote your articles?
Andrew: You mean, within the flow of uploading?
Tikhon: Yeah. I still remember to this day how excited Bill Allen was,
because he had an essay about why intelligent people are unhappy, which
apparently is not true, there’s no correlation between intelligence and
happiness, according to Wikipedia, but he had a whole essay on this topic
and it got on the front page of Dig, and he was blown away by the analytics
showing. I think he had like millions of people who had seen this article
of his, and he was just blown away by this, that he could find this kind of
distribution via Scribd and social media.
Andrew: So, it seems like your marketing process was very similar to the
process of coming out with that idea. You were just trying a bunch of
things and seeing what worked with the audience and what got you traction.
So, where did you get those ideas that you’re going to try? Because this is
really clever, I mean, some of it is really simple but it’s a lot more
clever than it seems.
Tikhon: Yeah. For all the ideas that I’ve mentioned there’s probably a ton
that we tried that also didn’t work at all. We had one project where we
encouraged people to actually physically mail us books and documents and we
would put them up on Scribd for them. That ended up being a terrible idea.
I’m telling you these success stories and not stories like that, ideas that
didn’t really work as well.
Andrew: So, you told me that one of the reasons why you were able to try
lots of different ideas at Y Combinator was that you kept them really
simple. You said all of the things that you left out that other people
might have included in them. What about when you’re trying marketing ideas?
What are you doing to keep them so simple that you’re not going out on a
limb and spending months on a project that ends up tanking, being like this
Tikhon: Yeah. The truth is I think a lot of folks including us did do that.
There are projects that we spent too much time on that we actually didn’t
end up pursuing. But just being able to execute really quickly and try a
lot of ideas was for us the key to having anything work.
Andrew: Can you give me an example of that? Of something that you guys just
had an idea, and then you executed the marketing on it quickly and it
either bombed or succeeded.
Tikhon: Yeah. The thing I was just talking about, mail us your documents,
that was not a good idea.
Andrew: How long did it take you to put it together?
Tikhon: Like 24 hours. The contest about who had the most interesting hard
drive, we did that in 24 hours too.
Andrew: So, how do you do a book… here’s what I would think. If you and I
were sitting together and you said, Andrew, we both run Scribd together for
this hypothetical example, I think what we should do is we should have
people mail us magazines and books and we’ll upload it for them. My
thinking would be, we should find a place where someone can scan it in for
us, maybe we find a company that we can partner with, maybe we hire an
intern and then we need something for him to cut the binding off, and then
do we have enough scanners? Maybe we should limit this, do we talk to a
lawyer about making sure that we don’t accept too many books, or let people
know that we’re not going to … look at all these things that I’m thinking
of. What did you limit yourself to so that you can do this within 24 hours?
Tikhon: Yeah, that particular idea came from reading about how Google was
scanning in books. And we were, like, “Wow, that’s amazing.” And it turned
out it was actually very cheap per page, much cheaper than we thought. And
so we found a way to do scanning, it was very cheap per page. And so we
just launched a story about it and kind of decided to wait and see if
people would mail us stuff. Some people did but it just wasn’t enough and
the project kind of … it was a nice little PR stunt, but it wasn’t
Andrew: Because you didn’t get enough books sent to you?
Tikhon: Yeah. Just not enough people who actually wanted to physically
package up and mail books and documents to us, surprisingly.
Andrew: But did you have somebody in house who was going to do this for
you? Did you spend any time on that?
Tikhon: Oh, we actually ended up scanning all that and putting it online,
it just wasn’t enough.
Andrew: Okay. So you just found a quick solution to do it. You realize it’s
not that expensive. So someone else might say, look at this, these Google
guys are sticking with this. And we’re not getting enough documents, but
maybe it’s because we’re not promoting it enough. What else can we do to
promote it? Why didn’t you guys go in that direction and spend even more
time and more energy on this one thing that didn’t work right away?
Tikhon: Yeah, it ended up not being a profitable adventure, also, is why we
didn’t put as much time into it. We discovered actually after executing it
that it was more expensive than what we initially thought it would be to
[??] these documents online. And the revenue that we would see from those,
didn’t justify it. We later, after putting ads on the site, we actually
added premium accounts to the website, which are now very, very successful
as well. The ads actually brought us the profitability. But the premium
accounts are now a huge source of revenue for Scribd.
Andrew: Okay. Here’s the other thing that I’m getting. You just have a
mindset of putting simple things out there and seeing how it works. But
there’s no sense that, well, we just had this big hit launch, we don’t want
to let people down by coming up with a half-baked marketing idea. You just
run it out there and if works great, if it doesn’t, you kill it. Right?
Tikhon: Yeah. I mean, for Scribd, at least in the first 3 to 6 months, it
was all about getting document in prominent places. Like on Digg, other
similar sites, and just getting links to Scribd and getting people viewing
documents. But a lot of these PR gimmicks and stunts were like a year, two,
or three in.
Andrew: Okay. So, you need to get people to embed your documents. I’ve got
friend who runs, Docstoc, Jason Nazar.
Tikhon: Jason’s a great guy. We’re actually good friends with Jason. He
called us back when he first discovered Scribd, before we launched it. We
just had a basic prototype up and we talked to him, and we’ve been actually
friends ever since.
Andrew: But he launched his company after he had this conversation with
Andrew: Did you feel like Jason screwed you over?
Tikhon: No, not at all. Docstoc is taking a different angle. They’re really
focused on legal and professional documents.
Andrew: Because they couldn’t go in your direction, right?
Tikhon: Not necessarily. Jason also wanted to do legal. He’s always wanted
to focus on that vertical. Just like SlideShare was really focused on Power
Points and presentations at that time, we had our own focus that was kind
of the general PDFs and poster files of the world, Word, Excel, all these
other formats that weren’t supported. But all this content that was just
difficult to share and making it easier to share and embed document . . .
Andrew: You wanted to go broader and he wanted to go narrower from the
Tikhon: Yes. He was always focused on that vertical that they’re still
doing right now. And they’re doing very well. They’re a great company.
Andrew: They are, right?
Tikhon: Yeah. Jason’s a great guy and a great CEO.
Andrew: But he’s really good at BizDev. Like, you can see him work a dinner
and get people, the next day, to embed their documents through his system,
on their sites? Who is in your company doing that BizDeveral [SP]? And what
did you guys learn from doing it?
Tikhon: Yes. I mean, we ended up hiring some really great people to help us
with that. But in the beginning, we had to learn it all ourselves. We wore
every hat. I mean, one day I’d be programming all night and the next day
I’d be trying to get people to embed documents in their blogs.
Andrew: Can you give me an example of how you did that? I remember, Matt
Mullenweg, founder of WordPress told me, when he wanted to get people to
use WordPress, he’d just go into the comments on their other blogging
platform and say, “Hey, did you know that I’ve got WordPress?” And he got
them to go.
Tikhon: Yeah. We found folks who were linking to Documents. Like a lot of
bloggers, especially financial bloggers, were including links to PDFs. And
you have to download that and try to look at that.
Andrew: So what would you do, when you saw that?
Tikhon: There’s a much easier experience than this. We’d go out and look
for those people and say, you can just embed that directly into your page.
You don’t have to that link to some enormous PDF. And people jumped on that
actually, because they really wanted their readers to find these documents
while not being taken off of their blogs, for instance, and not to lose
that engagement. So Scribd was a real instant win for them because, A, we
hosted file for them so they didn’t have to host it and bother with that.
B, they didn’t lose the reader because once someone clicks on the PDF link,
for instance, they’ve left your site and they’re gone and they might not
come back. But with Scribd, they’re still on your site. So selling embeds
to people was actually a really easy sell. And much like I guess Matt’s
strategy, yeah, we pushed anyone who was linking to [??] documents – hey,
you can just embed that instead.
Andrew: Now was it? And were you looking for the bigger players, the guys
who had bigger traffic?
Tikhon: Well, we talked to anyone we could find. Everything from cold
emails to phone calls. I would still try to get an introduction whenever I
could and the YC Network was invaluable for that. But there was just a lot
of reaching out to people and prodding them and encouraging them, much like
Andrew: I saw that Jason of Docstoc [SP] used to – he saw that Digg was
doing well for traffic, too. He seemed to, I don’t know for sure, but he
seemed to just hire people to create content that would basically be
uploaded to his site and then linked to from Digg. So you’d suddenly see
these top 10 funny lists, you know, top 10 funny things you can say to your
wife on your wedding night and then a picture for each one of those
sentences. And then it was sent to Digg, where it was dugg and sent them
traffic. You saw that and what did you guys do? How did you guys take
advantage of that opportunity that he saw?
Tikhon: Yeah, so that’s actually really interesting. We did do a little bit
of that in the beginning, but we really wanted our users to be the ones who
were developing content on the site. In the very beginning we did have to
create some of our own content. But to be honest with you, I mean, our
users actually rated a lot, I mean, all the best content on Scribd was
created by the users even from the beginning. And all of the stuff I got on
Digg was always created by the users, although we tried. I probably made
quite a few of my own top 10 lists, but I never got anywhere. The users
were just better than I was at making that stuff.
Andrew: What about [??]? I don’t want to spend much too much time talking
about Docstoc, I just happen to know him because I lived near him
California when I lived there.
Tikhon: We’re big fans of Jason’s. I stayed at his house for a week in LA,
Andrew: I don’t get that, because you’re clearly in similar business. He’s
calling you up for a reason because he wants to learn about how you run
your business and he’s not just going to do a school project on it. He’s
going to get into a similar business and you’re still so close with him
that you end up at his house. How does he do that? How do you do that?
Tikhon: He’s a great guy and we shared a lot of ideas. And I think both our
products are better because of the other person being there.
Andrew: Give me an example. What did Scribd get to do better because you
talked to Jason?
Tikhon: We thought a lot more about professional documents than we had
initially. We had started off with this YouTube entertainment idea because
YouTube was so hot at the time, like the way Facebook is now. We were
thinking a lot about consumer facing, entertainment stuff and less about
business-to-business type documents. But, you know, Jason has done a great
job with that sort of content and at Scribd we think a lot about that now,
probably in part due to conversations with him.
Andrew: So, someone who was listening to us now must be thinking that’s why
you shouldn’t talk to your competitors. You should have them talk to you,
but you shouldn’t talk to them. Why would you talk to someone who is in a
similar space knowing that you just got an idea from him and you’re going
to go do more of the same kind of business he’s after?
Tikhon: Yeah. I mean, document sharing is such an enormous space that
there’s room for people who focus on these particular vertical niches. So,
yeah, I mean [laughs], it’s an interesting question because obviously the
competitors don’t always . . . you won’t show them your six month road map.
But I think you can learn a lot and both be better off by talking in
certain cases about the market in general.
Andrew: I get that.
Tikhon: Maybe not about your product road map, but going forward.
Andrew: All right. I want to come back to this idea of how talking to other
people brings out ideas. But, let’s just spend a little time . . . one more
positive thing that I noticed that he did and then I’d like to hear one
negative from you. I know, I don’t want this to become a lovefest. I want
to learn. So, first email collection. As soon as you wanted to download a
document from him, early on, he was smart enough to say, “OK, give me your
email address and I’ll create an account for you.
Andrew: What about you? What were you guys doing in a similar way, to
collect email addresses so you can build those long-term relationships with
people? What was your attitude towards it?
Tikhon: Our attitude was always, let’s focus on making the best user
experience. I hate giving out my email address online. It drives me crazy,
and I always make up fake ones, or use a service like SpamBox. So, we
didn’t do that for a long time. Now, we actually do. It does help encourage
growth, and a lot of those users do end up actually coming back and
uploading their own content. So, it’s a big win for distribution and
growth, but initially we didn’t do it because we thought it was not the
best user experience.
Andrew: It’s a different mentality. Southern California, where he’s from,
is very let’s build this business as scrappily as we can. Let’s lock users
in with email addresses if we have to. Let’s find ways to bring in revenue
quickly. And northern California is more about user experience. The way
we’re going to grow is by creating a better user experience, which takes a
little bit of faith, right? A lot of faith.
Tikhon: It sounds like a distinction like West Coast and East Coast rap
back in the 90s. Like, NorCal does things this way, without collecting
Andrew: But you know what? You don’t see the same kinds of conversations in
northern California as you do in southern California. They are very much
about ‘where’s the revenue?’ down there.
Tikhon: I think, just in this case, he’s seen a lot of successful companies
around legal stuff, like LegalZoom, and so he was really inspired to stay
in that sort of vertical niche, whereas we kind of focus more on the
YouTube model, sort of, let’s get all the content under the sun.
Andrew: And if you had to do it again, would you go after email addresses
Tikhon: Yeah, I think I would. If I was starting over again, I would. It’s
really good for distribution. It’s a similar reason to why sites like
LivingSocial try to get your email address whenever you want to do
anything. So, email addresses are worth a lot for converting people later
Right. I don’t know your revenue, but I happen to know his. Going out for
drinks with people is really helpful. They trust me with their drinks. He
didn’t specifically, but someone else did. And frankly, I’m always shocked
by how many people trust me with their data, and I take it very seriously.
I know I’ve never revealed anyone’s data that was private.
Tikhon: [??] all of it.
Andrew: Let’s what?
Tikhon: Let’s tell everyone all of the data right now.
Andrew: No, I can’t. You should see what people tell me before interviews,
and also in person. But here’s the thing. My understanding is, yours is
higher. Scribd is higher. Why? How did Scribd become bigger? Why is
Tikhon: So, I don’t know Docstocs’ numbers. We’ve just focused a little bit
differently. Scribd is more about being a social publishing platform for
all kinds of content, and we’re not really focused on a particular niche or
vertical the way some other sites are. We just have broader regional
appeal, I think, due to that.
Andrew: It’s because you went broader, and the scrappy guy who doesn’t have
as much money needs to go narrower, and needs to go after a smaller group
of people that he can monetize more? Is that what I’m getting from this?
Tikhon: I think it’s, if you’re not the first player in the space, you have
to have a new angle. You have to either be faster, cheaper, better, or have
your own vertical or niche. And I think Jason wisely chose an interesting
vertical and really focused on that, because there were a lot of people who
tried to clone us exactly, and none of them were any better than us. It
just wasn’t interesting. So, to unseat the incumbent you need to be either
dramatically better or focus on some sort of vertical or niche. Or just be
much cheaper, but we were free, so it’s hard, unless you’re going to pay
people to use your website. That probably wouldn’t work.
Andrew: So, bottom line it for me. They are smaller, as far as traffic, we
can say for sure. Why do you think they’re smaller than you guys?
Tikhon: They’ve focused on a vertical, on a niche. But I’m sure they’re
doing really well in that vertical and niche. I don’t know their numbers or
metrics. But we’ve just gone after a broader market.
Andrew: All right. Then you charged. And you and I talked before the
interview. You saw the pain I went through with my audience when I charged
Mixergy. You were watching that. What’d you think of it, by the way, from
Tikhon: It was a wise move for you. I know your audience had mixed
reactions, and at Scribd, it was the same thing, where some users were
upset about premium accounts for some features that used to have for free,
but a lot of folks were actually very supportive, and much like with Reddit
Gold, we saw some folks that were like, yes, we want to help you, we want
to make sure that you’re around for a long time and that you won’t
And I think with Mixergy it’s the same thing, your best audience members
are probably happy to pay you, because they love the product you’re
producing and they want to keep seeing it done. And so at Sribd, much like
at Parse, we had people begging to pay us even before we were actually
charging, just because they wanted to encourage us and keep us going, and
help us move faster, and so…
Andrew: How’d you decide what to put into the paid version and what to
leave in for anyone who wants to come in for free?
Tikhon: Yeah, so we introduced a lot of new features that were just for
premium users, but also we took a couple of older features that were not…
we didn’t want to hurt our existing user base which was very large already,
so it was only the pro features that really heavy duty users would need,
people who were making money via the website. We also launched a store, for
instance, where we let you sell documents, Docstocs has that, too now,
where people can buy and sell documents using Sribd. We introduced other
stuff besides premium accounts like that which has actually worked out very
Andrew: But you did take some features that were already free and you said,
these we’re going to have to start charging for.
Andrew: How do you explain that to your users? How do you do that right? I
completely flubbed mine. I didn’t even realize that we were starting to
charge. I had outsourcers take care of it for me, and I didn’t expect them
to do it when they did. I completely flubbed it, I completely flubbed the
discussion of it, I let someone discover it and then talk about it instead
of introducing what was happening. It was crappy the way I did. But you did
it right. How did you do it?
Tikhon: Yeah. It was tough, we got a lot of blow back initially, too.
People don’t like when you take things that were free and then suddenly
charge for them, but eventually almost everyone kind of came around and
they didn’t mind. Our premium accounts are very cheap. It was tough though,
it wasn’t the best day, the day that we launched that. Some folks were not
thrilled about that but at the end of the day most folks came around.
Andrew: Teach us how to do this right. I know that there are people in the
audience that are thinking that they want to start charging for something.
How do you do it right?
Tikhon: The way I’m doing it this time around with Parse is that we’re
adding new functionality that larger apps would need and charging for that,
for our pro and enterprise accounts, but were not taking away features from
people and charging for stuff that already was free. I probably wouldn’t do
that again or I would try to come up with new killer features that people
really wanted to pay for.
Andrew: And how do you figure out what those new killer features are?
Tikhon: Well, for Parse it’s easy because people keep asking for them, so
every day we’ll get hundreds of emails asking for various features over and
over again, so that becomes clear. But also, we talk to a lot of the larger
apps on Parse like Band of The Day, and we get a lot of ideas from folks
like that, specific things that they want and then it becomes obvious very
quickly that many more folks than just them are going to want this.
Andrew: Let me see. I want to make sure that I get everything here at
Scribd before I fully move on to Parse. Headlines. You saw the way that I
introduced you, I told you I was going to introduce you as the drop-out who
went on to build two hit start-ups, and you said that’s good.
Andrew: And you used to do that too, you told me. How did you guys do it
and what was the impact of headlines for you guys at Scribd?
Tikhon: Yeah. That was really important, having good content in the early
days, and getting good content onto sites like Digg and Reddit. Having
good Linkbaby titles was really important and it still is to this day, this
is why it’s almost a joke now if you title anything “The Top Ten Ways to Do
This.” On Hacker News they’ll actually remove that from a title because
it’s so annoying now, but back in the day we looked what titles were
popular on sites like Digg and Reddit, and we actually did some analysis on
them and kind of figured out that if it says the top 10, a lot of those
stories do really well. And so we noticed just that stories or documents at
Scribd have titles like that, that have [??] titles did better on Scribd
also. So, yeah. We definitely did do that to our advantage. And also for
SEO as well. Words like best. That if you include those in the title,
that’s very valuable in terms of SEO.
Andrew: How did you learn SEO, Search Engine Optimization?
Tikhon: That was 100% from scratch. We had not done any of that before
Scribd, but we were kind of thrown into the fire and had to learn it very
quickly. Luckily we had a lot of great advisers who could help us, but I
know much more about SEO now than I ever wanted to know [laughs].
Andrew: You just started learning it internally by talking to one person at
Tikhon: And also just running experiments. We ran a lot of experiments.
Every day we’d test out new ideas.
Andrew: How do you run experiments for Search Engine Optimization? Because
it takes so long to get results, doesn’t it?
Tikhon: Yeah. You can A/B test. You can test based on random sections of
your documents, of your content. So, you can take a random sampling of your
content and, you know, do different things to it and see what happens to
that. And actually Google, for instance, was crawling us so quickly that
we’d get results very, very fast.
Andrew: Wow. Do you have an example one thing that you did that have
outsized [sounds like] results?
Tikhon: Just in terms of SEO?
Tikhon: We didn’t initially really understand what to put in H1 tags and
what not to. Putting the right content into, say, the H1 tags is now
obviously a no-brainer. But at the time we were still figuring that kind of
stuff out. We learned a lot about what to put in, for instance, it’s a bad
example, in the H1 tags.
Andrew: OK. I wanted to come back and talk about ideas and having
conversations with people and how that brings about ideas. Who do you – it
seems to me that yes, you want to try a lot of things and yes, stuff just
comes to you. And if you can crank it out within 24 hours, then you’ll get
feedback about whether to build it or scrap it and so on. But I know that
when I go to a conference and I sit down for drinks with a couple of the
speakers for the conference or get to have a conversation with guests from
the conference, I get all these ideas that I want to go back and try
because we’re having this conversation as two people who are trying the
same thing. You seem to have that a lot because you’re a Y Combinator,
because of where you live, right?
Andrew: How do you do it so that all of these conversations that you get to
be a part of are useful and not just, you know, a waste of time?
Tikhon: Yeah, it’s a good question. I think for Scribd we really talked to
the users who really loved the product and then the users who were eh about
it and got a lot of great feedback from both of those groups. But there
wasn’t as much great feedback from people who just say, I mean, there
weren’t that many, from people who just hated the product. Some people just
hated Flash, for instance, and their feedback wasn’t super helpful to us.
We weren’t going to rewrite all of that logic overnight. So, yeah, sure,
you hate Flash, thanks, don’t come back.
Tikhon: Sorry. Hope you come back, but we know you won’t. So, folks like
that who had Flashblock [SP] installed probably weren’t as helpful. In
terms of conversations about startups in general, I mean, lucky to know a
lot of great people thanks to Y Combinator and the various folks out here
in the Valley. That’s been just absolutely key to kind of, even running an
idea like parse, I got a lot of help from, say, James Lindenbaum from
Heroku. He’s a good a friend. And we actually helped introduce them to
Redpoint [SP], who did their Series A. So, it all kind of comes back. We
help someone out and a few years later they’ll help you out.
Tikhon: Why did you leave? Why didn’t you stay with it the way Phil was
going to stay a 100 years with Evernote? Why did you leave Scribd?
Andrew: Scribd is a great company. It’s doing really well. It’s over 40
people now and some. I love it there. I really wanted to get into mobile
stuff and into making a product for other people like me who were building
mobile apps at Scribd. I saw how difficult building mobile apps was for us.
We kept reinventing the wheel, rebuilding all the same logic and
infrastructure that everyone else was also building at all of my other
friend’s startups, from JTV to Hipmunk, to everywhere else. The situation
just seemed insane that we’re all asking each other how to solve these same
exact problems, so, together with James, Ilya, and Kevin, we really wanted
to build this platform that would solve the problems once and for all for
Andrew: I’m sorry, I want to understand where this idea came from and give
Parse the attention that it needs, but first I have to understand, if
you’re at Scribd and it’s going well, and you’re rocking it, when you have
a new idea, why don’t you say, hey, you know what? I’m going to find
someone and I’m going to give him that idea, and I’ll invest in their
business or I’ll find someone and I’ll be an advisor to them and I’ll make
sure that they understand Parse and they get to meet my friends that have
the same problem, and I’ll just focus on Scribd. I’m not ragging on it, I
just want to understand the way of thinking that gets you to build one
successful company and then move on to the next instead of focusing more on
the one that’s working.
Tikhon: Yeah, I mean, Scribd is awesome, it’s a great company, I just
personally wanted to get into the mobile stuff. At Scribd we noticed more
and more of our traffic was actually coming from mobile devices. Just the
pain in building our own mobile apps internally to me was inspiring in the
same way as like sharing documents had been so difficult before Scribd.
It’s always [??] to do something new too, after, I think it was over five
years I was at Scribd. So, I felt like the company was in great shape, we
had built up a great company with great people, a great infrastructure, and
I was able to leave in such a way that the company would still be in great
shape. The idea of building a mobile platform for me was really compelling,
it’s really fun, building something that’s kind of for yourself.
Andrew: Do you take money off the table at Scribd? You did the last round,
Tikhon: At Scribd, no, no.
Andrew: You didn’t. Still… all right.
Tikhon: Still chugging along. We’re obviously huge fans of the product and
there’s some really exciting stuff that’s going to be coming out of Scribd
soon. I’m really bullish.
Andrew: OK. All right, so then, you told us where you had the idea for
Parse. What’s the first thing that you did to test it out? What was the
basic prototype there?
Tikhon: Testing out Parse? That’s a good question. So, we built out a very
simple engine that some other folks at Y Combinator, in our own batch, were
kind of starting to use and we had folks at… that’s one of the nice
things about building products for other start-ups or companies, we had
hundred of guinea pigs right there in our batch at Y Combinator who we
could test things on. So, we learned a lot, just because a lot of the
companies were building mobile apps, just from having them kind of test out
our stuff for us. And then we dog-fooded our own product, too. We would
build apps on our own platform.
Andrew: What is this thing that you said you guys were all doing over and
over in mobile development? That you saw yourselves at Scribd doing over
and over, that you saw your friends at JustinTV and other companies do.
What was it that you saw that too many people were doing? The exact same
thing and re-inventing every few weeks the way that you were. I want to
understand the problem a little bit better.
Tikhon: Mobile development is so difficult compared to web development.
Mobile development today feels a lot like web development did in the mid
90s. None of the modern tools exist and it’s just really difficult and slow
and everybody is rebuilding the same infrastructure and guts over and over.
Andrew: For example, what kind of infrastructure and guts are you building?
Tikhon: Storing data to the cloud from the app on your phone. Storing data
to the cloud and then syncing it across devices so that if I’m halfway
through a document on my phone, my Android tablet should also know that.
Stuff like that is actually very difficult and everyone is building their
own half-baked implementation of stuff like that. Handling push
notification, so if user A sends a message telling… sending a push
notification to user B, stuff like integration with Facebook and Twitter so
people can log in with Facebook and Twitter via [??]. That’s actually a
real sticking point and it’s difficult. How does your mobile app know that
if you sign in with Facebook one day and then Twitter the next that you’re
still the same person? Stuff like that.
There are a lot of common problems in mobile development and also on the
client side. How do you handle the problem of someone being on a train…
there are problems that are totally foreign to web development. Say someone
is on the BART, and they’re online, suddenly they’re offline, and they try
to save something, and then they’re back online again. That’s a problem
that we never had at Scribd because with the web you assume that people are
Tikhon: With global there’s this whole new set of problems around online,
Andrew: So, of all these different things… I see now, I didn’t understand
the problem fully, now, just in that one answer you gave me, I’m
overwhelmed by the problem. So, now, I don’t know where I would begin if I
were to be in this business. Where did you begin? What did you narrow down
your first version to?
Tikhon: So, our MBP or our core first product was just storing data from
the cloud to your app, which was a really common problem that people had
and we wanted to make it dead simple. And that’s what we did. So, we
started off just with that, and that was actually already initially, that
got a lot of traction from various folks logging, really interesting kinds
of data from their apps. And then, we had people starting to build their
entire app just based around Parse, as the backend. So, using the entire
platform. Then we added stuff like user accounts because most apps had the
concept of users. But it’s crazy for everyone to build their own half-baked
version of that.
Andrew: So, you told me that one of the reasons that you decided to focus
on storage first was that it was such a common problem. Why else did you
pick it? I’m learning about how people learn, because at Mixergy premium
were doing courses, and one of the things that I’ve discovered is that you
need to give people that quick win. And if you give them that quick win
they want to know more and they want to continue further and so on. Were
you looking for something that you could do quickly and give yourself, give
your company a quick win on? Or were you looking for the toughest problem
to see if you had what it took to continue?
Tikhon: No. We looked for an easier problem that we could tackle early and
get feedback on, and that’s what we did. And also, you’re 100% right about
the quick win thing, we actually have a guide called “Quick Start”, which
in 60 seconds you can have an app that’s using Parse and that has stored a
piece of data to the cloud. That’s how we get you introduced to Parse, is
actually by using the product immediately upon sign-up. It’s kind of a
great feeling when suddenly you realize that you actually already have a
working app that is already persisting data to the cloud and can sync
Andrew: All right. And how long did it take you to build that? The first
Tikhon: We actually launched our beta within about two months of starting
Parse. So, roughly, two or three months.
Andrew: And how did you get, I mentioned in the beginning that you just
got… I had the number wrong, I thought it was 1,500, it was 15,000. How
did you get 15,000 apps? Was it one of your friends at Y Combinator who
decided he was going to launch 1,000 apps? Is it a collection of those
people? What was it?
Tikhon: [Laughs] No, we do have a lot of consultancies and agencies who use
Parse, and a lot of specific people are pumping out an app or two a week.
So, those are really interesting people, or clients. But, a lot of it is
thanks to the marketing of Y Combinator, we’re on Hacker News a lot, we’re
on Stack Overflow and Twitter. Mobile apps, this market, as you know is
growing exponentially, the number of apps that are in production is more
than doubling every single year. It’s such an enormous market and it’s just
such a huge common pain points, where people want to, just for example, to
store data to the cloud from their app, store some piece of data maybe, and
Parse makes that really difficult problem just dead simple. So, the
distribution has been frankly really easy in that sense. And, since we’re
all developers, the entire company, everyone here writes code, we’re all
engineers, I wrote code yesterday. We’re all technical, so building
something for ourselves we kind of know where to go, where people like us
hang out, where mobile developers are, so distribution’s been pretty easy,
Andrew: So, is there any business development that goes into it? Or is it
just because you happen to be at a time when this industry is growing so
quickly and this problem is so pervasive that just showing up means that
you’re going to well. Showing up and building the product for people.
Tikhon: Right now we do our own business development, so the folks here who
are also developers do business development.
Andrew: So, what do you do? Can you give me an example of some business
development deal that you put together?
Tikhon: We’re announcing a couple of really interesting partnerships soon
but I can’t talk about those yet.
Andrew: People trust me with their finances, you can’t trust me with that?
I don’t think you should. How about one that’s already publicized?
Tikhon: What’s a good one? The best three are coming out very soon. So,
those are the ones I’d probably stress the most.
Andrew: Don’t give me the name, just give me the industry that they’re in.
Can you do that?
Tikhon: There are some large public tech companies that we’re going to be
announcing some cool stuff with.
Andrew: OK. So, how do you get one of these large, public tech companies to
take your call, to trust you, to build a relationship, and then to use you?
Tikhon: The first one approached us, they believed that we had a really
cool product and they’ve been talking to their customers who had been
complaining about how difficult mobile development has been, and so they
approached us, actually, so we haven’t done, most of our [??] has been
responding to inbound request, actually. We haven’t had to do a lot of; we
haven’t really had to do any cold calling or cold emailing, that sort of
thing. So, it’s a nice market because it’s already really large but you
know that in a few years mobile apps is going to be, there’s going to be an
enormous number of mobile apps, there already are, but…So a lot of folks
are trying to figure out their own mobile strategy right now so they kind
of approach us.
Andrew: All right, you’re making me feel like we should be doing more, I
keep feeling like I’m doing too many mobile interview, with too many
founders of mobile app based companies, but I think we should do even more,
even maybe more courses. We just turned down another one, but I’m probably
wrong not to.
Tikhon: If you look at the graphs mobile is already on par, well, in a few,
there are already more smart phones than there are PCs out there and this
trend is just accelerating right now. There are billions of people who will
never own a PC right now, who don’t have one now, who have Android smart
phones, right. It’s just an amazing time, I think in 10 years we’re going
to look back on this time period and just be blown away, like wow, I can’t
believe how early that was, before people had mobile devices. It’s going,
the new opportunities and business, and companies that are going to be
started out of this mobile revolution, it’s a real exciting time period,
and it really feels like the web. Like I said, in the early 90’s before
Google, and Facebook, and Twitter, and all these monsters, so I think
you’re going to see a lot of new innovation thanks to mobile, all the new
opportunities that mobile has opened up.
Andrew: I feel it, too. I’m stuck in my ways when it comes to a computer,
for the most part, but on the phone I’m trying everything. I’m trying new
apps, I’m trying new sites, I’m trying new activities that I hadn’t done
before, and so yeah, my appetite, if it’s any indication of what’s to come,
I know that it’s huge.
Andrew: Let me do a quick plug and then I want to ask you a question, one
final question, and the plug is actually for Mixergy premium, I wanted to
introduce the audience to how an ambitious entrepreneur in our audience
solves a problem. In the comments about one of my recent blog posts about
systemizing, Antonio, the founder of Hootsuite, said that he found himself
working 90 hours a week and his business was in chaos, and I know because
mine was like that too. I was working 90 hours, I was feeling really
arrogant because I though, hey, if I’m working 90 hours I must be beating
everyone else who’s working only 40 hours, or even 89 hours, but my
business was in chaos too, and I didn’t have time to think, and it seems
like Antonio was in a similar situation.
What he says is, he says he implemented the lessons from five Mixergy
programs on systemizing your business, systemizing the way you think,
systemizing the way you organize your process for other people in the
company to work with you, and as a result he doubled both his profits and
customer satisfaction in one year. He says, quote, ‘I can’t wait to improve
upon this as I have so much more to systemize.’ And it’s true, once you get
some order in your life, you want everything to be organized because you
see the power of it, and if you implement this, if you take those courses
the way he did and you implement those ideas, it could be worth thousands
of dollars to you, but…
Tikhon: You’re cutting out.
Andrew: Oh, am I? Oh, that’s all right, the recording is coming in OK on
this side. I’m cutting out on the most important part too, which is just
that Mixergy premium doesn’t cost anywhere near thousands of dollars. If
you go to mixergypremium.com you’ll see exactly what it costs and I
guarantee it, it’s the same price for everyone, I just don’t want to reveal
it here in the interview, I want you guys to go to mixergypremium.com and
see, and I guarantee it, I guarantee that it’ll work for you or your money
back. Here’s what Antonio says in that comment that you can see on
Mixergy.com, he goes ‘for anyone considering Mixergy premium, you’re crazy
not to jump on this informative, not to jump on this right now, Andrew is
way undercharging for the value you receive.’ And I agree, so go to
mixergypremium.com, low raise prices, that’s why I don’t want to say them
now, but if you go now you lock it in forever, mixergypremium.com.
Tikhon: I agree, [??] to that.
Andrew. Thank you [??], I appreciate that. So, I got introduced to you from
the founder of Weebly.
Andrew: Fantastic freaking guy, I didn’t realize how bright he was until,
until I did this interview, and actually, until I prepped for it and Paul
Graham told me how bright he was.
Tikhon: You’re cutting out [??].
Andrew: I was going to ask you, who should I interview, who do you know
that I should have on here as a guest? Oh, we lost the connection. We were
Andrew: Go for it.
Tikhon: Oh, sorry, you were cutting out. Where you asking the question,
was, who should you have on as a guest?
Andrew: Yes, who do you know that I should be interviewing the way that I
just interviewed you?
Tikhon: I imagine you’ve already interviewed Dave Rusenko from Weebly,
they’re one of the most under hyped companies right now around. They’re
actually; they’re a goliath that no one even knows about.
Tikhon: Justin Kahn is a really entertaining interviewee from exact, Justin
T.V., Twitch. Steve Huffman [SP] is one of the best entrepreneurs I’ve ever
met, he was actually really helpful for me in starting off Parse. He
started Reddit and Hipmunk, just a brilliant guy. Also, his co-founder,
Adam. Emmett Shear [SP] from Twitch TV, great guy.
Andrew: Can I hit you up for an introduction to Emmett and to Adam?
Tikhon: Absolutely, anytime.
Andrew: All right, I will and I’m looking forward to having them on. I know
that people are already rushing from this interview to go to, you own
Parse.com, right, you have the .com?
Andrew: That’s impressive. All right, I won’t put anymore strain on our
internet connection, I’ll just say this, thank you for doing this
interview, I appreciate it.
Tikhon: OK. Well, hey, Andrew, thanks so much, it was a pleasure chatting
Andrew: Same here, thank you all for watching, bye.
Walker Corporate Law – Scott Edward Walker is the lawyer entrepreneurs turn to when they want to raise money or sell their companies, but if you’re just getting started, his firm will help you launch properly. Watch this video to learn about him.
Grasshopper – Don’t make the mistake of comparing Grasshopper with other phone services. Check out their features and you’ll see why Grasshopper isn’t just a phone number, it’s the virtual phone system that entrepreneurs (like me) love.
Shopify – Remember the interview I did about how the founder of DODOCase sold about $1 mil worth of iPad cases in a few months? He used Shopify. It’s dead simple and very effective. To get a longer free trial, use this code: Mixergy