start

How to Start/Create/Build a Software Company - Hamid Shojaee PHX Startup Week

thank you very much appreciate it so as

was just covered my name is hameed

shojai ii my twitter handle is at hamid

s so if you're going to be tweeting you

can tweet at me

you can include hashtag Phoenix startup

week and hashtag yes Phoenix if you're

really into hashtags you can also

hashtag chase Basecamp I'm not sure why

you would you could just do a collection

of a bunch of hashtags and call that

your tweet that's pretty popular these

days so I'm going to be talking about

how to build a software company and the

lessons that I've learned and let's get

started I want to leave a little bit of

room for Q&A towards the end so I'm

going to just run through a bunch of

different things that I want to talk

about quickly so first question to ask

is why even a software company and the

answer to that is duh because software

is the absolute best business in the

world to have or to be in it's super

easy to get into you can if you're a

software engineer you all you need is a

computer and your own skill set to get

into the software business very very

easy very low barrier to entry

it's literally hundreds of dollars to

get into the software business it can

change the world as all of our lives

have been impacted in one way or another

over the past few decades with software

it's highly scalable meaning you can go

from zero customers today to a million

customers tomorrow if you happen to have

a hit on your hand as has happened many

times in the App Store as well as

web-based applications that literally go

from zero users to millions of users

overnight to multi-billion dollar

valuations it's just an incredible

business it turns out software is also

one of the worst businesses in the world

for largely the same reasons because

it's so freakin easy to get into anyone

can get into it from anywhere in the

world so you could be competing with

your sort of living style requirements

that maybe multiple thousands of dollars

per month in living expenses with

someone in Bangladesh whose living

expenses or only 50 dollars per month

and guess what he can actually sell or

she can actually sell her software to

people in the United States who are your

neighbor

so that happens all the time and in fact

it's the one of the most competitive

businesses in the world for those same

reasons

just taking the ios app store as an

example there's over one-and-a-half

million apps in the App Store right one

and a half million apps think about that

for a moment because if there's even a

thousand different categories of apps I

can't even imagine a thousand different

categories of apps that means on average

there's 1500 different apps per category

so how are you going to compete in that

market space so it's very very

competitive it's very cutthroat the same

software can be sold for thousands of

dollars or even given away for free

there's no sort of cost of goods sold if

you will in the software business or

it's so low is sort of negligible so the

question we're going to try to answer is

how do you build a software company

that's going to sustain that's going to

grow and is going to have the most

likelihood of success that's what my

talk is about if you're not interested

in software this is probably not the

right talk for you to give you a little

bit background about myself I've started

three software companies everyone's

limited by their experiences so my limit

is based on these three experiences that

I've had I started a software company

with my college buddy in 1995 called vie

Trix we went out and raised a bunch of

money we raised hundreds of thousands of

dollars to start the company and over

the course of the next five years we

raised a couple million Burt through

most of it we made all sorts of mistake

it was a software company targeting time

and attendance software by the way do

not Google vie Trix anymore they gave up

on the domain name and yeah it's not

sufficient for work NSFW whatever they

call it anyone so the image searches is

terrible I was looking for a logo of vie

Trix and I had to like dig quite a bit

to find my old business cards but anyway

it was a funded company I after five

years I gave up on it and chalked it up

into this sort of failed but learnt

lessons learned category and went to

work for Microsoft I then started

another company called AK so soft in the

software development tool space in AK so

soft we grew that organically and

profitably it's going strong still it's

it's a multi-million dollar success

story and we

created a number of different products

in there including transfer big files

one of the soft side projects of access

oft was a product called pure chat that

has led to the sort of next thing that

I'm focused on right now and pure chat

is a funded company we split it off into

its own company raise money around it we

raise one and a half million dollars

it's still burning cash but it's growing

very rapidly we expect to be profitable

by the end of this year so that's my

background and that's sort of where I'm

coming at things from so of course the

first thing you need to have if you're

building a software company or any sort

of product based company is a product

and generally speaking a better

mousetrap will have much higher

probability of success than the same old

mousetrap but it turns out there's

something else that makes the product

more likely to succeed and that is this

concept of market timing fit different

than product market fit market timing

fit is just a timing issue of when

you're entering into a product market so

every market goes through this growth

cycle that looks something like this the

the number of potential users that can

use the product we'll call that a

hundred percent of all of the users in

the universe that can use it and that

number could be you know depending on

what category you're targeting it could

be in the hundreds it could be thousands

it could be millions or even billions so

if you were to look at email for example

today has billions of potential users in

the 80s email was at sort of the very

beginning in the late 70s and 80s it

took it a while to get into here maybe

this was sort of around the 90s

timeframe and growing rapidly in the 90s

probably by 2000s it was over here and

today I would say email is over here so

if you are sort of creating an email

product there's a good time to get into

this market and there's sort of a bad

time and there's a very difficult time

and those times look something like this

this is a great time to get into the

business into any market think of it as

you know like in any product category

you can sort of break it down in this

way and there's a bad time which is

right at this stage where the market is

already largely saturated

the people who could use this product

category already have a solution so

email today would be one of those market

and then there's a really hard time

which is at the beginning before people

even know that they need email you have

to go out there and convince them if

you're trying to build this sort of

market category you have to convince

them that this is something that's

really good and you need it before they

actually are like I don't know what

would I do with email seems kind of

stupid I could fax things and you know I

could print memos and give it to my

employees so it's really hard to get

started and building the market category

another example of this might be you

know early social networks like MySpace

and Friendster and others sort of built

the sort of need like helped get the

market started got it into the millions

of users then Facebook came along and

sort of really Facebook was sort of a

later entrant into the game but you know

like they came in at this great time

when the market wasn't already saturated

and they sort of took up a book of this

market share and now dominate the space

so today it would be too late to get

into the sort of social market social

media space right trying to compete with

with Facebook that doesn't mean you

can't disrupt things for example in

emails case something like a tool like

slack is coming at it from a different

angle it's not making a better email

client slack doesn't do email it's a

better communication client which is

sort of the same problem that email is

trying to solve so it turns out that

slack is disrupting email in in a sort

of coming at it from a different angle

and making a lot of headway as a team

communication tool eliminating the need

for email internally at least for

internal communication and interestingly

slack is now a team communication tool

which is probably sort of in this sort

of timeframe so if you were to go into

that space you know like right now is

probably at the later stages of when it

might be a good time to get into that

space just you know as an example of

another market space so part of the

reason why the sort of there's a great

time and sort of bad

is that at the beginning stages of it

you work really hard a lot of time so in

the case of email maybe a whole decade

before you build maybe 10% of the

potential users for that space over the

next decade you know 30 40 percent more

people come online over the next decade

in the case of email the remainder of

people are coming online or you know a

bulk of the people so you could sort of

you know in in most of these markets

that are going through these cycles and

it in most cases it's not decades most

markets will saturate in just a matter

of a few years you could not get into it

for the first year or to come in with a

much better mousetrap and still be able

to capture the market as long as you do

have a better mousetrap and sort of

solve all the problems that the early

people did not see like pricing issues

and other things that might allow you to

capture market share so so the next the

next thing and I'm not sure why this

goes to slides at a time or there we go

so the next thing that I hear all the

time is with respect to building a team

or outsourcing and there's a lot of

people who have great ideas for apps or

various different software product ideas

that they want to bring to market and I

have in the course of the past five six

years as I've been active in the tech

community I have spoken with a number of

people who have spent 50 thousand a

hundred thousand even two hundred

thousand dollars on the development of

an app and then ended up abandoning it

because the product never functioned

exactly the way it should and it turns

out that software is very difficult to

build especially if you're not building

it yourself

so if you're thinking about building a

team to build a software company which

is very very expensive to do or

outsourcing it which is relatively cheap

to do is probably a 1/4 1/5 the cost to

get to sort of a v1 a lot of people will

choose outsourcing but it turns out

outsourcing often times more often than

not is throwing your money away because

what happens is software is an iterative

process you have to build something and

then you have to improve on it your

first version is never going to be the

right version so it

it turns out that you'll be throwing a

lot of that money away and if you're

outsourcing the people will not have

that continuty of thought your users are

too far removed from the developers for

them to be able to get that feedback

cycle going rapidly and iterating

through it so if you're thinking about

it

software company think about it like

this what is the core competency of the

company that you're building if you're

building a software company the answer

to that question better be software

right so if you're outsourcing your

software development you're outsourcing

your core competency as a software

company so what are you exactly a sales

and marketing company that's ok if you

are then it's best that you know that

ahead of time you're not a software

company so you might want to partner

with a software company or bring in a

tech co-founder as a lead it if you guys

were here in the previous talk and and

make sure that that you get that portion

of your technology addressed otherwise

you're going to have a hard time

building a software company so as has

been mentioned in previous talks as well

the the clean router one was a perfect

example building incrementally Minimum

Viable Product and sort of moving on and

iterating quickly is one of the best

pieces of advice that anyone can give

with respect to developing software you

want to tackle on a small sort of set of

problems features that you can do really

well with your product solve that

problem really well get it into the

hands of users give it away for free and

I don't know why I keep going multiple

slides but give it away for free if you

have to not if you have to it's actually

a great practice to do so and get that

feedback build on it release again and

do that whole cycle over again and the

reason you want to give it away for free

and a lot of people sort of resist this

and there's a lot of advice out there by

people who say don't give your product

away for free people will not have value

for it you need to charge from the

beginning you need to have a business

model and those are great pieces of

advice I haven't found them to work very

well in the software business but the

reason is because of similar sort of

back to that graph of market growth is

that if you're entering the market let's

say at the right time between you know

like if you can release early get

feedback between the first year of

development so let's say that's between

that timeframe and then the first let's

go to here the number of users that come

online in that time frame is a small

fraction of the entirety of the market

so if you can establish yourself as a

leader capture market share you can keep

that market share going keep your

product free for all of the existing

users so you don't you know do a

bait-and-switch on anybody

or gradually increase their price to a

normal price but what's more important

is that you're going to capture a bigger

piece of the next group of people who

are coming online for your for your

product category and as you capture this

next piece that's when you start

charging and actually have a very

significant and fast growing presence

right so that's why you should give away

your product for free initially as

you're sort of going through it and it

also does another thing which is quite

interesting which is similar to putting

a beta sticker as the clean router guys

did is that people are a lot more

tolerant to free products not working

exactly as they should now even if

they're paying ten bucks a month for it

they are very very critical so it works

out really nicely the last lesson that I

want to talk about a little bit is

experiment with pricing with software

you can sell the same product for free

ten bucks a month $100 a month $1,000 a

month a hundred thousand dollars a month

if you if you depending on the number of

customers that you want to have and all

of those prices are actually valid for

software because literally it takes

millions of dollars to create most

products so if you're only going to have

tens of customers you should be selling

your product for hundreds of thousands

of dollars in order to make a profitable

company but if you're going to have

millions of users then you probably

could get away with very close to free

or near free five dollars ten dollars or

whatever and what you want to do is

experiment with pricing especially if

you don't know the size of your market

you could experiment with pricing all

the time

sort of get things right now there are

certain things that you can do like

experimenting with pricing starting from

low to going high which are relatively

easy to do going from high if you

already have revenues down to low it's

very very difficult to do so might be

good to start the other way around one

example of the company that went from

high to low is Apple Apple had a product

called web objects when they acquired

next from Steve Jobs back in the late

90s they sold this product for $50,000 a

license and overnight they decided oh

let's sell it for five hundred dollars

and see what happens

of course they cut the revenue from that

product significantly the existing

revenue but the volume went up very very

rapidly as a result and it gave web

objects more of a chance to be

successful eventually it failed and they

sort of candidate but just an example of

a pricing experiment Google has bought

companies before that have charged

hundreds of dollars for their product

and then made them free like google

analytics used to be another product

that was four hundred dollars a month as

an entry fee and now everyone has it for

free super interesting right so you want

to experiment with pricing or not what I

want to do is share with you some of the

experimentation we've done with pricing

and then we'll go into QA right after

that so with pure chat we started for

free and we did this sort of exact same

thing where we gave away our product for

free had lots and lots of users we had

over a hundred thousand sign up accounts

sign up for pure chat before we ever

charged a single dollar for the product

then we came in from the bottom we

wanted to charge as little as possible

so we charged $5 per user per month and

then established an average revenue per

customer of which is by the way

something you should know average

revenue per customer

a RPC or ARP you per user of eleven

dollars per month because our average

customers were getting two point

something users per customer after a few

months of doing that then we decided

okay let's let's add another pricing

plan eight dollars per month where we

tweaked some features we said okay if

you wanted the powered by pure chat to

not be there

we're going to charge you $8 per month

for that and minimum is two users per

per account for signing up for the $8

plan and our average revenue per

customer jumped to $16 per customer at

50% growth a increase in our growth rate

essentially and your question should be

well did that do anything to conversion

rates because if you converted fewer

customers then they might not be worth

it and the answer is that there wasn't

any noticeable change to our conversion

rate so we did that for a few months and

the next thing we did is we came up with

a sort of set of buckets where we said

okay if you have up to three users our

entry plan is $15 per month it's still

$5 per user per month but we charge you

for three right off the bat the next

bucket is six users and the market after

that is 15 users and in fact those are

newer experiments we did ten users and

25 users and we wanted to see what would

happen if we did this type of a plan

because if our average revenue per

customer was 16 we wanted to see how do

we move that average revenue up a little

bit and it turned out that our average

revenue per customer shot up to $30 per

customer almost double so our growth

rate overnight doubled just like that

and of course the question of whether or

not it affected our conversion rate was

answered by this it actually increased

conversions which we were kind of

surprised about now the increase in

conversions is kind of hard to explain

we don't know exactly the reasons for it

but we have a bunch of theories right

well we know the product guys thinks

it's the product the support guys think

it's the support the marketing people

think it's all the marketing and content

efforts that we're doing which is all

probably true but one other thing that

happened is that we were now giving away

more users per account type and it might

have been that $5 or $15 it's just such

a low barrier to entry that that didn't

affect business people's decision

because it's a business product so at

$15 if you already have three users

you're more likely to add your three

users and add a couple other people to

respond to chats and therefore you get

more usage out of the product

for the for a relatively low price

rather than having to think about

whether or not you want to add your next

employee as another user of pure chat so

we had this sort of like interesting

things that have developed as a result

of pricing experiments and we have many

pricing experiments we're going to do

but just sort of giving you a taste of

things that we've done one other thing

about pricing experiments is that you

want to do them without affecting

existing customers so because we're

moving price inching it upwards in some

ways we're keeping all of our existing

customers on the old plans we call them

legacy plans without affecting them so

our average revenue for total customers

it hasn't doubled but our growth rate

has doubled meaning like then new

customers we're adding are paying us

more and as a result the growth rate has

doubled so in summary there's five

things that I talked about market timing

fit building a team as opposed to

outsourcing developing products

incrementally giving it away initially

to get sort of your initial early

adopter feedback and cycling through

that iteratively building the product

there and then experimenting with

pricing all right thank you