Fog Creek Software
Discussion Board

eternal business cycle

I always find it amazing how businesses remain
in business (well some). A friend of mine works in a retail
store selling clothes all day long. It is a small
store on a relatively high profile street in a
city of about 500,000. On average they sell anywhere
between $2000 to $4000 daily. On special days of the year,
they sell up to $15000 in a single day!

What is amazing is that I just can't believe how much
people spend buying clothes in that tiny store. This is a local store with other similar ones elsewhere in the city. Not to mention the ones right next door selling more or less similar items. I don't know how many unique customers
they have. They surely have a lot of competition even on the same street.

It appears they sell anywhere between 100 to 400 items per day. There are enough many people buying that many
things in a day all year round. Some could be compulsive
buyers, while others rip their clothes all the time, but
I just wouldn't think so many people would shop all day
long. Do these people have time to work?  :)

It could be said that this store has a repeat business
since when you buy clothes, you could still buy more.
Repeat business is great since you keep selling the same
thing over and over. Like the Gilette blades that I love.
But what about FogCreek? CityDesk is not necessarily something you keep buying over and over. Once you make a sale, until the customer upgrades or something, you are done. You need to keep pulling in new customers to buy your product. The cheapest CityDesk is $299. If you were indie, you'd need to make -let's say- 80,000/year to live in New York. That means you need a little over 20 sales per month which doesn't sound that bad, but you have other employers, business expenses, etc.. I don't know how much all that is, but let's say you have to sell a copy of CityDesk every day. Even that puzzles me... You'd have to convince new people every day to buy CityDesk. Granted there are hundreds of thousands of people out there, but still, it is mind boggling to me that every day (until you have another product I guess), you have to keep convincing new people to buy this one. It is extra tough
that you are not selling something people need for survival
like food or drugs.

This all makes more sense when I am in a plane flying low
over a city. Then I get to see all the hundreds of little
houses side by side with about 2-4 people living in each. Then the math of big numbers come into focus a little. I try
to picture how much toothpaste all these people would use
or how much garbage is collected every week.  :)

There is no question here really, but maybe you can comment on my ramble and bring some more sense into the whole process. This is the kind of thing that would keep me away from starting a business. Would you mind sharing your thought process before you started FogCreek?

Wednesday, April 21, 2004

I guess there are more people in the world than you might think :)

Every business model is a combination of "multiply a very large number by a very small number."

E.g. 1,000,000 people in my city; if only 1% buy a $1.99 toothbrush from me I've made $20K!

Or what you always see on powerpoints intended for VCs:

The world market for middleware according to Gartner will be $3b in five years. If we can capture just 0.1% of that business we'll be making $3million a year!!!!

The trouble with multiplying a large number by a small number is that tiny variations in the small number, which you pulled out of a hat, result in large variations in the result. So every time I see a claim that business X can make lots of money because (multiply large number by small number) I know somebody is grasping for straws.

The truth is you can't really tell how much money you're going to make selling something until you start doing it.

Joel Spolsky
Fog Creek Software
Friday, April 23, 2004

> Like the Gilette blades that I love.

The " ..., pausing only to shave." interjection was pretty funny.

Friday, April 23, 2004

What never seizes to amaze me is that people will bitch and moan incessently over the prize of a 299$ software package that will last them 3 years, even to the point where they think stealing is justified, but fork over 50-100$ EACH MONTH on cellphone bills alone without hesitation, and think nothing of spending 150$ on a pair of shoes, 300$ on a handbag for crying out loud.
What is it with software that seems to bring out the Scrooge in mankind?

Just me (Sir to you)
Friday, April 23, 2004

In that particular case, it's a misunderstanding of the time value of money and an inability to quickly calculate the value of a series.

For example CityDesk is probably comparably priced with TypePad (a fine product for blogging) but TypePad charges monthly and CityDesk charges once-up-front.

Joel Spolsky
Fog Creek Software
Friday, April 23, 2004

Personally, any product for which I have to pay a monthly subscription fee repels me unless I really need it. Tivo is out of the question for that reason. I don't watch TV that much anyway even though I fork the $40 every month.  :)

I was just ranting the other day about how much I pay to the phone company just to keep a dial tone alive at home!

That kinda ties in with having to get new customers all the time. Obviously subscription based businesses are more attractive to business owners since they get a pretty steady stream of cash coming in every month without getting a single new customer.

Maybe that should be part of the sales pitch for software... Pointing out the fact that you don't have to keep paying a monthly fee for the rest of your life, but perhaps offer that option if the customer wants it.  $50 up front, and then $5/month for the rest of your life. $8/month for high risk customers.  :)

I think people have less trouble paying small amounts of money spread over a long period of time than paying large amounts of money all at once even though in the end it sums up to the same amount. I certainly have this syndrome.

Friday, April 23, 2004


but if that is the case do you believe e.g. CityDesk would sell more at a 10$ a month licence price?

Just me (Sir to you)
Friday, April 23, 2004

At work, we use lots of software (EDA, simulators) that has expiring licenses. Each license costs anywhere between $5K - $20K. The licenses are good anywhere between 1 year and several years.

Perhaps if software licensed this way was instead licensed with a monthly subscription, that would bring down the instantaneous amount of money received from the customer per transaction which might make it easier to sell the software as well. This comment is in reference to one of Joel's articles about how much it _really_ costs to sell software.

I don't know if it makes any sense for CityDesk to adopt a monthly subscription based pricing method given its already low price...

Friday, April 23, 2004

Big up front charges can be a big hinderance to people. A pay as you go system can also be a great way to lock people into your company by making it hard to compare two products' pricing.

Where I work we sell insurance, the type of insurance we sell has a big upfront cost followed by 9 smaller payments ($3,000 up front $700 a month would be an example). Now that pattern repeats every year, with the big upfront and then 9 smaller payments. We will start this year setting up the upfront as payable ahead of time in three payments if they wish to renew, thus no big hit, just a relativly constant out of pocket cost. The small and big business owners we deal with are loving this. Because they don't need to save up a big sum of money. The benefit to us, when they go to get a quote from a competitor they face that big upfront bill, by staying with us the bill stays constant. (Government regs require the big upfront payment system.) Sure in simple math it is all a mind game but for a guy who has a slow month before his renewal he is greatful for that smaller payment that can be fit onto a credit card or borrowed from a friend rather then trying to find that big chunk of cash.

Friday, April 23, 2004

Maybe Joel wants to stick around for few years and sell the whole thing to Microsoft?

He has repeatedly mentioned Charles H. Ferguson and Front Page story.

It's my theory.

Jack in Yankee's stadium
Friday, April 23, 2004


You might want to mention that those EDA simulators are likely on a support contract that runs 15-20% of the up-front purchase price, per year.

That's not too much different from a subscription.  In fact, today, EDA customers prefer the "time-based licensing" (TBL) which is essentially renting the tools on a monthly basis.

An EDA developer
Monday, April 26, 2004

> Big up front charges can be a big
> hinderance to people.

This is what I HATE about web advertising.

Want to advertise?

Well, on most sites, it costs from $700 upwards. With $700 you buy a lot of clicks or impressions.

However, I can't risk the $700.

Why don't they offer an option of "start with $100, see if it works, and if it does, advertise more"?

That's why I love Google AdWords - you can start with $100, and if it doesn't pay off, quit, and if it does, scale up your expenses!

Monday, May 3, 2004

*  Recent Topics

*  Fog Creek Home