April 19, 2008 at 01:01ArsDigita: From Start-Up to Bust-Up (or Icarus Fall History) - Part I
- Transfer
Many, many years ago (in the web calculus — long before our era 2.0),
once upon a time there was one very interesting company. This was back in the days when
there were no Google campuses, the Apple apple still shimmered with many colors, and Sun and
SGI were still considered super-machines.
At that time, for many geeks and nerds, ArsDigita was about the same beacon
as Google is now. Many grew up on the book by Philip Greenspun (Philip Greenspun)
Philip and Alex 'to the Web Publishing Guide Review and short papers
SQL for Web Nerds and the Tcl for the Web Nerds . For many, the craze for the photo was
inspired by Phil .
ACS (ArsDigita Community System) was the first open implementation of a platform for
building sites, whether it be CRM, portal, CMS — in general, any application
focused on the user community. The choice of architecture for implementing
this platform may seem strange in part (especially for the Russian
reader) —ACS was written in Tcl , and used AOLServer as a
web server. However, if you read the articles above, you may
understand why.
For those who are interested, ACS still lives mostly as OpenACS , completely
migrated to PostgreSQL instead of Oracle (although Oracle is still
supported). All ArsDigita assets were bought in 2002 by Red Hat , but at the
moment, what was to become Red Hat Contacts and Content Management
System (CCM) no longer exists.
The history of this company, its formation and its death is dedicated to this translation (part
one of three).
***
If you happen to be near the Delaware court in the last few
days, then you may have heard about the lawsuit of venture capitalist investors with the
founders of ArsDigita. With this letter I want to explain the current
situation (from the point of view of one of the sides of the defense).
99% of the information below is not the least
process. The lawsuit filed by the company's shareholders is trying to determine the rights and
degree of control over the company’s management on the basis of the rules of law and signed
agreements. In other words, the conclusions about who knows better how to
manage the company and whether all decisions were made correctly are beyond the scope of the
current process.
In the courtyard of 1993. Around this time, the basics of that knowledge were laid that
later formed the basis of the ArsDigita Community System (a system of building online
communities). Most of the people who managed to earn money on software development
did this due to the fact that they were the first to find a solution to a particular urgent
problem. An attempt to create a company with the bare enthusiasm “our programmers are better than
the Microsoft and Oracle programmers” is doomed to failure — both companies
have in their arsenals quite high-class specialists capable of
burying any competition. But you can count on success if you
build it on the idea “we will solve this problem a few years earlier than
Microsoft and Oracle.”
A good example of how a small company can live and prosper, despite
the strong competition from large companies, is Adobe. Read the
biographies of its founders:
http://www.adobe.com/aboutadobe/pressroom/executivebios/johnwarnock.html and
http://www.adobe.com/aboutadobe/pressroom/executivebios/charlesgeschke.html . You
will see that these two guys who founded Adobe spent many years (at Xerox PARC,
Evans and Sutherland) over the very subject areas for which
Adobe now offers turnkey solutions.
The engineers and founders of Adobe are not smarter than their counterparts at Microsoft, they just
started thinking about graphics and printing earlier than Microsoft.
Fast forward to the year 1998. We have contracts with several large
companies (America On-Line, Hewlett-Packard, Levi Strauss, Oracle. We are
known as ideological leaders (see Macmillan's article in Database
Backed Web Sites) and as market leaders (our Open source software for building
learning communities.) In the words of Jack Welch, CEO of
GE, working for company # 1 or # 2 is easier and more fun. All things being equal, customers are
knocking on your door, not vice versa .
a year later, in 1999, customers simply bursting at our door. a good example
is the Siemens. They had to solve a very important problem for the which
ArsDigita Community System is ideal. By understanding how the task and
the solution fits together, the Boston Consulting Group brought them to us in our
old headquarters quartet on Franklin Street, 603 and two weeks later the contract was
signed.
We were growing up. But we were still a small company and tried to avoid direct
confrontation with competitors who were funded much better than
us. They built “closed” decisions, while we preached
openness. We wanted to replay them by creating a global application using
open standards and full access to our code, instead of defeating them
in the pages of Business 2.0 magazine. We laughed at most small
companies building closed solutions, we asked: “What is their marketing
strategy? Are we both Microsoft and Oracle, simply not market leaders and without money? And
does this help to search for and hire the best programmers? ”
By March 2000, we had grown to 80 people. I was still General, but I
began to worry that I could no longer answer who in the company
worked on each specific task. But we still earned a good
profit, the monthly revenue on an annualized basis was $ 20 million per
year. We paid almost $ 1 million in income tax for the year 1999. Not bad at all,
given that we started investing just $ 10,000.
We never tried to attract venture financing, but our turnover and
profit did it for us. Many of the best East Coast venture capital firms
wanted to invest in us, not least because of the super active
investment climate of the time, they wanted to invest in the company,
and how and where to make a profit can be found out later. ArsDigita
looked much better than most start-ups. Despite our growth of 1000% per
year, we had our own reserves in our accounts. Most of our revenue was
constant, and most of our customers were satisfied with us and loyal to us.
Typically, large companies do not like to buy enterprise-level software from small
developers. The risk of bankruptcy is too great. Using open source software
source code, in a sense, neutralizes this risk a little, but the desire and
tendency to order from IBM, Microsoft or Oracle is strong. We tried to
demonstrate our financial stability and strength. We bought a Ferrari,
which was provided to everyone who would bring 10 friends to the company. In reality, the
car cost us only $ 2,000 per month; the employee whom she was given
could use her only on condition that he continued to work with us. Well, the
cost of Ferrari is much lower than the cost of head-hunter services in
terms of ten employed employees. The presence of such a car in the
parking lot created a sense of extravagance — despite the fact that in the office we drove
modest enough, even economical life. And it could hardly have been otherwise if
we paid $ 50-85,000 base, plus an annual bonus based on the results of the
company and the employee himself, and paid the labor of professional
developers trained at MIT. We had a couple more tricks similar to
Ferrari. One is an oceanfront home in Cape Cod where the development team
could leave to work. The other is ArsDigita University, a free, one-year
professional training program for software developers. Although all this may
seem unnecessary, causing excesses, in fact, such an approach
to people constantly reminded us of humanity and care, and gave rise to
Journalists write about us (and they wrote), while standing no more than our
net profit of 1999 (i.e. practically nothing, provided that our revenue
continued to grow).
At the end of March 2000, we signed financing from two venture funds:
Greylock and General Atlantic. With the arrival of two checks from the funds, a
substantial amount of $ 38 million was formed in our bank . Having this stock, we decided to invest part
in the prospective development of our product and in marketing. Under the heading “
product development ”, the idea was that partial funding for development should be done by
teams involved in client projects. The idea was that if a
client’s request for one or another functional more or less coincided with what we wanted
add to ACS, we were ready to significantly reduce the price, thereby giving ourselves the
opportunity to more thoroughly study the problem and write code so that it
could be used on other projects, thereby improving ACS. Under the
marketing heading was meant the further expansion of our educational
activities (a kind of "educational marketing"). Well, of course, we needed
working capital. A company earning $ 20 million a year should have about
$ 10 million in the bank, in case one of the customers does not pay, an
economic downturn sets in , a large and important project will take longer than
planned, etc. Due to the fact that until that moment we were growing at 1000% per year,
we have never had more than a couple million in the accounts.
Under the terms of venture financing, VI (venture capital investors,
approx. Translator ) bought shares of the company, which gave them about 30% of the
vote. Under standard corporate governance conditions, minority
shareholders of this level should have received very little (or no)
strategic control over the company. In this regard, we also
signed an agreement between the shareholders, according to which the current shareholders (I and
Jin Choi) will vote for the board of directors of the company as follows:
Those. VI got 2 out of 7 seats on the board of directors. Current shareholders will have to
elect the rest. In addition to this, VIs received a veto on certain types of
significant transactions, e.g. the purchase of expensive equipment, the sale of a company, the
acquisition of another company. Finally, in the event of a sale of the company, they were
entitled to receive, on a priority basis, the amount of their investment of $ 38 million (which of
course means that if the sale was for less than $ 38 million, the remaining shareholders
would not receive anything). The conditions offered to us by three other VI firms were
equivalent. They wanted a “seat on the board,” but no one wanted
absolute control over the company. Their suggestions came down to
help the founders of ArsDigita do what they already did great.
In parallel with the negotiations on venture financing, we were actively searching for an
“external CEO”. Now, after I talked to many
successful people , I understand that this was a fundamentally bad idea. Even the
most capable person will need a couple of years to understand the culture of the
company, its people, the market, tasks and goals. The CEO involved can be
very successful in a stable 50-year-old company, in which its bureaucratic
system itself manages (see George W. Bush at the head of the Federal
Government (referring to the father of the current (spring 2008) US president,
George Bush seniorapprox. translator )). But in a young company there is no such internal
stability and strength.
However, we hired Allen Shaheen on the recommendation of Chip
Hazard, a Greylock employee who represented the firm on
our board of directors. Allen's previous place of work was Cambridge
Technology Partners (CTP), a company that provided IT services. In it,
he managed a large group of consultants in the foreign branch of the company. I knew
that Allen had no experience in the field of software development and that when working in CTP his
tasks were neither the development of a general strategy for the development of the company, nor the need
extraordinary, innovative approach to solving problems. In other words, he
always worked for someone in a much less competitive environment than the field of
software development. However, the other candidates we interviewed were either
very poorly trained (like one from Lotus), or they were too
aggressive and intrusive and my team of managers of that time could not
work with them. Allen seemed like a man who could work together with a number of
complex people scattered across different company offices, thanks to his experience
managing a multinational business in the provision of IT services. In general,
we agreed that I will continue to deal with engineering issues, issues
education and propaganda, while Allen will do the
rest.
And a few weeks after Allen’s passage, several people pointed out to me
that I had almost no power left: Allen and two representatives of the
VI could block any decision on the board of directors. Given that we did
not fill out the two invited positions on the council, there was a lot of
truth in these matters . 3 out of 5 = absolute power. And the point.
Part two has already been published .
once upon a time there was one very interesting company. This was back in the days when
there were no Google campuses, the Apple apple still shimmered with many colors, and Sun and
SGI were still considered super-machines.
At that time, for many geeks and nerds, ArsDigita was about the same beacon
as Google is now. Many grew up on the book by Philip Greenspun (Philip Greenspun)
Philip and Alex 'to the Web Publishing Guide Review and short papers
SQL for Web Nerds and the Tcl for the Web Nerds . For many, the craze for the photo was
inspired by Phil .
ACS (ArsDigita Community System) was the first open implementation of a platform for
building sites, whether it be CRM, portal, CMS — in general, any application
focused on the user community. The choice of architecture for implementing
this platform may seem strange in part (especially for the Russian
reader) —ACS was written in Tcl , and used AOLServer as a
web server. However, if you read the articles above, you may
understand why.
For those who are interested, ACS still lives mostly as OpenACS , completely
migrated to PostgreSQL instead of Oracle (although Oracle is still
supported). All ArsDigita assets were bought in 2002 by Red Hat , but at the
moment, what was to become Red Hat Contacts and Content Management
System (CCM) no longer exists.
The history of this company, its formation and its death is dedicated to this translation (part
one of three).
***
If you happen to be near the Delaware court in the last few
days, then you may have heard about the lawsuit of venture capitalist investors with the
founders of ArsDigita. With this letter I want to explain the current
situation (from the point of view of one of the sides of the defense).
99% of the information below is not the least
process. The lawsuit filed by the company's shareholders is trying to determine the rights and
degree of control over the company’s management on the basis of the rules of law and signed
agreements. In other words, the conclusions about who knows better how to
manage the company and whether all decisions were made correctly are beyond the scope of the
current process.
Start
In the courtyard of 1993. Around this time, the basics of that knowledge were laid that
later formed the basis of the ArsDigita Community System (a system of building online
communities). Most of the people who managed to earn money on software development
did this due to the fact that they were the first to find a solution to a particular urgent
problem. An attempt to create a company with the bare enthusiasm “our programmers are better than
the Microsoft and Oracle programmers” is doomed to failure — both companies
have in their arsenals quite high-class specialists capable of
burying any competition. But you can count on success if you
build it on the idea “we will solve this problem a few years earlier than
Microsoft and Oracle.”
A good example of how a small company can live and prosper, despite
the strong competition from large companies, is Adobe. Read the
biographies of its founders:
http://www.adobe.com/aboutadobe/pressroom/executivebios/johnwarnock.html and
http://www.adobe.com/aboutadobe/pressroom/executivebios/charlesgeschke.html . You
will see that these two guys who founded Adobe spent many years (at Xerox PARC,
Evans and Sutherland) over the very subject areas for which
Adobe now offers turnkey solutions.
The engineers and founders of Adobe are not smarter than their counterparts at Microsoft, they just
started thinking about graphics and printing earlier than Microsoft.
Fast forward to the year 1998. We have contracts with several large
companies (America On-Line, Hewlett-Packard, Levi Strauss, Oracle. We are
known as ideological leaders (see Macmillan's article in Database
Backed Web Sites) and as market leaders (our Open source software for building
learning communities.) In the words of Jack Welch, CEO of
GE, working for company # 1 or # 2 is easier and more fun. All things being equal, customers are
knocking on your door, not vice versa .
a year later, in 1999, customers simply bursting at our door. a good example
is the Siemens. They had to solve a very important problem for the which
ArsDigita Community System is ideal. By understanding how the task and
the solution fits together, the Boston Consulting Group brought them to us in our
old headquarters quartet on Franklin Street, 603 and two weeks later the contract was
signed.
We were growing up. But we were still a small company and tried to avoid direct
confrontation with competitors who were funded much better than
us. They built “closed” decisions, while we preached
openness. We wanted to replay them by creating a global application using
open standards and full access to our code, instead of defeating them
in the pages of Business 2.0 magazine. We laughed at most small
companies building closed solutions, we asked: “What is their marketing
strategy? Are we both Microsoft and Oracle, simply not market leaders and without money? And
does this help to search for and hire the best programmers? ”
By March 2000, we had grown to 80 people. I was still General, but I
began to worry that I could no longer answer who in the company
worked on each specific task. But we still earned a good
profit, the monthly revenue on an annualized basis was $ 20 million per
year. We paid almost $ 1 million in income tax for the year 1999. Not bad at all,
given that we started investing just $ 10,000.
We never tried to attract venture financing, but our turnover and
profit did it for us. Many of the best East Coast venture capital firms
wanted to invest in us, not least because of the super active
investment climate of the time, they wanted to invest in the company,
and how and where to make a profit can be found out later. ArsDigita
looked much better than most start-ups. Despite our growth of 1000% per
year, we had our own reserves in our accounts. Most of our revenue was
constant, and most of our customers were satisfied with us and loyal to us.
Typically, large companies do not like to buy enterprise-level software from small
developers. The risk of bankruptcy is too great. Using open source software
source code, in a sense, neutralizes this risk a little, but the desire and
tendency to order from IBM, Microsoft or Oracle is strong. We tried to
demonstrate our financial stability and strength. We bought a Ferrari,
which was provided to everyone who would bring 10 friends to the company. In reality, the
car cost us only $ 2,000 per month; the employee whom she was given
could use her only on condition that he continued to work with us. Well, the
cost of Ferrari is much lower than the cost of head-hunter services in
terms of ten employed employees. The presence of such a car in the
parking lot created a sense of extravagance — despite the fact that in the office we drove
modest enough, even economical life. And it could hardly have been otherwise if
we paid $ 50-85,000 base, plus an annual bonus based on the results of the
company and the employee himself, and paid the labor of professional
developers trained at MIT. We had a couple more tricks similar to
Ferrari. One is an oceanfront home in Cape Cod where the development team
could leave to work. The other is ArsDigita University, a free, one-year
professional training program for software developers. Although all this may
seem unnecessary, causing excesses, in fact, such an approach
to people constantly reminded us of humanity and care, and gave rise to
Journalists write about us (and they wrote), while standing no more than our
net profit of 1999 (i.e. practically nothing, provided that our revenue
continued to grow).
At the end of March 2000, we signed financing from two venture funds:
Greylock and General Atlantic. With the arrival of two checks from the funds, a
substantial amount of $ 38 million was formed in our bank . Having this stock, we decided to invest part
in the prospective development of our product and in marketing. Under the heading “
product development ”, the idea was that partial funding for development should be done by
teams involved in client projects. The idea was that if a
client’s request for one or another functional more or less coincided with what we wanted
add to ACS, we were ready to significantly reduce the price, thereby giving ourselves the
opportunity to more thoroughly study the problem and write code so that it
could be used on other projects, thereby improving ACS. Under the
marketing heading was meant the further expansion of our educational
activities (a kind of "educational marketing"). Well, of course, we needed
working capital. A company earning $ 20 million a year should have about
$ 10 million in the bank, in case one of the customers does not pay, an
economic downturn sets in , a large and important project will take longer than
planned, etc. Due to the fact that until that moment we were growing at 1000% per year,
we have never had more than a couple million in the accounts.
Under the terms of venture financing, VI (venture capital investors,
approx. Translator ) bought shares of the company, which gave them about 30% of the
vote. Under standard corporate governance conditions, minority
shareholders of this level should have received very little (or no)
strategic control over the company. In this regard, we also
signed an agreement between the shareholders, according to which the current shareholders (I and
Jin Choi) will vote for the board of directors of the company as follows:
- 1 representative from Greylock
- 1 representative from General Atlantic
- 3 senior managers from ArsDigita, including CEO
- 2 invited
Those. VI got 2 out of 7 seats on the board of directors. Current shareholders will have to
elect the rest. In addition to this, VIs received a veto on certain types of
significant transactions, e.g. the purchase of expensive equipment, the sale of a company, the
acquisition of another company. Finally, in the event of a sale of the company, they were
entitled to receive, on a priority basis, the amount of their investment of $ 38 million (which of
course means that if the sale was for less than $ 38 million, the remaining shareholders
would not receive anything). The conditions offered to us by three other VI firms were
equivalent. They wanted a “seat on the board,” but no one wanted
absolute control over the company. Their suggestions came down to
help the founders of ArsDigita do what they already did great.
In parallel with the negotiations on venture financing, we were actively searching for an
“external CEO”. Now, after I talked to many
successful people , I understand that this was a fundamentally bad idea. Even the
most capable person will need a couple of years to understand the culture of the
company, its people, the market, tasks and goals. The CEO involved can be
very successful in a stable 50-year-old company, in which its bureaucratic
system itself manages (see George W. Bush at the head of the Federal
Government (referring to the father of the current (spring 2008) US president,
George Bush seniorapprox. translator )). But in a young company there is no such internal
stability and strength.
However, we hired Allen Shaheen on the recommendation of Chip
Hazard, a Greylock employee who represented the firm on
our board of directors. Allen's previous place of work was Cambridge
Technology Partners (CTP), a company that provided IT services. In it,
he managed a large group of consultants in the foreign branch of the company. I knew
that Allen had no experience in the field of software development and that when working in CTP his
tasks were neither the development of a general strategy for the development of the company, nor the need
extraordinary, innovative approach to solving problems. In other words, he
always worked for someone in a much less competitive environment than the field of
software development. However, the other candidates we interviewed were either
very poorly trained (like one from Lotus), or they were too
aggressive and intrusive and my team of managers of that time could not
work with them. Allen seemed like a man who could work together with a number of
complex people scattered across different company offices, thanks to his experience
managing a multinational business in the provision of IT services. In general,
we agreed that I will continue to deal with engineering issues, issues
education and propaganda, while Allen will do the
rest.
And a few weeks after Allen’s passage, several people pointed out to me
that I had almost no power left: Allen and two representatives of the
VI could block any decision on the board of directors. Given that we did
not fill out the two invited positions on the council, there was a lot of
truth in these matters . 3 out of 5 = absolute power. And the point.
Part two has already been published .
