Hello, my name is Bruno Lowagie, I’m the original founder of the iText Group and co-Founder of ThinkFree, a joint-venture between iText and the South-Korean productivity software vendor Hancom. Although this is a Master Class called “Investor Readiness”, I’m going to introduce you to an alternative approach called “Bootstrapping.” More specifically, I’ll give you an example of a successful tech company that was built without any money lent from a bank, without any investments by a business angel, or without any funding from a VC.
Incidentally, I am the original developer of the product, one of the original founders, and one of the original owners.
For those who are curious about the technology that is being developed by the iText companies, allow me to present a single slide that summarizes what our product is about.
iText is a PDF engine that is built into web and other applications and that allows those applications to generate, manipulate and process PDF documents. To the left, you see sources of data such as XML files, XHTML, data from web forms and databases, in-between you see an application on a desk top, on the web, in the cloud, on a mobile device. This application has “iText inside”, our engine that runs on Android, .NET, Google App Engine and Java. This engine produces output such as PDF/A (for archiving), PDF/UA (for accessibility), digitally signed PDFs, and much more…
In this presentation, I am not going to talk about iText as a product. Instead, I’m going to introduce you to iText as a business.
But before I do so, allow me to introduce two other companies you might have heard about.
One is called Hare Corp., it’s a Belgian Social Media startup started by Jack Haas, a high-school dropout who is often coined as the Belgian Steve Jobs. The company was selected by the Belgian incubator BelgoLab and received €50K in seed money.
The other company is called Tortoise LLC. It’s a technology startup founded by two engineers who found a niche in the B2B market. The company is fully funded by the founders who invested their own savings.
Back to iText: iText was first released on February 14 of the year 2000. It was a personal hobby project of mine and it was published as an open source library using a “free as in free beer” type of license. In the Summer of that same year, a developer in Portugal joined the project and soon the library got traction. If we browse the web, we discover that the library was already used by IBM in 2002.
At that time, there was a lot of Fear, Uncertainty and Doubt about open source. Big corporations such as Microsoft and Oracle claimed that open source wasn’t documented (as opposed to their products), that the IP of open source wasn’t clear (as opposed to their legally safe products), and most importantly that open source wasn’t backed by a company (as opposed to their products for which they ensured continuity and long-term support).
When I looked at my project in 2004, I knew that I would have to solve these issues one by one if I wanted my project to be successful.
I made “Documentation” my first challenge.
At that time, I had to support a family of four (myself, my wife and two children) with the salary of a single day job, so instead of going on vacation, I decided to use my spare time to start writing a free online tutorial.
I discovered a thing called Google AdSense, and I decided to allow Google to put ads in a column next to my tutorial. As you can see, this was my first source of revenue. In 2005, I didn’t have a company yet, but I made more than $1,000 a month thanks to Google.
If you look closely, you can see that this is not a sustainable business model. In 2006, I had more content, more page views, more clicks, but less money: other people had also discovered that you could make money with ads. Even worse: if you look at the ads, you can see that most of the companies advertising on my site were direct competitors of iText. Isn’t that ironic? In a way, my competitors gave me the seed money that was needed to bootstrap.
The free online tutorial was a success and soon I received offers to write a book. First from O’Reilly, later from Manning Publications.
In 2005, O’Reilly was the bigger publisher: they released about 500 books a year. Manning was the new kid on the block: they released about 50 books a year.
I decided to go for Manning because I wanted my book to be a success. This was what I thought: if my book fails at O’Reilly, that failure represents only about 0.2% of the titles they produced that year. If my book fails at Manning, it would represent 2% of their production. It was in their interest to make the book a success, so I went for Manning even if that meant more work for me.
It turned out being the right choice. As you can see, one doesn’t get rich from writing books, but this was my second source of seed capital. More importantly: the book was great marketing for the product. James Gosling, the inventor of the Java Programming language blogged about it, and Manning publications asked me to write a second edition. Today, Manning is one of the top publishers in the IT world.
This is how I solved problem #1 “Documentation” for iText. The problem is also solved in general: there are currently more books being written about open source products then there are book being written about closed source applications.
On to problem #2: let’s fix the legal issues.
In 2006, I was contacted by The Eclipse Foundation, which is a foundation that is heavily funded by IBM. They asked me if I could change the open source license I used for iText to the Eclipse Public License. This would allow IBM to do whatever they wanted with my code. At some point, the pressure to do so got really high, but I was as stubborn as a mule: I continued to refuse.
“Then we have to do it the hard way,” was the message from the IBM lawyers. So I asked: “What is the hard way?”
As it turned out, the hard way consisted of a thorough IP review funded by companies in the Eclipse Foundation. These companies paid the equivalent of what I costed to my employer for a full year. I didn’t get any of that money directly, I just continued doing my day job and I was paid my regular salary, but whenever IBM’s legal team in Canada found an IP-related issue, I had to drop everything and solve the issue.
The end result was a full list of all the people who had ever contributed code to my open source project. It was one of the most boring jobs I’ve ever done because I had to go back in time up to five years in time to find the information that was needed. Nowadays, we have a very disciplined book keeping for the IP. To the left, you see the full list. In the middle, you can see who worked on a specific source file. To the right, you can see a page listed all the source files a specific developer has worked on.
A word of advice: if you’re just starting a new project: don’t forget to document, document, document! You don’t want to do an IP review on code that goes back five years.
This is another thing we should have done from the start: registering our TM.
We registered iText as a Trademark with the USPTO in 2010, and they accepted February 14, 2000 as the date of “first use of business.”
As for patents: filing, getting and keeping a patents can be very time-consuming and expensive.
I published my code as open source, because I didn’t believe I would ever have the time and the money to write patents. By bringing the code and the concepts out in the open, I prevented others to patent our technology.
Today, I regret that I didn’t start writing patents earlier: when I read my books, I’ve already discovered several ideas that could have been filed as a patent. Unfortunately, it’s too late to do that now. Or is it?
In 2013, I started writing some patents that were accepted in Belgium. Having Belgian patents is important for many reasons: a company that has a similar patent in other countries will want to be able to sell its product in Belgium too. One way to achieve this, would be to challenge our Belgian patent in court. However, it may be less expensive to make a deal instead. That can save such a company a lot of money in legal fees.
As for us, Belgians, having a Belgian patent can result in a significant tax exemption.
After the IP review, the ownership of the code was finally 100% clear. It was time to start a company and to transfer all the IP into that company.
Most people would have started by founding a company, just like most people would have started by writing patents or by registering a Trademark, but that’s something you do when you have sufficient means. If you’re bootstrapping a company, you often have to postpone essential steps in the process of building a business. This can seriously compromise your growth and even jeopardize your chances at survival. It takes quite some heart to be a bootstrapper.
The year we founded our company, we tried making money with our technology, but we didn’t have any success.
I’m sorry, but I have to interrupt the iText story for some breaking news about Hare Corp.
…
There is no news about Tortoise LLC.
Back to iText: as I explained, we didn’t have any significant success in business in 2008. We were a Belgian company, looking at the European market and although we detected many iText users (even the Belgian Government), nobody was interested in becoming a customer.
This could mean two things:
Either or business model was wrong, or
The market was wrong.
I’m an engineer. That means that one can save plenty of time just by assuming that I’m never wrong. In short, there was essentially only one option: the market was wrong.
To prove that the market in Europe was wrong, we started a company in California: iText Software Corporation. The company was founded for me by a friend, Andrew Binstock, and we worked on two deals: one on the East Coast (The Hartford) and one on the West Coast (Charles Schwab). It took quite some time to make these deals, because we didn’t really know what we were selling and the customer didn’t really know what they were buying. However, the end result of the negotiations was our first End User License Agreement.
For a long time, I truly believed that the software library known as iText was our product. Only when the first deal was closed, I realized that I was wrong. The EULA, not the software, was our first product.
With two deals closed, we signed up two sales people on commission. If they didn’t sell anything, they wouldn’t get paid. But if they succeeded in closing a deal, they’d get a 30% commission.
We gave them a list of companies that had shown interest in the past and they worked on that list. At the end of 2009, this resulted in a total of $300K in revenue and that was that: all the companies on the list had been contacted. The low hanging fruit was picked and we had no more revenue in Q4.
That’s when we decided to change the “free as in free beer” license into a more strict license. The technology would still be offered as open source, but companies using the software in a commercial context would need a commercial license. The license change resulted in a slow increase of sales starting with 10Q1, but as long as the “old iText” had more or less the same functionality as the “new iText”, there was very little reason for customers to buy a license for the “new iText.”
The only way to convert users to switch from using the old “free as free beer” version to the new “pay for commercial use” version would be to add new functionality. Hence I quit my job in September 2010 and I started working full time on iText. You can see the result: the 2010 revenue was about $700K.
In 2011, we reached $1M at the end of the second quarter.
Isn’t that exceptional: we suddenly had a company making a million dollar and we didn’t even have any employees on payroll. We only had a self-employed part-time managing director in the US (Andrew Binstock), a self-employed full-time developer in Belgium (me) and two sales people on commission in California. It may surprise you, but I made my first trip to the US in 2010, a year and a half after iText Software Corp. was founded.
The fact that money was pouring in, made us confident that we could give Europe another try. In August 2011, we founded iText Software BVBA and we hired our first employee on payroll. We also cut the commission for the sales people on commission in half. This resulted in a very strange effect: they started selling twice as much as before.
Hiring people was very scary for me at first, but in 2011 and 2012, we hired more and more people in Europe. We noticed some huge differences between employees on payroll and people working on commission or as consultant paid by the hour. Thanks to the dedicated sales people, we finally started to gain more insights about our customers. That was new for me: whenever I used to ask the sales people on commission for information about our customers, they were reluctant to share that information as “they weren’t paid to do that.”
In August 2013, we opened an office in Cambridge, Massachusetts. iText Software Corp. is still incorporated in California, but the Boston area was selected for two simple reasons:
Our top 2 verticals are #1 government and #2 financial. Government in the US is mainly based in Washington DC. The headquarters of the major banks and insurance companies are mainly in New York. It made sense to have our office in the same time zone as those customers.
The time difference of 9 hours between Europe and California was killing us. It’s much easier to run a company remotely when there’s only a 6 hour time difference.
We didn’t “fire” the sales people on commission, but we changed them into being a reseller. This has a significant impact on the revenue: before this change, their commission was counted as revenue. Our gross margin was calculated as the revenue minus the commissions. After the change, the commission was changed into a reduction and that reduction doesn’t count as revenue. Today our total revenue is (almost) equal to our gross margin.
In any case, the company was growing and so was the revenue. Since 12Q2, we’ve never had a quarter with less than $1M in revenue.
Once again, I have to interrupt the iText story for some big news.
…
There is no news about Tortoise LLC.
This was the situation at the start of 2013: we had a staff of 3 self-employed people, 5 employees, and 1 reseller in a company making more than $1M a quarter.
How did that happen?
And more importantly: what to do next?
In 2013, we hired an M&A consultant and we asked him: what are our options?
After a project that took about 6 months, we distinguished 4 options:
Option 0: GET OUT! We worked with an M&A company for a couple of months to probe if we could sell the company. We soon found out that we weren’t ready to do this. We had been professionalizing, but too many things weren’t in place yet. We were a company all right, but we weren’t a real business yet. We wouldn’t have survived a due diligence process.
Option 1: MILK IT! We were profitable from day 1, so we could continue working the way we had always been working, without investing in new employees, without investing (too much) in new development. We’d have a nice “mom and dad business” and we’d be able to extract some profit out of the company every year. Why be ambitious when you can be comfortable? However, this wasn’t in our nature.
Option 2: PEDAL TO THE METAL! We could grow the business with the help of a VC. We talked to different investors, but it didn’t feel right. I like riding my bicycle, but what the investors proposed felt like: “We’re going to give you an injection and next year you’re going to win the Tour de France!” For instance: we made a spread sheet that showed us that we’d make $50M a year after about 2 year. I didn’t believe that spreadsheet and I wasn’t really interested at creating a $50M business at all cost, so we went for option 3.
Option 3: DRIVE IT! During our talks with VCs, we asked them: “what would you do for us if we accepted your money?” We added their responses to a checklist and turned it into a priority list. We installed a board with people of our choice and we started professionalizing: we changed the company structure (1T3XT BVBA became iText Group NV; iText Group NV bought iText Software Corp and iText Software BVBA), we switched accountants (BDO as the global accountant firm instead of different accountants in different countries), we appointed an auditor (Deloitte), we started working with an HR specialist (Hudson; all employees need to do an assessment),…
The main goal of the “Drive it” strategy was actually very simple: “How do we convince Bruno that it is possible to turn his current business into a $50M business, and why does it make sense to do so?”
This was the situation before. My wife and I each owned 50% of 1T3XT BVBA, which we saw as the IP company. We also owned 100% of WIL-LOW (aka Willaert - Lowagie). WIL-LOW owned 100% of ISC and 99% of ISB (we personally owned 0.5% each). This was the first slide in our fact book during the short M&A project. I can assure you: many companies browsing the fact book, didn’t get past that first slide: it looked simple enough to us, but it was too chaotic to many.
WIL-LOW sold ISB and ISC to 1T3XT. We then liquidated WIL-LOW (with the 10% liquidation bonus), which means that we’ve cashed personally (our first “exit”).
1T3XT was changed into iText Group, the BVBA became an NV. Earlier this year, a new subsidiary was founded in Asia: ISA.
The situation became very straight-forward:
There are two share holders who each own 50% of iText Group NV.
iText Group NV has 3 subsidiaries: ISA, ISB and ISC.
Each subsidiary is responsible for a specific geographical territory.
Having cashed personally also gave us the level of comfort to make some important decisions for the companies.
We could afford more risk and there was no pressure whatsoever to get cash out of the companies.
Oops, I have to interrupt once more for some bad news.
…
There’s still no news about Tortoise LLC.
But enough about Hare Corp. Let’s take a look at the financials at iText Group. As you can see, we stopped thinking in dollar, but we started counting in euro.
2014 was a transition year. At first sight, you might think that the growth slowed down significantly. That’s not entirely true. We had less revenue because we switched from paying our independent sales people a commission. Instead we gave them a reduction. The commission used to be counted as revenue. The reduction isn’t counted as revenue. Furthermore, we changed the revenue recognition, so that we comply with Belgian GAAP.
We also fired and hired many people, which resulted in some serious dips in sales. We were executing a business plan that involved change. We knew that change would cost us on the short term, but that is was needed for the long term.
Since January 2014, we have a two-day board meeting every three months. The first board meetings were a very painful experience for my wife and I, being the founders/owners: we were confronted will all the mistakes we made in the past and with all the essential steps we had skipped due to lack of time, lack of experience, lack of interest, and so on.
However: soon every new board meeting became more and more rewarding. It’s amazing, but as founders, we slowly start distinguishing the shape of a real company in 2015. I know this must sound strange, knowing that we founded our first company 7 years before, but that was how it felt.
This is our operational spreadsheet for 2015. It is the “tool” we use to follow up our progress in “real time”, month after month, quarter after quarter. The blue line shows the billed revenue in 2014 (about €5M by the end of the year). The dashed gray line shows the target revenue. The orange line shows the billed revenue in 2015. We were aiming for €6.5M by the end of the year. We exceeded the target and ended at €6.7M. Note: some of this revenue will only be “recognized” in 2016, hence the 6.5M on the previous sheet.
Note that we also changed the bonus system in 2015: instead of working with an individual bonus for the sales people in the US (and only in the US), we replaced all bonus systems by a company-wide bonus in 2015 (for every employee, whether (s)he’s a sales person or not) based on the revenue of the group as a whole (all the invoices from all the subsidiaries). The bonus is determined on a quarterly basis. Coincidence or not, the third month of Q1 and the third month of Q2 were record-breaking months. When we didn’t reach the target in Q3, we introduced an extra bonus if the final target for the year would be reached. Sales broke every record in December.
In the meantime, things have gone from bad to worse at Hare Corp.
It should be obvious that Hare Corp. isn’t an existing company either. It could have existed though.
Hare Corp. was inspired by the stories I read in the CB Insights startup failure post-mortems. Some newspapers only bring you news about companies getting funded as if success is guaranteed. You may argue that reading startup failure post-mortems is borderline masochism, that it’s depressing and that it harms your optimism. I don’t agree. I look at this differently: a mistake made by someone else is a mistake I don’t have to make.
There is no Tortoise LLC either.
When I talked about Tortoise LLC, I was actually referring to iText, and up until recently, iText didn’t make the news papers very often.
Nevertheless: in October 2014, iText won the BelCham Entrepreneurship Award in the category “Most Promising Company of the Year.” A couple of weeks later, the company was recognized by Deloitte as the fastest growing Tech Company in Belgium. In the EMEA Fast500, iText obtained the 28th place. In 2015, iText won:
a Gold Hermes Creative Award for the project “iText Gets a New Look”,
a Silver Best in Biz International award in the category “Product of the Year (small and medium business)”,
place #1,666 in the American Inc. 5000,
a Bronze American Business Award aka a Bronze Stevie in the category “Most Innovative Company of the Year”
place #2 in the Belgian Fast50 (#143 in the Fast500)
a nomination for the European Business Awards (Belgian Champion in the category “Entrepreneur of the Year”)
Because of all these awards, we are getting phone calls, e-mails, even paper letters from all kinds of investors who “want to get to know us” and who “are very interested in investing in companies like ours.” The year before we won the Fast50, we obtained third place in Belgium. The Belgian #1 of the Fast50 in 2013, Cmosis, and the #2, LayerWise, sold (part of) their shares to private equity. Needless to say that we also received many offers. We could easily have sold our business too, but we didn’t.
Some people probably thought we were stupid, but… we had other plans in mind (world domination, muhahaha).
As you may remember, the goal of the Drive It strategy was “To convince Bruno of the feasibility of creating a $50M business (or more).”
I didn’t have the time to watch the HBO series Game of Thrones, but during my many travels, I read the five “A Song of Ice and Fire” books.
As things were going well at iText, there was a risk that we’d get to comfortable. It felt as sitting on a throne.
But if Game of Thrones learns you anything, it’s that there are always contenders who want to push you aside and claim your throne as theirs.
In 2015, I started developing 10 (eventually 11) ideas for the future. Each idea was given a name of one of the characters in the GoT series.
These ideas were presented internally and everyone was very excited about those ideas. (As a side-effect, some people started binge-watching GoT, but that only increased their enthusiasm). Everyone at the company worked on these ideas (is there a market? Do we have customers who are looking for this solution today? Which resources do we need to build this idea?) The ideas were presented to the board and the board made an official priority list. Arya and Tyrion were selected as the most important ideas for 2016.
This was how it was presented to the complete staff. At the board level however, we again defined four strategic options.
Option 0: BUSINESS AS USUAL. We could continue and try to grow the business with 30% every year without implementing any of the GoT ideas. That was the easy option: do (almost) nothing and see where we get. Whether we’d succeed wouldn’t depend as much on the staff, but on external factors. In a world that doesn’t change, we’d have a chance at success. However, some of our competitors are very eager to conquer our throne and to take our place in the PDF space. The risk that we’d grow too slow and lose the race against our competitors was high. Moreover: who wants a “business as usual” business? Where’s the challenge?
Option 1: MAKE! We could try to implement all the GoT ideas ourselves. We’d need to invest in more development, more employees, maybe sublicense some existing technology, and we’d also need to prioritize. We had a nice war chest, so financial resources weren’t really an issue. However, due to the lack of human resources (it’s not easy to build a team), it was inconceivable that we’d be able to implement all the GoT ideas at once. We’d have to choose wisely, if we wanted a chance at success.
Option 2: BUY! We could try to acquire companies that own technologies that are in line with the GoT ideas. This would allow us to move faster than in the MAKE! scenario. However, we might not have sufficient financial resources to acquire the companies we needed (or we might want to spread the financial risk), so we were thinking of teaming up with a VC. Actually, one specific VC in the Bay Area was eager to invest, so we almost chose this option. But then opportunity knocked.
Option 3: GET IN! This is also an exit scenario similar to the GET OUT! option in 2013. But GET OUT! had a negative ring to it. It sounded as if I wanted to abandon iText, as if I wanted to quit. Quite the opposite was true: I wanted to find an interesting partner for iText. Wouldn’t it be great if we found a company that would acquire us and:
Provide developers to help us execute on the GoT ideas,
Provide technology we can use to get a jump start with respect to some of our ideas,
Invest in the development of our ideas.
We found such a partner in Hancom.
Let’s take a look at a real newspaper article for a change (De Tijd, December 23, 2015).
Hancom is a South-Korean productivity software company. It is big in Asia where it competes with MS Office products. It was a customer of iText.
We met Hancom for the first time in July 2015 in Seattle. We discovered that our product offering was very complementary (which was nice), but also that there was a huge overlap when looking at our future plans (many of the GoT ideas were also under development at Hancom).
Early September, Hancom visited us in Ghent with a draft for a Letter of Intent (which we signed).
Early October, we visited Hancom in South-Korea for in-depth discussions on how we would work together. We left with a term sheet (which we signed).
The due diligence process started soon after (October - November).
On November 13, Hancom founded a new company in Belgium, named ThinkFree NV. This company received a license to sell all Hancom products in Europe and the US.
On December 7, my wife and I signed a Stock Purchase Agreement and I signed a Shareholders Agreement.
The license agreement described an exit for iText Group NV: all iText companies are to be sold to ThinkFree NV.
The Shareholders Agreement described a joint-venture: ThinkFree would become a company owned by Hancom and myself.
Closing of this deal took place on December 29.
People with experience in M&A tell us that this deal was closed extremely fast. Usually, it takes much longer to come to an agreement.
There are probably different reasons why we were able to make this happen in less than half a year, but I’m convinced that preparation was key.
For my wife and I, this process started in 2013. It was the goal we had been working towards all along.
I was asked by Hancom to be the CEO of ThinkFree and iText Group, but I refused.
I accepted to be the CEO for a maximum of 3 months, which should be sufficient time to find someone to replace me.
This doesn’t mean I am looking for a way out. As I said: this deal wasn’t a “Get out” kind of exit. The goal was to “Get in!”
I proposed to be the Chief Strategy Officer and that proposal was accepted.
As a shareholder in ThinkFree, I am also a member of the board (the board of ThinkFree as well as the board of iText Group).
In the past, I was able to combine a seat in the board with my role as CEO because we were a relatively small company.
With ThinkFree, it’s the goal to grow to a 50M euro company, and I think it’s better to separate the board and the CEO.
The CEO will be responsible for the short and medium term. The board is responsible for the medium term and the long term.
I want to be the person in-between. I want to be a major influence on the board regarding the long term goals, and I want to assist the CEO to make sure he has the freedom to make decisions for the short and medium term. I will guard the roadmap to make sure that these short term decisions don’t lead us astray.
Another reason why I refused the role of CEO is simple: when we grow, we need a manager. I’ve worn many hats, but I’m definitely not a manager. I’m an entrepreneur and a leader. The role of CSO should allow me to try untrodden paths (as a reconnaissance officer) and clear the way to further growth for the company to grow (as a field marshal). I need a strong CEO next to me (an XO) to execute the strategy, to organize the company in such a way that the short and medium range targets are hit.
To be continued?
Morale of this talk.
…
I hope you’ve enjoyed listening the iText story, as much as I enjoyed bragging about our achievements ;-)
Please understand that I’m not here to tell you how to run your business. That would be wrong: every business is different.
Moreover: I realize that I’ve made several mistakes that have costed me valuable time and money. I don’t regret those mistakes: they were part of the learning process. If you’re to be an entrepreneur, you’ll have to go through your own learning process and you’ll discover that school doesn’t prepare you for every obstacle you’ll encounter. I can assure you that it’s a very rewarding learning process, hard at times, but rewarding.
I conclude this session with a quote from a book that I read when I was 18 years old.
I wrote this quote down in my diary, and I’ve been implementing that quote ever since:
…