Back in 2001, during my first year studying Electrical Engineering at University, I was a devout and passionate proponent of Open Source software. I was a purist, and did not want to use any proprietary software. (I’m still very fond of Open Source software, but I’m not a purist any more). My laptop’s operating system was Debian Linux and I absolutely loved the terminal (and still do!).

Unfortunately, the University faculty did not share my passion, or at least they took a more practical (non-purist) approach and kept sending course assignments in Microsoft Word format (.doc). The only way to open these files on Linux was by using OpenOffice.org. This violated my purist principles. “Why should I, an undergraduate student on a very limited budget, be forced to buy Microsoft Office (and Microsoft Windows to run it)? I decided to do something about it. This ‘something’ was the beginning of my own little business.

The Customer Problem

“How can I quickly convert a Word file to PDF?”

Microsoft Word files (.doc and later .docx) make use of a format that is proprietary, binary and extremely complex. It’s a format meant for editing — not reading. Viewing Word files without Microsoft Word (or Word Viewer) was and still is difficult. Even now I have yet to find software that is fully compatible.

The PDF format (PDF stands for Portable Document Format) is the exact opposite. It’s purely for viewing and reading — not editing. And, like the name says, it’s portable. This means that it doesn’t matter on which device/OS/monitor you view it, it will look the same. This is not as trivial as it may sound.

Converting a document, such as those created by Microsoft Word, to PDF (Adobe’s Portable Document Format) was a luxury back then, only available by buying expensive software. There was no “Save as PDF” in Microsoft Office back then.

The Solution

OpenOffice, however, had the capability to save any supported document as PDF. And, most importantly, OpenOffice was free. Since it could open Microsoft Word files, I thought it was only a matter of automating the process of having OpenOffice open a .doc file and subsequently saving it as a PDF, and build a website front. (I was certainly not going to be manually converting these files myself).

This is exactly what I did. I hosted the website and my patched together OpenOffice ‘server’ in my parents’ basement (I was living at my parents’ house back then). I registered a domain and doc2pdf.net was born.

Initial ‘Marketing’

The next step was easy. I sent an email to the head of the university’s department and said if faculty members wanted to distribute their assignments in PDF format, they could use my website. He promptly forwarded the link to his staff.
Soon after, the website started to register some traffic, although it was not more than a dozen a week.

The Website

To say the website’s frontpage (which was the only page) was Spartan would be an understatement. It was made with Microsoft Frontpage (!!!) which had output a bewildering web of HTML tables — nested in such a way that it was almost impossible to edit and maintain.

There was no database, no user management, no need for users to register and sign in and users did not have to submit their email address. No ads were embedded in the PDF. There were no parameters to tweak. It was as simple as it could get.

Upload the Word file, and click Convert. Done. Completely hazzle free. It “just worked” and later this turned out to be the key differentiator.

Business Model

Generating revenue from a free service

The website was a great tool, and the very few people that happened to stumble upon it back then were very grateful. Someone even sent me $20 via snail mail!
In 2003 Google launched its AdSense program and a revolution was born. Being a big fan of Google — I immediately registered for an account and added some ads to my website. It was easy, fun, and kind of exciting. I was ecstatic when I received my first click on an advertisement, earning me about 1 cent.

This was still a loss making business — but I didn’t care.

The website happily ran in my basement, chugging along on my out of date server in my basement (The ‘server’ was a 6 year old desktop PC). It only needed the occasional restart, and every now and then I had to delete the pile of temporary files OpenOffice created.
One day, my home broadband connection became very slow. My router’s light indicators were blinking fast — which meant there was a lot of network activity despite nobody was using the internet at home.

Scaling up

It was clear that doc2pdf.net had outgrown its modest network capacity, and I needed to host it somewhere else, with a bigger network pipe. I found a great hosting service, Linode, where you could buy a Virtualized Linux instance for a very low monthly fee ($9.99). It was entirely self-managed — exactly what I needed. I bought an instance with 64mb of RAM and migrated the website over to it. It was great (soon after this would be called “cloud hosting”)

Over time, traffic increased slowly but steadily, reaching a few thousand visits a day. It was easy to upgrade the server instance to accommodate the increasing traffic.

Discovered by Google Search

One dark winter day, many years later, when I was casually reviewing the website traffic metrics on Google Analytics, I noticed a huge increase. Visits per day had multiplied. So did the number of clicks on ads.
After looking a little deeper, it seemed that Google’s search algorithm had suddenly ‘discovered’ the website. When searching for “Convert Word to PDF” or something similar, doc2pdf.net was the top result. Over the next few weeks daily visits reached over 20,000.

To Be Continued ???

Want to hear the rest of the story? In the spirit of gathering early feedback and fail-fast principles, I will publish the second part to this story only if someone is interested in reading it. There’s a lot to cover, such as

• Incorporating a company
• Surviving an epic financial crisis
• Expanding and differentiating
• Fighting copycats
• Dealing with increased competition
• Hiring contractors
• The importance of experimentation
• Selling the company
• Lessons learned

Let me know what you think. I’m on Twitter.