Before Facebook There Was Grono.net: How our CTO Built Poland's First Social Network

our press

Before Facebook, in Poland there was Grono.net - a social network with 2.5 million users and a $2MM investment from Intel Capital. In an interview with Justjoin.IT founder Piotr Nowosielski, our CTO and co-founder Wojtek Sobczuk tells all on what it took to build Poland’s first social network and how that led him to create VentureDevs.

Piotr Nowosielski: It is no secret that you are the founder of Grono.net — one of the biggest Polish social media sites to this day. However, it is only one of several projects that you have developed so far (don’t worry, we will have the opportunity to talk about Wojtek’s other projects later!). I wonder how it all started and what your life looked like before you came up with Grono? What did you study, what did your first steps in the programming world look like?

Wojciech Sobczuk: Everything started with the idea of writing computer games over 20 years ago; I bought a book called ‘Language C in 21 Days’, and after a week, I learned enough to write some simple first-person shooter and Tetris clone.Then came the years of spending countless hours at the Lublin University hall where the only Internet connected computers were available. I spent my time on IRC or various interesting online activities related to security, and it’s there where I met the people who largely created the Polish Internet in its present shape — e.g. Carp. He built the Atman network and is now head of security at Orange Poland. At that point, I had been dealing with server administration, writing low-level code in the Linux / FreeBSD kernel, and with security — including writing and protecting from exploits and trojan horses.One of my best memories with my comrades from UMCS was when we executed a successful penetration test of the main GSM operators in Poland. We managed to hack their servers using a specially crafted text message and then gain administrator access to the whole machine. Thankfully, since that time security has improved… but only somewhat, honestly.Then the time of the WWW websites came, and I just switched to it completely. I wrote something similar to Ceneo.pl (Polish price comparison website) before anyone came up with the idea of creating something like that in Poland, but I gave it up because the idea of creating Grono.net appeared at the same moment. In any case, when Grono started working, every other idea was thrown into the trash — even my IT studies at the University of Warsaw. To be honest, I hope that one day the success of VentureDevs will be so great that the Grono project will look like a small part of my career.

Piotr Nowosielski: It started innocently, with meet ups of a couple of programmers over beer. Only a few years later, you became the most popular Polish website with nearly 30 million page views per month. How did the idea of the Grono come up and what services did you implement in the beginning? Facebook was created in parallel, and the trend seemed to be Myspace at that point; I don’t really associate the two, though.

Wojciech: An acquaintance of my friend Bronek got the idea for this page from another friend. The acquaintance told us that in Hungary an experiment was carried out to prove that there are six connections between every human in the world. We didn’t think long, and we just came up with the name Grono and started our work on the website. Bronek is a front-end developer and graphic designer, so he dealt with the graphics and front-end development, and I wrote all the mechanics in Java.

Piotr: How long did you work on the first version of Grono.net? It was probably an intense time. Did you do it after hours, or rather did you all focus on it full-time, shut yourselves somewhere together to release it as soon as possible? I never dug up information on how old you were at the time?

Wojciech: I was 22 years old, Bronek was older. We both started after hours, and it did not take long, maybe 2–3 months because at the beginning the functionality was very simple. At some point I quit the job that I had at the time to focus 100% on Grono.

Piotr: Start at February 2004 and the first serious problem — launching. How fast did the user base go from zero to one thousand? I couldn’t believe that all you had to do was to invite your friends — who so far were only using emails to communicate with their family and colleagues. Back then, those who were more modern and progressive, were already using the first Polish instant messenger — Gadu Gadu, but the idea of using some kind of social network website was definitely still something “exotic”.

Wojciech: Contrary to appearances it was very simple. At the beginning, we invited our closest friends, asking them not to invite anyone because we are testing the service and correcting mistakes. When we found out that the underlying functionality worked well enough, we sent an email to everyone asking them to invite their friends, and then a continuous avalanche of users began. The service was new, also everyone was curious how it worked. At the same time, it had that note of elitism because people needed to be invited, and it is known that if something isn’t for everyone then everyone wants to have it.

Piotr: This approach of being an invitation-only service turned out to hit the bullseye and basically became known as a great growth hack. Had you thought the service would be closed to a select community from the very beginning? Did you make Grono more open to anyone who wanted to join at the beginning and then make it more selective after a certain time after you had gained some users?

Wojciech: The service was invitational from the beginning, but believe me there were no great discussions and philosophies — we just stated that we’ll check the principle of 6 handshakes from the Hungarian experiment. In order to do this we needed to add the invitation to eliminate any accidental people who would not have friends inside. Then it turned out that this was not the best assumption in the context of the popularization of the site because after all, even people without an invitation could subscribe and already find friends inside.

Piotr: By the way, I have another, somewhat peculiar question. Which do you think is easier — to turn an open service into a selective one, or the opposite, open up a service to the masses that was previously available only by invitation?

Wojciech: One of the fund managers at Intel Capital once said at a Grono investment meeting, “He who is the most open wins.” I think that aside from a few exceptions, this rule makes sense. Also, I think that it is better to open something up as time goes on rather than to shut it down.

Piotr: So were there any other growth drivers, besides inviting friends?

Wojciech: No, only friends who invite friends.

Piotr: The first year after your launch brought you some popularity, but it was far to go. We are talking about 30,000 users here. After just over a year, we could talk about 500,000, and at the peak — about 2 million. What problems did you have in the day-to-day?

Wojciech: The main problem was of course performance and for a very long time users were experiencing this problem. There were also a number of problems not visible to the users — mainly related to knowledge of how to lead and develop a startup — in Poland it was new and few had experience in this. Things like how to get financing, how to build a sales department, how to manage a growing company — these are the questions we often had to answer ourselves.

Piotr: How did you go about getting funding in those days? There were obviously users; it was the most popular portal in Poland. You were in the position of strength. Tell me more about the process of investing in startups at the time. Were investors more reserved in making these investments? Or was everything faster and less formal than it is now?

Wojciech: Investors will always show great interest in you if you have success and do not need them. And if you do not have the success and you need them — then you have to run after them and ask for funding because you are competing with thousands of other founders. Because of our very fast growth in the beginning, we were able to choose from our offers and set our conditions.

Piotr: How was the IT department organized at Grono? How many programmers did you have once you reached 2 million users? How many people have been involved in the development of new functionalities, and how many of you were needed to keep it all in check?

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat. Aenean faucibus nibh et justo cursus id rutrum lorem imperdiet. Nunc ut sem vitae risus tristique posuere.Wojciech: At our largest, we were over 80 people, most of them were programmers plus a team of server and database administrators. We broke it into functional modules, some dealing with forums, others with profiles, and new modules. The team was amazing; they delivered basically our entire vision from product conversations in record time. I very much regret that we were not able to lead the team to long-term success so that they could continue seeing the results of their work today.

Piotr: In retrospect, what other decisions would you have made when it comes to the development of key Grono.net functionalities? Do you think that any of these functionalities could make this website exist today?

Wojciech: It consists of many factors, but the main ones in my opinion are:

Technology — Python + Django enabled very fast builders.
Team Skills — We had managed to gather a lot of talented people because of the fact that we were one of the biggest and most famous Python sites.
Motivation — It's known that when something is built for millions of users, and friends enjoy it, people are more eager to do it.

Piotr: Let’s look at some of the technological developments at Grono.net. You have a dynamic experience of many different development languages, you know a dozen, and you learned a lot while building Grono. What technologies have you used over the last 10 years while you were working on Grono?

Wojciech: The first version I wrote in Java, but it quickly turned out that under heavy load the JVM (Java virtual machine) was a big problem. I was definitely a JVM tuning expert, but that did not help either. Around then, I was introduced to Python and Django and decided to rewrite the entire site for these technologies. It’s been Python ever since. When it came to front-end technologies, there was not much choice at the time — what we did was in jQuery.

Piotr: How quickly did you rewrite the product in Python?

Wojciech: I’ve always had a good sense for the effectiveness of development. For my version written in Java, I used the author’s library, which accelerated writing web apps in this language, so it was created very quickly. At the same time, it did not have the best performance because, in my opinion, performance is completely irrelevant until it becomes a problem for users. At the time, Grono was unusual for me, though, as the avalanche of users appeared immediately after launch, and with them the performance problems.When it turned out that the current Java architecture did not meet our performance expectations, I found that it would be faster to rewrite in Python + Django than to change the framework to any existing one in Java. We did this, if I remember correctly, in 2 months, with the first Python programmers who joined the team.Here I was reminded of the interesting performance situation. At some point we found that the service was running slow and eating at the CPU resources of the 50 servers we had available. In a few hours I was able to find the cause and speed up the service by 40% (!). It turned out that one of our developers used the ‘eval’ function of Python in the template tag — it was very slow and executed every refresh of the page, so its elimination brought a dramatic effect.

Piotr: In retrospect, what other decisions would you make when it comes to developing the key functionalities of Grono.net? What could have made it live on to even today?

Wojciech: Two things:

Less functionality: The fliers were too much; they were both a big expense and caused us to lose users both active and potential.
Focusing on our forums and building the community just around that: We had great thematic groups where you could find lots of interesting people. That was something unique — something that could be further developed into a forum for business. Grono had it and in this case Facebook wasn’t even our competitor.

Piotr: If you were to create Grono now, what would you write it in?

Wojciech: If it would be a new startup then I would write it entirely in Python + Django. In this technology you can produce a service in record time. I’d build the front-end in Vue.js or React, but it’s only because of the fact that today such service would have to be intricately complicated internally, reactive and high-performance. Normally I would stay with Python + Django in 100% without front end technology, because it is the easiest and fastest, and simple technology is the key to success.

Piotr: What is the biggest lesson you have learned from these 8 years of working on such a giant site, one of the biggest in the country? It was certainly a huge challenge and opportunity that many people would like to have. Many young developers and startup people will come to this material, certainly some words from someone like you, will be valuable to them, say something from yourself!

Wojciech: About the lessons, I could probably write a book because there were many and often painful. I think the most valuable for developers will be the following insights:First — programming is generally not as important in startup as programmers think. I used to think that as a programmer I am the most important, because I can build any product and will soon be successful because it is so great technologically. This is misleading — even if you can rewrite Twitter in one weekend, doing this alone isn't enough to succeed.The product is one thing, but building a business based on this product is a completely different story that usually requires incredible persistence, often strong nerves and skills that are far from programming (such as sales and marketing). And the most important thing in all of this is to conduct thorough market validation to verify if anyone will ever use our product before you write the first line of code.Second — the less the better. The larger and more complex your product, the lower your chances of success are because you:

Increase the cost of construction: Even if you are only investing your time, time is also a cost.
Increase construction time: This slows down the launch time, causing you to find out if the product really makes sense later after you’ve already sunk an investment.
Reduce the number of potential users: Yes, more features means fewer users. Because, the simpler the product, the easier it is to understand it and thus share it with others who also have an easier time adopting the product.
Reduce your chances of funding: The easier it is to convince someone to use a product, the easier it will convince the investor that he will gain from investing in it.

Also, if your product has a key idea or feature, just build it and integrate it with ready-made products (if possible) instead of rewriting what’s on the market.I made the above mistakes several times, sometimes it led to a complete failure, and sometimes it didn’t.

Piotr: Grono.net was created in 2004. In 2012, after 8 years it was closed. What has happened to Wojtek Sobczuk for the last 5 years to this day? You do not look like a man who likes to do nothing.

Wojciech: Unfortunately, due to my programming talent and my love for creating new startups, I’ve always had a problem with taking some time off to have a proper vacation. At the same time I am the type of person who focuses more on doing something than on talking about it during conferences or on blogs — that’s why you haven’t heard anything new about me for some time.During this time, I have worked on a few startups, one of which was quite successful. I built a billing system for a San Francisco startup that is now an enterprise-stage company with many users and employees. They still use that original system I built. I was also a CTO for another successful startup from the US and led that startup to an acquisition. In short, I was not bored.

By doing so, I got to know a lot of great people, and it came to me that the best solution for both me and the team I worked with on various projects would be to create VentureDevs. By creating one startup, the fate of the entire team depends on its success; and the team’s ability to build is limited by financing, but with VentureDevs, it’s a much different situation.At VentureDevs we can work on many projects for clients and friends from all over the world (often from the West Coast of the US because that’s where our CEO, Joe Gardner is from). The fate of the company does not depend on the success of any one of the products we may build or invest in. At the same time, we have full control over who we work with, and we are very selective here because we want to maintain a unique atmosphere despite the growth of the company. Also, we have an amazingly skilled team that allows us to build any venture, it’s just a matter of having the right ideas to build.Of course we are a young company now, a de facto startup, and there is so much work ahead of us. However, we are determined to be the best employer in Poland and hope that more and more of our customers and our own products will succeed in global markets.