Usually through the process of writing down thoughts and discussing them with friends and advisors, the correct answer comes naturally, often when you expect it the least. Sometimes it takes days, and sometimes months, but it will come if the decision you are considering is binary. This typically involves a choice between continuing what you are doing or moving on to something else. For instance, I wrote the email considering leaving OLX on July 30, 2012, and I left on December 17, 2012. I laid out the logic of that decision in Why I am leaving OLX which was much more fleshed out and thoughtful than my general sense of malaise and followed a long personal cost / benefit analysis.
Once the decision was made, I felt relieved, happy and confident that I had made the right choice. What was less clear was what to do next, which leads me to Step 3 in the decision-making framework.
Step 3: Explore many paths, find what sticks, iterate
As entrepreneurs we throw a lot of spaghetti on the wall to see what sticks. Once we find things that stick, we iterate on all their aspects: design, product, funnels, business model, team composition… Ultimately disruptive product change and success are the result of 1% improvements done 1,000 times over (though until you are at scale you need to find things that move the needle more to get statistical significance).
We don’t do enough of either of those steps in our personal and professional lives. Coming out of OLX, I decided to open my mind and free up time to try as many things as possible. To increase my flexibility and not tie myself down to my existing choices, I gave up my house in Bedford, my apartment in New York and sold my car. I gave 75% of my non-financial material possessions to charity and the rest to my friends and family. I detailed this in The Very Big Downgrade and explained how I went down to 50 items which fit in my carry on, backpack and tennis bag. This allowed me to experiment in my personal and professional lives.
On the personal side, I had two objectives: reconnect in a deeper and more meaningful way with my friends and family and find a work life balance that works for me. As I detailed in Update on The Very Big Downgrade, I tried lots of different approaches. I had the feeling that my friendships were fraying and becoming shallower. First, I tried to live on my friends’ couches and guest bedrooms. It was an utter failure and total disaster. As Benjamin Franklin once said house guests, like fish, start smelling after 3 days. The issue is that embedding yourself in other people’s lives while they still have obligations with work, their kids etc. only creates incremental burden for them. My vision of hanging out and remaking the world for hours on end was just not realistic given the constraints in their lives. That said I did not let that discourage me and kept trying. I organized vacations with friends testing every variable possible: time of year, group size, location etc. It took me a long time to find the solution because the ultimate answer went against all my core instincts.
My preference is to go on remote adventure trips when no one else is going on vacation. However, to really have my friends and family present during the vacation and not stressed or distracted we had to meet in a safe, comfortable, yet inexpensive place that is easy to reach, during traditional school and work vacations with lots of activities and supervision organized for their kids. Once I figured that model, it was easy to reconnect in a more meaningful way especially as it connected well with my new model for work life balance.
I started emitting ideas for life design in the “other considerations” section of Step 1 of the decision-making framework. I love New York. It’s the place where I feel at home. I love its intellectual, social, artistic and professional opportunities. As a result, my time in New York is always extraordinarily intense personally and professionally. Over the years, I realized I also needed to get away from the hustle and bustle and be in a place where the 15-year-old in me can run wild.
Sands Point and Bedford played that role: I installed a paintball field, padel court and remote-controlled car race track in the garden. I setup a movie room, console gaming room and PC gaming room. Tennis and go-kart racing were easily available. However, the setup was not ideal. Spending Friday, Saturday and Sunday outside the city was socially isolating. It took a lot of effort and organization to get people to come visit and I disliked the transaction costs inherent in going from one place to the other every 3 or 4 days.
The Mediterranean boy in me also yearned for warmth and sun. I started to suspect that a setup where I spent an extended period in New York City proper would combine well with spending an extended period in a place where there was little to do – ideally where I could work during the day, kite surf and play tennis in the evenings and that was it. It would be a place where I could be reflective and spend time reading, writing and thinking. After a pleasant period of “island shopping,” my heart settled on Cabarete (read Why I Chose Cabarete) which was suitable both for my work / life balance setup and as a gathering point for friends and family.
I then iterated a lot on time split between New York and Cabarete, how often to be alone versus have people come visit, group sizes and much more. Traveling from one to the other every week proved too tiring. Only making the trip every 2 months made me lose ties to the place I was not in. I ultimately settled on the following model: spend one month in New York, then go to Cabarete for a month and alternate every month. It gives me enough consecutive time in both places to have the time to live the full New York life, and then to properly disconnect in Cabarete.
During my Cabarete stints, I settled on splitting my time between being fully alone and having small groups of colleagues, friends and family come visit. This allows me to reflect, think through trends, read, and rebuild my energy half the time, while strengthening my relationships and being social the other half. In both cases a typical day includes several hours of emails, Zoom and Skype calls, an hour of kiting and an hour of tennis. The Dominican Republic also started serving as the gathering point for friends and family several times a year during school holidays.
I complement this setup with a few real vacations in different locales every year. I go heli-skiing the last week of January in the Revelstoke area. I organize a friends and family ski trip for President’s Day weekend in February. I spend a week in June in Nice visiting my family. I also go on two other week-long trips in exotic locales with a few close friends (no more than 4 of us typically) every year.
Now that I had my set up, the time came to figure out what I was going to do next professionally. Given that I settled on Cabarete, I started working on my Silicon Cabarete project. However, it was just never meant to be a full-time project, but just a fun development and gathering point for entrepreneurs, my friends and family. Having left OLX, I went back to the July 2012 email to try to pursue the seven ideas I had emitted in Step 1 of the decision making framework.
Try to convince Craigslist to let me run them
I had first approached Craigslist back in 2005 after I sold Zingy. I was impressed by their liquidity but had been disappointed by their horrible product and non-existent content moderation. I asked them to let me run product, content moderation, and their international expansion efforts. They were not interested at the time, nor were they interested in selling the company. Admittedly, their reluctance was understandable as I did not have credibility in this space, despite being a successful entrepreneur (I had grown Zingy into a profitable company with $200M in annual revenues).
I approached them again in 2013, after I left OLX, thinking that by now, I had earned the credibility. I had built a classifieds site that was larger than Craigslist with over 300 million unique visitors per month, 5,000 employees and with a much higher net promoter score. We had built what Craigslist should be: a mobile-first, spam free, murder free, prostitution-free, personals free, classified site, catering to the primary decision makers in all household purchases: women, creating a safe environment for them to trade.
I reached out to Craig and Jim and offered to run Craigslist for free for a year. Should they be happy with my work, they could compensate me in equity after a year’s worth of work and should they not be happy, they could let me go at no cost to them. I was sadly not able to convince them that this made sense. They frankly didn’t care to improve Craigslist. They were also not interested in selling to me despite the multi-billion dollar offers I was putting on the table. It was also amusing to note that every time we interacted, they seemed to have forgotten we had ever talked before.
Try to extract the non-strategic countries out of OLX to have a dividend stream + try the classifieds 3.0 strategy
Failing to convince Jim and Craig, I moved on to my next idea. I tried to convince Naspers to let me buy out their non-core countries, like the US, to launch a classifieds 3.0 strategy. Sadly, they were also not interested. They felt it was complicated to extract some countries and it did not move the needle enough for them to be worth considering.
I considered independently launching a new mobile classifieds site to go after Craigslist in the US. I reached out to my head of mobile at OLX after he also left OLX. Unfortunately, he was tired of classifieds and decided to pursue other paths, so I set the idea aside.
Think through the new idea I should build or company I should run
Inherently I prefer building companies to investing in companies. I get a stronger sense of fulfillment by creating something out of nothing and directly impacting the lives of my employees and users. In fact, I had never meant to sell OLX. I would have been happy to run it forever given the positive contribution we were making to millions of people’s lives and the scale we had reached. Sadly, when Schibsted, a publicly traded competitor, started aggressively attacking us in our core markets, we needed the capital to fight back which led us to partner with Naspers. They proved to be the right partner and we won the war but had lost our independence. After I left OLX, I started thinking through new companies I could build. As I mentioned in Step 1 of the decision-making framework, I considered building a competitor to www.justanswer.com. I put a lot of thought into 3D printing. I also considered bringing www.made.com to the US.
Ultimately, I decided that none of these opportunities were large enough to warrant my full-time commitment and decided to stay open to new opportunities as they presented themselves.
Try to buy the IP for Rise of Nations
Rise of Nations was my favorite real time strategy game after Age of Empires II. Big Huge Games, the parent company, went bankrupt trying to build an RPG game and its various assets were placed for sale. I bid on Rise of Nations but was outbid.
Frankly I am not sure it would have been easy to improve the game to fulfill the vision I had for it of improving the graphics, creating multiplayer auto-match and adding the tactical unit control with cover of a game like Company of Heroes 2.
The buyers just re-released it as is on Steam which was a bit disappointing, but easier to do.
Become a credible public intellectual
Because of my intellectual curiosity and voracious reading, I fancy myself a bit of a polymath. While I sometimes worry that having so many interests might make me a bit of a dilettante, I’ve often felt I had the knowledge to opine intelligently on many issues. This is especially true when it comes to economics.
I’ve always been both an economist and an entrepreneur at heart. I studied economics at Princeton because it created a framework that I felt fit best the world we lived in. It’s my fascination with markets as an economist that led me to focus on building and investing in marketplaces as an entrepreneur.
Over the years, I have written a lot about economics and spend a fair amount of free time thinking about macroeconomics and microeconomics. I thought it could be interesting to have my thoughts reach a broader audience.
I reached out to public intellectuals to understand the journey they took to get there. There seemed to be two components to the answer:
- Write op-eds for well-respected newspapers such as the New York Times or the Washington Post. Op-eds have a reasonably rigid format and are around 800 words long.
- Write on a reasonably narrow range of issues to associate your brand with that issue. This is generally true for the likes of Paul Krugman, Nouriel Roubini, Nassim Taleb and Niall Ferguson.
I started going down the path, but upon writing a few op-eds, I realized that I didn’t like the exercise. I preferred the flexibility and freedom offered to me by writing a blog like this one. It also occurred to me that given that people have a limited mind-space and associate one site with one topic, I would have more traffic if I only wrote about entrepreneurship where I am considered the most legitimate.
That said it’s not the way my mind works. I like sharing about all my adventures, travails and random thoughts, whether it’s book or gadget reviews, thoughts on entrepreneurship, economics or how to make important decisions 🙂
I also took the time to write an 11,000-word article: The Economy: An Optimistic Thought Experiment
While I came up with a total of 8 options in Step 1 of the decision-making framework, they were just a starting point. As I went down certain avenues and exchanged ideas with friends, I came up with many other opportunities and went down many a rabbit hole.
Age of Nations
For instance, while I was bidding for the Rise of Nations IP, I talked to lots of game studios who could help improve the game. When I failed to buy the game, my brother Olivier and I decided to hire one of them to build the real time strategy game of our dreams. As I detail in 2018: Giving Up and Moving On, it ultimately turned out to be a failure, but it was an interesting experience nonetheless.
Build and Run Special Economic Zone in Cuba
In 2013, as I was thinking and writing about economics and starting to work on Silicon Cabarete, the Cuban government started to liberalize the economy and announced they were interested in having Special Economic Zones (SEZs) in the country. At the time the average Cuban salary was $19 per month and the annual FDI was $110 million. Given the quality of the education in Cuba, its geographic proximity to the US, and how badly the formal economy was being run, I felt there was a real opportunity to create a zone which would be a beacon of growth and hope which would better the lives of millions of Cubans.
I collected commitments for investments of $300 million to launch the project, with the intent to invest over $1 billion over 4 years. I originally pitched them for an SEZ in Mariel, and in 2014 pitched other locations when they decided to do a SEZ in Mariel with Brazil.
Given the economic situation in the country, I felt we could do no wrong provided, we had the right to operate freely. The issue is that from a technical perspective it really needed to be more of a Special Administrative Zone (SAZ) rather than a SEZ. I pitched it as a SEZ for marketing purposes but included all the elements we needed: a convertible currency (not the Cuban peso) and an independent constitution. This constitution protected private property ownership, personal and corporate business rights. The idea was that it would be valid and prevail for the granted territory over the national constitution for a list of pre-agreed bylaws that covered the rights we needed. The national constitution would be valid in the SEZ for everything which was not specified in the SEZ constitution.
The zone was to be called Nueva Libertad or New Liberty. You can find the last version of my pitch to the Castros below.A-framework-for-making-important-decisions
Ultimately, they were not comfortable with setting up a region that had a different currency and legal system, and I was not comfortable following through with the project unless it did, so I regretfully moved on to other projects.
Trying to buy eBay Classifieds in 2015
After my attempt to build a special economic zone in Cuba fell through, the most ambitious project I worked on was to try to buy eBay Classifieds from eBay in 2015. Over the years, I had become friendly with John Donahoe who was CEO of eBay from 2008 to 2015. As eBay was splitting its core marketplace assets from Paypal, I partnered with a private equity firm to try to convince them to sell us their classified assets which I was interested in running. We did a lot of work on the project, but sadly eBay ultimately decided to retain its classifieds assets.
The Genesis of FJ Labs
While most of the ideas I tried failed, two ideas from my Step 1 email from 2012 really worked: angel investing in marketplace startups and helping build a new startup every year.
- The Accidental Creation of a Startup Studio:
To improve deal flow and processes, I had already started doing all my angel investments with Jose Marin. When I left OLX, we had already made over 100 angel investments. In fact, despite being a full time CEO, I was often more known as an angel investor, than as a founder. Given how busy I was running OLX, I decided to only invest in things I knew innately, marketplaces, and I created a set of heuristics to decide whether to invest in a company in a 1-hour meeting. Despite this, I was stretched very thin simultaneously running a large company while investing in many startups.
Fortunately, I was approached by Morgan Hermand-Waiche, a hungry, young entrepreneur. He originally wanted feedback on his idea, but after a number of iterations (read Grit, Perseverance, and Happenstance: The Genesis of our Apprenticeship Program) he joined me part time while in this second year at HBS to help me filter my inbound deal flow and to find a new idea for him to launch. This forced me to explicitly codify my heuristics and test whether my processes could be imparted unto others. It turned out they could.
It also led to the creation of AdoreMe. Given the success of having an apprentice to help filter the crazy amount of inbound deal flow and the success of AdoreMe, Jose and I created a formal apprenticeship program. We started having apprentices from the top business schools every year. We taught them venture capital during the summer between their first and second year, and part-time, 15 hours a week, during their second year until they graduated. The idea was that upon graduation they would become full time Entrepreneurs in Residence (EIRs) with the objective of becoming co-founders with us of companies we would build together.
As we were ideating for the next company to build, we went to Indonesia to consider launching an OTA like Expedia.com there. While the opportunity was real, we ultimately decided we did not want to deal with the Indonesia-specific legal issues and regulatory framework or deal with the 12-hour time difference. While we abandoned the idea of building that company, we kept ideating and iterating on new ideas to bring to the world and ended up creating several new startups in addition to AdoreMe:
- Lofty, an art marketplace
- Beepi, a managed car marketplace
- Rebag, a managed handbag marketplace
- Poncho, an insurance marketplace
- Instacarro, a consumer to dealer marketplace in Brazil
- Merlin, a blue-collar job site
- Properly, a transactional real estate marketplace in Canada
Not all succeeded, in fact Beepi infamously blew up after raising $149 million, but our startup studio model proved successful in attracting amazing talent, led to the creation of several successful startups and was fun to partake in.
We iterated a lot on the model before settling on the model as it stands today. We train the apprentices for a year, then they join as EIRs. We jointly ideate until we come up with an idea we want to build together. Jose or I join as co-founders and executive chairmen and invest $750k for 35% of the company. The team collectively has 65%. We also have the right to invest the next $2 million at either $8 million pre-money or at market terms if the entrepreneurs can get a higher valuation.
We help de-risk the project given the time we take to ideate and test models. We allow the entrepreneurs to focus fully on executing for the first two years by allowing them not to have to seek pre-seed or seed funding if they don’t want to. We also help actively with marketplace design, recruiting and fundraising. This model has allowed us to attract amazing entrepreneurs in the program. Every year 250 first-year MBA students from Harvard, Columbia, MIT, Stanford and Wharton (and we’ll add Stern next year) vie for 2 to 3 spots.
There are many companies we want to build. A recent brainstorm at our Turks & Caicos offsite led us to come up with over 140 ideas! I am sure that one of them will germinate soon enough and I look forward to being able to tell you about it soon.
- The scaling of our angel investing activities:
At the same time, angel investing took on a life of its own and we ended up investing in 60-130 startups every year. As an entrepreneur, I hated that I never knew where I stood with VCs. I would have a meeting which seemingly went very well but would not hear back from them for weeks, if ever. Meetings would be spread across many months. I suppose most VCs don’t want to pass explicitly on a deal to preserve optionality, but it’s extraordinarily frustrating as an entrepreneur. As a result, we decided to evaluate startups in no more than 2 1-hour meetings that take place in less than 2 weeks and to always tell the entrepreneurs where they stand: why we are investing and more importantly for most, why we are not investing and what we would have to see to change our mind.
That approached proved appealing to many entrepreneurs and our deal flow kept increasing, as was our reputation as friendly value-added investors. We don’t lead, price, take board seats, have reporting requirements, or have minimum ownership requirements which allows us to be very flexible. We are also comfortable investing at every stage and internationally. 65% of our investments are pre-seed and seed, 25% are Series A & B and 10% are late stage. 70% of our investments are in the US and Canada, 20% in Europe and the Nordics and 10% Brazil and India.
While we invest in every geography at every stage, we do have a specificity: we focus on marketplaces and are thesis driven. 70% of the companies we invest in meet our thesis and they all meet our heuristics. That approach of extreme flexibility while being thesis and heuristics driven proved to be very successful. To date we invested in 475 startups, had 163 exits. On these exits we had a realized average multiple of 4.4x and IRR of 61% including all the failures.
- The formal creation of FJ Labs
While we scaled both our startup studio operations and our angel investing activities, it was still not a given that it required the creation of a venture fund. While our approach has been extremely successful, there is a relatively low maximum to the amount of capital we can deploy with our strategy. Because we are not leading and pricing rounds, this puts a maximum check size that we can invest in any given round. We don’t want to compete with traditional VCs, but instead be valuable thought partners for them and a source of differentiated deal flow. With today’s round sizes we can invest a maximum of $225k at pre-seed, $450k at seed, $1M in a Series A and a few million in a Series B but are often limited to smaller investments.
Given that we don’t start many companies because our hands-on approach is not very scalable and only invest the first $2.75M, that also does not create an opportunity to deploy a huge amount of capital. Back testing our model, it looked like we could perhaps deploy $100M per year without changing our strategy. We were not successful enough to deploy anywhere near that amount of money, but it was not an issue. We were happy investing in our own name, working with just the apprentices. It would have been nice to have more firepower, but we did not want to lose our flexibility or bother fundraising from institutional LPs who would run away in the other direction when they heard of our crazy investment approach.
On top of that the economics of small funds are not particularly appealing. If you deploy $100M and do a 3x, you have $200M of profits. With a 20% carry, you have $40M in profits which you split between the partners. That’s pretty good, but not particularly impressive. However, if you deploy $100M of personal money and do a 3x, you get $200M of profits for yourself. In other words. small funds only really make sense if you are deploying a lot of your personal capital alongside the fund because that’s where the real returns come from. It’s nice to have more firepower and some fees to cover our costs by raising external capital, but it’s not worth going out of our way to pursue.
As we were having this reflection, we were approached by Telenor, a Norwegian telecom operator with a large presence in South East Asia and 174 million subscribers. To my great chagrin, Telenor had funded Schibsted in its war with OLX which is ultimately what led me to sell OLX to Naspers. That said, OLX’s merger with Schibsted in Brazil and other markets was very profitable for Telenor and gave them direct ownership of several classified assets in South East Asia. Given Telenor’s digital and marketplace ambitions, they reached out to us to see if they could invest in us in order to have a looking glass into the future by having exposure to US tech trends to either defend against them or bring them to their markets.
As they were comfortable with our crazy strategy, while it gave us more firepower, and provided a small fee base to start building a real team, we decided to pull the trigger and FJ Labs was formally born in January 2016 with a $50M investment from Telenor complementing our personal capital and the small entrepreneur’s fund we run on Angelist.
As the relationship proved successful all around, we decided to scale up the partnership and open it to other strategic investors and family offices. Our second institutional fund should close in the coming months with $150M of external capital.
Sellit / Wallapop / LetGo
The beauty of the FJ Labs setup from my perspective is that it has the flexibility for me to step in operational roles should the situation call for it as it might have had the Craigslist, Cuba or eBay Classifieds opportunities had materialized. In 2014, I circled back to my 2012 idea of building a mobile classifieds site to attack Craigslist in the US. OfferUp had since launched and it would have been clearly better to launch in 2012 when I first had the idea, but they were not following the OLX playbook or the best practices in classifieds and I felt the market was still open. I scaled back my involvement in FJ Labs to the weekly investment committee calls and set about building and running Sellit.
We launched in New York, proved out our playbook and started fundraising. In the meantime, the category had gotten very competitive with the launch of several well-funded players such as Close5, VarageSale and most importantly LetGo, backed by OLX, which was launched by Alec Oxenford, my OLX co-founder using the playbook I had helped put together. While we got several term sheets, I felt it would be best to align with a well-funded competitor, so we merged with Wallapop, the leading horizontal classifieds site in Spain. They were backed by Accel, Insight, DST, NEA, 14W and others and had the firepower to properly take on the opportunity. I became chairman of Wallapop and CEO of the US operations.
We aggressively competed and grew the business. However, having witnessed firsthand the massive benefits of consolidation at OLX when we merged with Schibsted in Brazil and with Avito in Russia, it quickly became apparent it did not make sense to have two players pursuing the exact same strategy. We merged Wallapop’s US operations with LetGo’s in May 2016, giving the majority of the company to LetGo (and hence OLX). I retired as CEO, left LetGo in Alec’s capable hands, and returned to my normal FJ Labs duties.
All that to say that since that Step 1 email, I threw a lot of spaghetti on the wall both in my personal and professional lives. There were epic failures along the way, but I stuck to what worked and kept iterating. In the end I found both a personal life setup that worked for me and a new exciting professional adventure.