How To Kickstart A Marketplace Business
Building a marketplace and lessons learned from Uber & Timberhub, community building in enterprise software, a $300m funding round, DevOps scaling, jobs, events, and more
Welcome back to another issue of Startup Pirate! The newsletter with tech trends, startup insights and tactics behind successful companies - along with all the latest updates, jobs, and resources from Greek startups and Greeks in tech. I’m Alex and you can find me on Twitter or LinkedIn. Join 4,020 readers by subscribing here:
How To Kickstart A Marketplace Business
Today, I’m really excited to have Ion Sergis in the newsletter to discuss how to kickstart growth in marketplace businesses. After launching Uber around the world, from Greece to Pakistan and South Africa to Colombia, Ion co-founded a b2b marketplace startup in the timber industry, Timberhub, with Dimitris Pagkratis, Giannis Androutselis and Thomas Androutselis. They started a year ago and they already have customers in 15 countries and raised funds from prominent VCs and angel investors. We learn:
stories from launching Uber in four continents
how to build trust in a new marketplace business
doing things that don't scale in the early days
top reasons marketplaces fail
the cold start problem in marketplaces - focusing on supply vs demand
how Timberhub aims to make a dent in the timber industry
Let’s get to it!
Ion, it’s great to have you on Startup Pirate today. You have been running marketplace businesses for the past years, first at Uber and now with your own startup. You were part of Uber’s early team here in Greece. How did you drive initial growth for Uber in this new market?
IS: I started working for Uber in Athens back in 2017. I was recruited as the Head of Operations and part of a team of three. During the next 1.5 years, we grew the network of drivers from 30 to 7,000 and the number of trips/week from 200 to 110,000. The Greek market was particularly challenging. Greece was dominated by Taxibeat and taxi fares were relatively cheap compared to other countries. We unlocked demand (passengers) with a very aggressive referral mechanism, while also using promo campaigns with reduced fares e.g. during night hours. But first of all, we had to grow the supply (drivers). That’s where we rolled up our sleeves and did lots of the things that don’t scale. There was no relevant regulatory scheme in Greece to cover Uber drivers (a special driving license was required for taxi drivers and this was very expensive to obtain), so we had to start from scratch; organise driver seminars, partner with the Hellenic Manpower Employment Organization (OAED) and the Ministry of Tourism, use fleet partners to employ drivers, etc. After lots of iterations, we managed to unlock supply and reduce the time from driver sign-up to hitting the road from 120 days to 30 days. Uber was growing nicely in the market until 2018 when it shut down operations due to changes in regulations. After a couple of months, it went back to business, but with a new model (yellow cabs only) and at the moment it’s probably as big as Beat in the city.
After Greece, you went on and launched several new markets for Uber across the world. How did you decide which new cities to unlock and what were the most important metrics you tracked for every city?
IS: After Greece, I helped Uber launch or scale operations in Pakistan, South Africa, Ukraine, Finland, Colombia, and Spain. The decision to launch a city was taken based on its population density (the more densely populated a city the better), regulatory context (is there available legislation that permits Uber drivers to hit the road?), and existing taxi experience for passengers (a mediocre taxi experience in the city leaves room for new entrants). The population density was the #1 factor. But every city had its own needs, so we had to create separate business models that catered for those. For example, in Pakistan, we had motorcycles, and tricycles, while certain two-seat cars were our most premium vehicles; in South Africa, all available cars on the platform were large and spacious, etc.
Nevertheless, what truly mattered and determined whether we were successful or not in a city were the taxi fare, taxi availability, and how fast a taxi would get to the passenger (so-called ETA). These were our high-level metrics, but we also tracked multiple steps of the passenger and driver experience that reflected the overall health of the marketplace. How many users open the app to look for a taxi? How many users request a taxi? How many users complete a trip? What was the drivers’ utilisation? How much money did drivers make? etc. We used those data to optimise every part of the process. The harder thing (because it sounded counterintuitive to many) was to persuade drivers to reduce fares in order for them to make more money. I have some crazy stories to share here… Drivers burning cars, drivers going after Uber employees, etc. Yes, Uber is a tech-enabled company, but there are humans behind every transaction. It’s not just optimising funnels and metrics and you have to respect that.
What are some of the things that don’t scale and you had to do in the early days of a city launch with Uber to find early adopters?
IS: During the first year of Uber in every city we literally had to do such things every day. For instance, in Athens when we were still a team of three, we used to have a dashboard with a map and the location of each taxi on the platform. I remember calling drivers on the phone that would drive away from the city centre, up to the mountains, so they don’t get assigned to a trip - while they were still receiving the guaranteed driver’s incentives, which was part of our strategy to kickstart supply and recruit more drivers! We had to do all sorts of things to recruit drivers from distributing leaflets outside OAED to creating a vehicle solutions company with private investors to supply cars to our drivers. To some extent, every marketplace has to start as a managed marketplace, to address the cold start problem. A marketplace needs buyers and sellers. So in the very beginning, when you have 0 buyers and 0 sellers, how do you begin? Oftentimes you need to subsidize one of the two sides of the network and do things that cannot be done in an automated way or are not scalable in the long run, just to get the marketplace started.
Talking about the cold start problem, how do you decide which side of the marketplace to put most of your resources behind — growing supply vs demand?
IS: This depends on the marketplace characteristics. Is it oversupplied or undersupplied? If it’s undersupplied, it means that the buyers have a problem finding a supplier (a driver, a sawmill, whatever that is). In that case, the best way to start is to find some buyers with common characteristics, understand their needs and then provide them with some supply that satisfies them. On the other hand, if the market is oversupplied, then the buyers have high optionality and the suppliers are not able to sell. In that case, you follow the opposite approach. Either way, it’s crucial to pick a niche within the industry. For example, with Timberhub in the beginning we had to focus on a specific part of the timber industry - we couldn’t kickstart growth by selling all types of timber. All in all, you should understand which of the two sides of the network needs you more in every stage and provide enough value to get them hooked.
A year ago, you decided to launch your own marketplace in the timber industry. How do you evaluate a marketplace business idea? What made timber interesting?
IS: With my co-founders Dimitris, Giannis and Thomas, we evaluated several ideas before starting Timberhub, as we had been looking for an industry that ticked certain boxes. What we evaluated - and what I’d recommend everyone to consider before launching a marketplace business - were:
fragmentation of demand and supply. If one of the two sides is not fragmented, then there’s no need for a marketplace to exist or the potential is limited.
transparency. Are there enough public data for prices and availability? The fewer the available data, the more critical the role of a marketplace in that industry.
price volatility. High price volatility of the goods/services between buyers and sellers means more uncertainty, which increases the need for a marketplace to exist and facilitate transactions.
Despite the fact that none of us had prior experience in the timber industry, we all got very excited about the problems it faces. Timber is a $600 billion industry and it’s forecasted to grow 2x in the next 10 years. There are over 30,000 sawmills and 120,000 enterprises that buy timber in Europe - from construction companies to wholesalers, and from carpenters to pallet manufacturers. Therefore, both supply and demand are extremely fragmented, which means none of them has significant power to set their own rules in the network. Let’s not forget that this is an industry where transactions happen at a global reach. Logistics are not a deterrent factor to trade (the cost is relatively low). If you’re based in Greece, you can buy timber from Scandinavia, Austria, Turkey, Lithuania, etc. Hence, this contributes to further market fragmentation. Furthermore, timber prices are volatile with significant fluctuations every week and there’s no single source of truth to inform buyers and sellers about the price. We think that this industry has some great characteristics and enough space for successful marketplace businesses to emerge. But of course, it’s all about execution at the end!
What problems are you solving for the timber industry? What’s your vision with Timberhub?
IS: At Timberhub, we focus on craft enterprises, manufacturing sites and construction companies that use wood as raw material. Usually, these are SMBs where 60%-70% of their revenues equals the purchasing cost of timber. So we help them purchase timber from the right supplier, at the right price and payment schedule, as well as with the right inventory and delivery time. At the same time, we help timber suppliers increase their orders, while expanding their geographical reach. We have suppliers based in Romania that sell to France, suppliers from Bosnia that sell to Germany and Greece, etc. Think that 95% of timber suppliers around Europe today don’t have an online presence, therefore they receive requests via WhatsApp, phone calls, and emails, from a limited range of 300km-400km away from their factories.
We’re very excited to operate in this industry, first and foremost, because we believe we can contribute to the decarbonisation of construction through timber. For every cubic meter of concrete that’s replaced by wood, we save 80% of emissions. And a well-run marketplace that adds transparency, higher price predictability, and guarantees quality, can contribute towards that goal. There’s also the misconception that forests are destroyed by logging. The problem is not logging per se, it’s eliminating forests for agricultural purposes. By empowering people to maximise their returns on timber sales through our marketplace, the use of available wood for manufacturing, construction, etc becomes much more efficient. We’re just getting started on our journey!
What’s your traction so far?
IS: We raised a pre-seed round last March from Speedinvest, the founders of Sennder (one of the largest logistics companies in Europe), alongside other prominent angel investors/founders such Alex Argyros, Manolis Manassakis, and more. Since then, we have built a clientele across 15 countries with 80% of them being in Germany and sold over €2m in gross merchandise value with positive margins. Our team is 12 people strong with plans to grow to 25 within the next six months. We are based around Europe, while also having an office in Greece. Soon, we’re going to announce a seed funding round led by very prominent European investors. You can learn more about us on our website. Oh, and did I mention we're hiring?
Unlike Uber, with Timberhub you had to start from scratch - I mean there’s no brand awareness the way you had with Uber in most new cities. How do you build trust in a new marketplace?
IS: This is a great question and in fact one of the hardest parts of new marketplace ventures. At Timberhub we did two things to address this:
we ask buyers to proceed with the payment after they receive the product and check its quality (to do this we had to structure our processes in a way that we buy the timber from suppliers and sell it to buyers, as soon as the transaction between both parties is confirmed - we aim not to maintain any inventory, but of course, there’s some risk here). If they’re happy, then they pay us.
we hired a sales team with prior experience in the timber industry, who already had established relationships and were familiar with the problems of both buyers and suppliers. This helped us a lot to start conversations.
What do you think are the top reasons marketplaces fail?
IS: I think there are reasons that have to do with how the industry, where the marketplace operates, is structured. We touched upon this before. Fragmentation, transparency and price volatility are key. If an industry has few large buyers or sellers and the purchasing power is concentrated, or if there are mostly established relationships and high transparency between buyers and sellers, or if the price of the services/goods is not volatile, then chances are that this industry is not the most fertile ground for very successful marketplace businesses to emerge. Another important reason why new marketplaces fail is the lack of focus. Not going after a particular niche within the broader industry, when you try to unlock demand or supply is usually a recipe for failure. But again, it’s all about execution at the end of the day!
Very good. It was great to talk to you, I appreciate you taking the time and we should talk again sometime soon.
IS: Thanks Alex, congrats on the great work that you are doing!
Are you looking for your next career move? Check out 960 job openings from Greek startups hiring in Greece, abroad, and remotely.
Mysten Labs raised a $300m Series B round led by FTX Ventures valuing the company and creator of the Sui blockchain at over $2b.
Dyania Health raised $5m Seed to advance their NLP technology to drive better outcomes in clinical research, led by Innospark Ventures and Big Pi Ventures participating in the round.
Rated Labs announced a $2.5m Seed round to accelerate its mission for greater transparency and rich context in Web3 infrastructure data. 1confirmation, Semantic, Placeholder and more investors participating in the round.
Clicktotherapy, an online mental health platform, raised a Pre Seed round led by Genesis Ventures.
Cohere Network, a member-owned network of co-living communities, was accepted into Techstars Anywhere Summer Class of 2022.
Interesting Reads & Podcasts
Communities, enterprise software and building moats - some thoughts I wrote on different types of communities, reasons to invest in community building, and how to get started, here.
Thanos Papangelis, co-founder & CEO of Epignosis, on his journey building an edtech startup, fundraising, and more, here.
Lessons from performance athletes and building with impact in startups, by George Hadjigeorgiou, co-founder & President at ZOE, here.
Prioritising work and learning to say no constructively, with Zaharenia Atzitzikaki, Design leader and management consultant, here.
A post on scaling from zero to (DevOps) hero with Atlantis, from Dionyshs Tsoumas, Head of DevOps at GWI.
A post on navigating in fast-paced environments as a Product Manager by Manos Kyriakakis, Head of Product & Growth at Simpler.
Creating oracle networks with the Chainlink protocol from Stelios Gerogiannakis, Senior Engineering Manager at Ignite, here.
Finding your market fit with Dimitris Kalavros-Gousiou, Partner at Velocity Partners Venture Capital, here.
A podcast with Rob Long, CHRO at Workable, discussing his unconventional journey to Chief Human Resources Officer.
How universities can foster an entrepreneurship mindset with Katerina Pramatari, Partner at UniFund, here.
“UX Greece welcomes David Bland” by UX Greece on Sep 21
“Opportunities to Work in Greece” by Starttech Ventures on Sep 21
“Greeks in Tech: PARIS” by Marathon Venture Capital on Sep 23
“26th WordPress Thessaloniki Meetup” by Thessaloniki WordPress Meetup on Sep 24
“Voxxed Athens 2022” on Sep 30 - Oct 1
If you find this newsletter interesting, consider sharing it with friends, or subscribing if you haven’t already.
Thanks for reading and see you in two weeks,