Garrett Camp, who came up with the idea for Uber, never drove a taxi. Joe Gebbia and Brian Chesky, who devised Airbnb, never worked at a hotel. They’re all outsiders who used software to massively disrupt established orders they were never a part of and had no interest in preserving.
If a Silicon Valley startup were to tackle the bulk shipping spot-chartering business – the short-term rentals of giant tankers and bulkers around the globe – the outsider-looking-in strategy would be a no-brainer: eradicate the ship-broker middlemen and democratize the data to provide supply-demand transparency, neutralizing information asymmetries.
Shipping insider Ioannis Martinos, the Greek founder of Signal Ocean, is following a different strategy. As he explained during an in-depth interview with FreightWaves, his company is now successfully using software to transform the spot-chartering business, but not in the way Uber, Airbnb or another “move fast and break things” disruptor would do it.
Augmenting not undercutting
Martinos was born into shipping. His family ranks high in the annals of renowned Greek shipping names, with the various branches owning the companies Thenamaris, Eastern Med Maritime and Minerva Marine. Ioannis is a ship owner himself and the brother of Thenamaris CEO Nikolas Martinos.
The Signal Ocean Platform is a software-as-a-service (SaaS) engine using technology to provide more timely and accurate situational awareness for spot-chartering decisions. It debuted with crude-tanker coverage and is expanding to dry bulk through a partnership announced on Oct. 28 with ship-brokerage giant Simpson Spence Young (SSY).
The platform is specifically designed to not be a threat to brokers, owners and charterers. It’s not trying to break the long-established mold, not out of some ingrained respect for the family tradition, but rather, because Martinos’ insider knowledge of how the industry works has convinced him that the most profitable course is to augment the status quo, not undercut it.
“We are not democratizing information – and one of our guiding principles is that we respect the market structure”, asserted Martinos.
“Uber’s ‘value added’ is that it connects people who do not know each other. That’s actually true for many of the platforms we all use at the retail level”, he said.
“But in the case of shipping, everybody already knows everybody. If, for example, you’re talking about Aframaxes [tankers than can each carry 750,000 barrels of crude oil], there are about 500 ships in the spot market today and you don’t need a new platform to find a ship or know who the owners are. Also, a lot of the added value brokers bring to the table is not actually booking the ships; it’s following up and executing the contracts.
“So, the Uber analogy does not work – but we think there is a lot of value to be added without attempting to be the Uber of shipping”.
How the platform works
The worldwide spot chartering business is like a massive three-dimensional multi-player chess game with millions of dollars at stake upon every move. The ships are the pieces on the board. Each decision to book a ship, called a “fixture,” has huge potential opportunity costs. Long-haul voyages can take months to complete; if a two-month voyage is booked at a time-charter equivalent (TCE) rate of $20,000 per day and one month later, the market spikes to a TCE of $100,000 per day, that ship has lost out on the higher rate.
The best way to play this high-stakes game is to know where all the game pieces are, when they will be in position to bid for various fixtures, who the players are that are moving each of those pieces around the board, and what the fixture bids are likely to be. And, of course, to have all of this information immediately.
The Signal Ocean Platform uses technology in two primary ways to do this, explained Martinos. First, it takes in all of a SaaS customer’s unstructured data in the form of broker fixture reports, emails, port agent reports and other information and uses narrow AI to swiftly decipher that data flow. This process is analogous to how computer algorithms “write” articles on quarterly financial reports and baseball games with no human editor. “It’s about ingesting all of the information and creating the most up-to-date version of the truth with the data that your company [the SaaS customer] receives,” said Martinos.
The second aspect poses a much more complex AI challenge. It involves using programmed shipping knowledge to make decisions based on incomplete data and ever-shifting market conditions. Martinos, who studied engineering and robotics at Tufts University and MIT, compares the difficulty of this aspect to the challenges posed to autonomous cars, which “have to synthesize one version of the world based on sensors even if one sensor stops working and conditions change”.
The second aspect of the Signal Ocean Platform engine “intertwines shipping knowledge with machine learning to create an agent that is intelligent enough to synthesize the information as if with the mental paradigm of a human shipping professional with at least 10 years of experience. We’re not talking about a magician. We’re talking about someone who has the experience to connect the dots and disambiguate conflicts”.
The output of this two-step process is different for each SaaS subscriber because their inputs in the form of broker fixture reports, etc. are different. A shipping outsider from California building such a product would probably seek to ingest all available data in the world from brokers and other sources and then provide the same outputs to all subscribers. Not Signal Ocean. “Under our software agreement with our clients, any information they send to their accounts is for themselves only. It’s like a Dropbox”, Martinos emphasized.
In-house versus third-party sales
Martinos launched a commercial ship-management company called Signal Maritime in 2014. Signal Maritime’s goal was to charter out ships, primarily Aframaxes owned by Martinos at first, and seek to use technology to optimize returns.
It launched a separate software development company called Signal Ocean in 2016. Signal Ocean’s SaaS platform was developed by first focusing on very large crude carriers (VLCCs, tankers that can carry 2 million barrels of crude oil each) because they have the simplest trading patterns, then Suezmaxes (tankers that carry 1 million barrels of crude each) and Aframaxes.
Subscriptions sales began to third parties in June 2018 and uptake has been rapid.
“More than 50% of the VLCCs, Suezmaxes and Aframaxes in the spot market are controlled by companies that now use our software,” said Martinos, who added, “Historically, I don’t think there has been any software in shipping that has been adopted by so many companies so quickly”.
He maintained that the benefits can be seen in the bottom line. Signal’s Aframaxes using the platform earned $18,000 a day in 2018, 20% above the $15,000 per day global average. This begs the question: If the upside is so lucrative, why not just keep the software in-house for the sole benefit of the family tankers?
The answer, Martinos told FreightWaves, is that there is ultimately more money to be made from SaaS subscriptions. “Obviously, we debated this a lot internally, and there were several things that drove us to that decision [to sell to third parties]”, he recounted.
“The first reason is that it’s like improving a competitive bicycle by making it out of carbon fiber so it’s stronger and lighter. It gives the rider a competitive advantage, but you still need to be a good rider in the first place. It doesn’t mean that everyone who buys the bicycle will immediately be able to use it to its maximum potential. Even though we’re not the only ones using our software, we believe we can maintain an edge because we built the software and we’re experts in using it.
“The second reason is that if you look at cases where software companies have been more open, and the open-source software movement in general, it tends to help the industry overall. And personally, for emotional and philosophical reasons, I am in favor of seeing the whole shipping industry develop and become more modern and become more attractive to younger people”.
Signal began with a staff of 11 and now employs 101, and has plans for significant expansion ahead.
“There are only so many companies in the dirty [crude and fuel oil] tanker market,” he explained. “There are maybe 100 companies in the world that are important from an ownership point of view.
“If we’re selling to 70 of those companies, we’re going to run out [of prospective clients] soon, so we need to go into dry bulk and then there is LNG [liquefied natural gas], LPG [liquefied petroleum gas] and containers. We can continue to grow as long as we keep expanding the vessel categories. Our ambition is to cover all of the vessel categories in the next couple of years”.
The recent deal announced with SSY is a major move. SSY is the world’s largest privately held ship brokerage; the biggest brokerage is publicly listed Clarksons, which has a software product that competes with Signal Ocean.
Signal had already begun work on the dry bulk module to complement its crude-tanker module before the new partnership was announced. The participation of SSY – which is also investing in the software product – will accelerate the process. “It will help us fill in a lot of the gaps [on the dry bulk side]”, said Martinos.
VC funding ahead?
When you think tech start-ups, you think Silicon Valley funding models – seed funding via convertible notes with valuation caps followed by preferred equity rounds in Series A, B, C, etc. Yet you’ll find no post-money valuation for Signal Maritime on the CrunchBase website.
“We haven’t had any external funding until now [the SSY partnership funding]. So far, the necessary funding has come from me personally for the most part”, said Martinos.
That said, he’s no stranger to the VC scene. He was the lead seed-round investor in Nutonomy, a startup that provides software for self-driving cars. He served on the Nutonomy board with Chris Thomas of Detroit Mobility, who is now a senior advisor to Signal Ocean. Nutonomy was bought by Delphi in 2017 for $450 million.
Asked whether he’d consider external VC funding for Signal Ocean, Martinos replied, “We have been approached by some growth funds and we are entertaining some of the discussions, but we are not actively seeking a financing round at this stage. If we decide to become a lot more aggressive, I wouldn’t rule it out. I’m not 100% sure the hypergrowth model is the right one for a shipping software company, because it’s not a consumer type of model where you do a ‘land grab’ – but I want to keep my options open”.
Also published on Medium.