The top 25: Saemin Ahn, Rakuten
There might be relatively few non-Japan-based corporate venturers investing in the world’s third-biggest economy, according to analysis by GCV Analytics, but one of the most interesting to do so is Saemin Ahn, managing director of Rakuten Ventures.
Singapore-based Ahn runs the corporate venturing unit of Japan-based Rakuten, a media to financial services internet group, and the distance gives him perspective as well as independence for the structure he has set up.
This distance allows him to ask hard questions. In a questions and answers profile in GCV last month, Ahn said: “One constant way that I look at investment targets is that I try find the right question. Creating the right problem statement takes the most energy. And once that is done, I simply research the industry and the company online.
“One example is when Cloudera got investments for their big data offering and Horton Works was IPO-ing on the stock exchange. This industry was seen by many as the next value goldmine. When I saw that much money being shoved into those pipes, I had a couple of predictions and those predictions came true.
“Number one was that it is not lack of technology output from big data that is going to be the issue, it is going to be that you will not find enough data scientists to build out the outputs that you want. It is because those data scientists are being employed by Cloudera or the likes of Google, or they are actually building out their own startups. That is what market abnormalities create. It creates giant vacuums of talent that go into different black holes or, secondarily, they themselves actually want to dream bigger and build out their own companies. Based on that problem statement, there are not going to be enough data scientists and algorithmic engineers to go inside companies to solve problems that these big corporations need.
“With that I asked how Google solves those problems. Google has one really important thing that they do – they use an algorithm repository, company-wide, for engineering teams that enables them to retrieve code in a consistent manner. I said to myself: ‘Why is there no algorithm marketplace that does the same thing?’ I literally typed in algorithm marketplace and Algorithmia popped out. We invested in the seed round with Madrona, and their series A was done by Gradient Ventures, which is Google’s AI fund. If Google’s AI fund is convinced enough to invest in Algorithmia, then my problem statement was material and durable.”
But having asked the tough questions, Ahn keeps Rakuten Ventures to a small portfolio. He said: “The reason I would say 17 to 15 companies is because we are not a commercial VC. The way that commercial VCs would activate is to ensure their risks are disbursed, and for them not to deploy capital to their widest extent is doing a disservice. We suffer no penalties for being patient. This is the biggest superpower that a corporate VC has. I do not want to focus on spending too much time investing in many companies but simply make sure that Rakuten Ventures can grow as a brand with good reputation and good exits. You cannot go full sprint from the very start.
“As with anything, you have to take baby steps first. I am very much reticent about large capital deployments. Once we start the flywheel of dealflow, those will opportunistically and organically generate one or two clips that will provide us with opportunities to invest a lot of money in a few companies.”
In 2016, Rakuten doubled the size of its Global Investment Fund to $200m and promptly led the $120m round for Spain-based ride-hailing app Cabify, while also launched a ¥10bn ($85m) Rakuten Ventures Japan Fund to fund invest in startups in its home country.
These new commitments are all built on the original $10m Southeast Asia-focused fund launched by Rakuten Ventures in 2013, when Ahn arrived from search engine provider Google (now Alphabet), and this initiative was joined the following year by the original $100m Global Investment Fund for investments in startups based in the US, Israel and Asia-Pacific region.
A native of Korea, Ahn had joined Google as an online partnership manager in Korea after some marketing work and following degrees at Sogang University. He then moved to Singapore with Google in 2009 and from there stepped up to cover the whole Asia-Pacific region.
This year, Ahn said Rakuten had already closed one deal. “It was a follow-on for an e-book subscription company. We are probably going to do one or two more deals in London in the third quarter. Annually we are trying to do at least four to six investments, whether it is follow-ons or new deals.
“The unit was set up for Southeast Asia, but, to be honest, I did what I wanted to. I was just hoping not to get fired the next day. That has been my modus operandi.
“When you look at a strategic investment department, it has all the bells and whistles of what a perfect Power-Point looks like for a department. It is going to be super-intelligent, and “we are going to invest in the best strategic stuff”. It has never worked out historically unless you have absolute control and material freedom of liquidity. When I got in, I said that this is not going to work and I ripped it all out and simply said let us be a true-blue VC firm with a single LP.
“Thankfully, it has been a very good run. If you look at us just by the numbers, we are a statistical anomaly. We have 14 companies and not one has folded. Several of our companies are profitable – they are running at a very nice clip of growth. We are very lucky to have such motivated and high-quality founders to work with.
“We just made sure that, number one, we are the correct umbrella to drown out all the corporate noise, and then look at adding value to the business.
“One of our strongest suits, weirdly enough, has been our insight into the founders’ products, the company’s products, and that is simply how I fed back to the companies. This is how I feel about your products, this is where I think the problems are, this is where I think the growth is. I like your product but I think if we could pivot it a little bit this way, it will be much more desirable for corporations to take up.
“It depends on the company. Some companies have hiring deficits, so we try to bring in head-hunters that we have known very well from one area to different areas. We try to deploy them and help them on those issues. If there are business planning issues, we have really smart people on our team that can help out. Our goal is to be flexible to their needs but to be pragmatic and sensible as well.
“Being more of a contrarian has made for a lot of exciting verticals to focus on. If many investors head in one direction, I automatically look at what they have left behind. If you look at artificial intelligence (AI), right now a lot of the effort and energy is going into autonomous vehicles and looking to solve generic wider problems. This is the cycle I observe with the startups looking at narrow applications and looking at wider stencils.
“So, the question I would ask myself is: If you want AI that is able to navigate exigent circumstances with many real-life variable elements – like autonomous vehicles – what kind of or how many millions of tests do you actually have to create and run to train those neural networks progressively and make them effective enough to deploy? How much does it cost, in time and actual dollars, to run those tests? Can those tests be replaced with simulations in virtual environments?”
“With the above questioning, I look at simulation startups. That approach is much more interesting to me. What are the niches those sorts of pressures create organically?”
It has led Ahn to back groups as diverse as ride-hailing service Go-Jek, bots developer Run Dexter, invoicing software MakeLeaps, advertising technology providers PocketMath and AdsNative, notifications service OneSignal, children’s books service Epic, file sending app ESTmob, payments business Coda Payments, online market Carousell, as well as Algorithmia and Visenze.