Mårten Mickos, the CEO of HackerOne (2019 update) and formerly the CEO of Eucalyptus is passionate about leadership (with Scandinavian influences), open-source software, modern organizational models (distributed teams), modern business models (collaborative ones), business ecosystems, and infrastructure software (cloud, databases). In his spare time, he tries to be helpful to young CEOs of high-growth startup companies and to female executives in the tech industry.
Q: What makes Scandinavian leadership different from say “Silicon Valley” style leadership?
A: I talk about Scandinavian-inspired leadership, but I must start by noting that it is nothing uniquely Scandinavian. You find it all over the world. There is a notion of “people first” and a principle of egalitarianism. We are all human beings, we should be respected for our opinions, and we all have something valuable to contribute. For practical purposes, we specialize and we are given various job titles such as “assistant”, “developer” and “CEO”. We also have a different salary and bonus levels.
But we are still just a bunch of people who are trying to accomplish something together. We speak up. It’s OK for one in a senior role to be wrong, and one in a junior role to be right. It’s OK to contradict someone and it’s OK to challenge authority. At the end of the debate, we all need to align behind a decision. But before such a decision, dissent is welcome. What’s not specifically defined as closed will be open. There are always pieces of information that need to be kept secret. But there is a ton of things that can be shared. We need to be as open as we possibly can. We must share the true state of affairs with our employees.
We must have a fully open dialog with our customers, not hiding or obscuring facts. We are not victims and we don’t put blame on others. We know that success is up to ourselves. We know that we may live in a harsh climate and that we may have to work a little harder than the rest. If we fail, we pick up ourselves, learn from the experience, and have a new go at it. We take broad responsibility. Perhaps we were hired just for one task. Perhaps we are paid a bonus just for specific accomplishments. But we all need to think about the company as something that we are collectively in charge of. For that reason, we must step up and step in whenever the situation calls for it. The CEO must be ready to answer the doorbell. The accountant must be ready to answer the phone and engage with customers. We are “us”, not “me”.
Q: What is your favorite book?
A: I love books. Reading is one of my hobbies. I have an aspiration to read 100 books per year. But given that I am running a startup company, I don’t really have the time for so many books. I’d say I am currently at one book a week.
I just finished The Zero Marginal Cost Society by Jeremy Rifkin and before that My Short Century by Lorna Arnold (1915-2014). I tried to read Energy – Myths and Realities by Vaclav Smil but I didn’t like it. I am thoroughly enjoying a long read: Empress Dowager Cixi by Jung Chang (although the book goes into too much historical detail for my taste). A week ago I finished The Courage to Create by Rollo May. A definitive highlight of the past months is The Hard Thing About The Hard Things by Ben Horowitz. From a philosophical viewpoint, books are one of the best ways to enable human minds to connect and collaborate.
One single book can change history for the better. As for favorites, I don’t have one. Some books are so good that you can read them multiple times. Some are so important to mankind that you should read them even if you may not like them. Some are so useful in your own situation that you shouldn’t leave them unread. Some have meaning at a particular point in time, some have eternal meaning. Some are poorly written but deal with a topic you absolutely want to learn about. Some are so well written that they are worth reading even if you don’t care that much about what the book is about. There are endless variations, and I believe that the human brain has a practically endless capacity to consume books. But no matter how much you read, there are always orders of magnitude more books that you will never read, many of them are wonderful books. There just isn’t enough time.
For that reason, I have become what I call a ruthless and reckless reader. If I find a paragraph in a book that doesn’t seem to bring me any value, I’ll skip it. If I find such a page, I’ll skip it. Or a set of pages. Or chapters. Or the remainder of the book. I don’t want to waste my reading time on the text that doesn’t give me something. I’d rather read another book. I also keep reading multiple books in parallel. Depending on mood, I can pick up the book that I feel most disposed towards at the moment. If some book doesn’t catch my interest in a long time, I simply give up on it. This doesn’t mean that I wouldn’t read long or difficult books.
On the contrary. Sometimes you find a book that is so densely and well written that you must read every word and every sentence In order to get it. You must be highly focused on reading it. Those books can be quite delightful. Although I said that I don’t really have a favorite book, I have created a list of useful business books. And digging into my formative years, reading GEB in the early 1980s was an amazing experience. GEB is Gödel, Escher, Bach: An Eternal Golden Braid. It was written by Douglas Hofstadter in 1979.
Q: OpenStack vs. Eucalyptus – who wins?
A: When OpenStack and Eucalyptus compete, open source wins.
In situations where OpenStack wins a deal over Eucalyptus, these are the most common reasons we hear from customers:
- OpenStack has so much momentum
- OpenStack is backed by large IT vendors
- OpenStack has a very broad feature set (current or on the roadmap)
When Eucalyptus wins a deal over OpenStack, these are the most common reasons:
- AWS compatibility
- Eucalyptus just works
- The customer doesn’t want to spend a lot of money or a lot of internal IT time
You can also look at the risks and negatives. Here is what we hear from customers when it comes to Eucalyptus risks:
- Small company – will it be around for the long term?
- Can Eucalyptus keep up with the advances in AWS services?
- Is compatibility on the IaaS API level really so important?
And here are the OpenStack risks and negatives:
- “No lock-in” is not true. It can be true only if you build your own OpenStack cloud (like eBay is doing) and for that, you need a large and dedicated IT team
- There is no meaningful public OpenStack cloud to build hybrid deployments with
- It’s a balkanized ecosystem and OpenStack maturation is slow
- Every OpenStack deployment is a complex project
The way we look at it, OpenStack is by now a project too big to fail. It will continue to consist of a loosely coupled set of cloud projects that hardware vendor will certify against. Red Hat will make OpenStack work. As they do that, they will become the main OpenStack vendor. Red Hat’s control over Linux, KVM and OpenStack will appeal to large enterprises who don’t mind paying a premium.
There is a massive business opportunity outside the area where OpenStack is aiming. AWS has a market share of over 80% (according to Gartner) and only Eucalyptus can build meaningful hybrid clouds with AWS. Thanks to its tight packaging and ease of operation, Eucalyptus is also the ideal cloud for agile and distributed environments where the customer does NOT have a giant IT team.
We win in the market segments where Eucalyptus has an unbeatable value proposition. Over time, we will integrate with the most useful OpenStack projects in order to expand our footprint in the market. We will drive commoditization of the operating system and hypervisor layer, providing superior value over any old vendor for whom the operating system or the hypervisor is the cash cow.
All of this we do with a cloud platform that is free and open-source software. We provide our tested and certified production-grade binaries to the entire world and you don’t have to pay for them (just like MongoDB, MySQL and others do). Our product is configurable and we don’t customize it. This means that you can truly create a Eucalyptus cloud that is identical to what someone else is running. In OpenStack, this is not possible. Firstly, vendors don’t give out their tested production binaries to everyone. Secondly, all main reference deployments of OpenStack are so heavily customized that it is practically impossible for someone else to get the same stuff.
All in all, we love to compete with OpenStack and we also love to integrate the best parts of it with Eucalyptus. You can already use Swift, Ceph, and RiakCS as the object store for Eucalyptus. Because it’s open-source, it’s quite possible to mix and match to create the best cloud platform for the customer.