This past Saturday, a group of fans and aspiring creatives of all kinds gathered together in the basement level of Silicon Valley Comic Con. Outside this room, the convention bustles noisily with costumed characters, video games, and the murmur of human voices. Inside it is a rather drab and gray setting for one of the most prolific and influential creators in the comic book industry, who incidentally, is running just a few minutes late.
Michael Golden is probably best known for creating the X-men character “Rogue” while at Marvel, but his comic works also include “The ‘Nam,” “Micronauts”, “G.I. Joe Yearbook”, “Dr. Strange”, and much more. His diverse commercial portfolio ranges from Nascar to Nasa, to Universal and Warner Brothers. In a few moments, the man himself approaches the stage. I suddenly recall that a wizard is never late, nor is he early, he arrives precisely when he means to.
“Rule number one is: people are stupid,” Golden says in his deep voice that is both gravel and soft butter all at once. This is not a man that wastes a lot of time in coming to his point. He is older, gray-bearded with glasses, and firmly clutching a Starbucks cup in his right hand. He smiles knowing that he has hooked us with surprise and continues, “Whoever you are trying to sell is ignorant of your story, and it’s up to you to give them the information. Make it dramatic, concise, involving, and understandable. People get bored easily, and they don’t come back.” There are only two explicit rules for success, apparently, and the second is a rule of no rules. “You hook them by sticking to who, what, when, where, why, how.”
At the age of 65, with a career spanning over 40 years, Golden’s longevity in this industry is unusual. “This industry will chew you up and spit you out,” he says. He credits his ability to stay ahead of the twenty-somethings to his discipline to keep learning and stay productive. “With commercial work, you no longer have the option to be in the mood.” A typical work day starts at 4:00 AM and ends at 8:00 PM. Golden forces himself to restrict creative work to the early part of the day. “By around 2:00 I really start to slow down. By 8:00 PM I am done, tired, wasted, and needing a drink.” He used to take Saturday and Sunday off without fail, but thanks to the Internet, Golden now works seven days a week.
When asked how he learns and draws inspiration, Golden says that he learns by doing, but he admonishes the audience to do as he says and not as he does. “Learn everything you can, go to school. Learn design, learn layouts, and learn color theory. There will be plenty of time later for Photoshop and Illustrator. Technology is just a tool; it can’t make you great.”
Although he does not say this, I start to realize there is something else going on here too. It is evident to me that, from his full work schedule to his penchant for learning and problem-solving, Golden is immersed in his work. You begin to get the sense that his favorite part of the day is in the pure flow of the creative process. This seems like it could be the real secret — being in love with your work — even after years and years.
Of course, being uniquely talented and globally recognized probably helps in the love-your-work department, and it may even account for some of Golden’s lack of burnout. But you can hear the underlying truth tinged everywhere in Golden’s story. Talent is only part of the equation. Success came after a lot of sweat, dedication, discipline, hard work, and long hours. “There is no formula!”, says Golden
Developer interest in C# has declined sharply since its peak in 2012 according to data from the TIOBE index. As one former .NET developer wrote back in 2015, this precipitous decline in due to the rise of cloud-native application frameworks and mobile platforms where Microsoft developers are not first-class citizens.
The reach of Microsoft’s developer ecosystem has declined in the past five years due to the rise of non-Microsoft web frameworks and mobile platforms. Android and iOS control 90% of the world wide smartphone market and .net developers aren’t first class citizens on those platforms. – Justin Angel, former .NET developer
Microsoft’s mobile device market share slipped below 1% again earlier this year. Yet there are growing signs that Microsoft’s long ball strategy of adopting open source and doubling down on developer tools may be starting to pay off. Last year Microsoft surprised everyone when it announced a partnership with longtime rival Red Hat to deliver Linux on the Microsoft Azure cloud computing service. Earlier this year Microsoft acquired Xamarin, which helps .NET developers build mobile apps for Android and IOS.
Microsoft seems to be listening to IT leaders, who have been under tremendous pressure to create agility for the enterprise application developers who are driving innovation. The move to open source .NET Foundation and the creation of Windows Server Nano is designed to put a stop to enterprises re-writing .NET applications in order to migrate to Linux.
Microsoft has also moved quickly to adopt Docker and containerization with native support for Docker containers in Windows Server 2016 and recently hiring Google’s lead developer on Kubernetes. The widespread adoption of containers is a boon for Microsoft because it enables developers to easily move applications from one environment to another. As developers increasingly use container-centric tools they become less reliant on the management interfaces of cloud rival Amazon Web Services and can more easily move applications to Azure. At the annual Ignite conference in September 2016 Microsoft showed a new demo of Docker in Visual Studio.
The results seem to suggest that Microsoft’s strategy is working. According to a recent article data from Synergy Research Group puts Azure cloud growth at 100% year-over-year, making Azure the #2 pubic cloud behind Amazon AWS. In 2017 enterprises with large investments in .NET may find these advances very compelling as they continue to seek a viable cloud strategy. Amazon’s lead in the public cloud has been so far been insurmountable. Do my eyes deceive me or does the most recent data from Indeed show a modest increase in C# job postings over the last 12 months? In 2017 we may finally see a real contest.
Indeed.com C# Job Postings
As I was reading the recent article by preeminent cloud expert Bernard Golden predicting the long, slow demise of private cloud, I was reminded of the “Bring out your dead” scene in Monty Python’s Holy Grail. In that scene, the old man who is still very much alive objects to being placed on the death cart, before being summarily dispatched to by a sudden blow to the head. Bernard makes a great case for why public cloud computing has been and will continue to be so successful. Indeed in my lifetime I do not think that I have seen a company the size of Amazon growing as fast as AWS is growing now. Yet lifetimes are on a relative scale.
On a long enough time scale, death and immortality are indistinguishable for the purposes of human beings. For example, it is perfectly accurate to say that the sun is also dying. Our star, the source of all life on our planet, will ultimately exhaust its supply of hydrogen. In about 5 billion years, the sun will ultimately grow so large as to engulf the Earth, totally destroying the planet in the process. While we know for a fact the sun is dying on a very long timescale, its not a good use of our time to plan for this extinction event. Much can happen in 5 billion years that is unknown. We may not even be here when the time comes. Or we may not even be exactly human anymore.
Let’s not equate private cloud with private infrastructure because these are not the same thing. Technology has become so embedded in the lives of people and companies that we are not even sure when we are using it anymore. There is privately owned infrastructure literally everywhere, and the trend is for more, not less of it. In the early days of the Internet, we used to joke that it was only a matter of time before your toaster had an IP address. The so-called Internet of Things, the proliferation of internet connected devices, is occurring on a scale that is almost unimaginable. The need to manage these legions might easily lead to a kind of data center renaissance. I would be shocked, by the way, if those data centers didn’t look suspiciously like private clouds.
I wish to offer you an alternate prediction, and it is this: The death of privately owned infrastructure, if it happens, will be so far in the future that your great-great-great grandchildren will still be running data centers of some kind. Private cloud is not dying or even sick – it is undead. It will go on living indefinitely seemingly unaware of its deteriorated state. What is dying is not private cloud or private infrastructure, but IT itself. For the last 25 years IT was a kind of factory converting the raw materials of technology into services consumed by the business. The fact is that IT was an inefficient factory that required a lot of manual effort. The era of custom built IT is being replaced by utility models. Public cloud is one such utility model, but there are others emerging even as we speak. What remains of IT is leaving the infrastructure era and entering the application era. To repeat a catchy phrase from a Microsoft executive, “Cloud is a model not a place”.
The evolution of the electricity-generation industry is an often used as an analogy for what is happening in IT with public cloud. I myself have used this analogy on many, many occasions since reading Carr’s prophetic “Does IT Matter?”. Yet the consumption of electricity and of computing are not close cousins. People have a great many preferences around how they consume computing but the same is not true of electricity. Some people value the privacy of their data while others value performance or the ability to scale. Public cloud is well suited to the consumer preferences of the moment, but that is subject to change. For most of us, electricity is just electricity, but even it is not immune. Just look at the impact solar and other alternative energy.
We haven’t even talked about the Grim Reaper of Public Cloud. When the Reaper comes he will come in the form of a catastrophic outage of one or more massive public cloud providers. Whether this outage is created by a security compromise, human error, or a bug or other technological failure, the result will be drastic changes in consumer preferences. We will all collectively learn the meaning of “too big to fail”. When the Reaper finally exposes his face it will be the Federal Government. Congress, greatly annoyed by their depravation of cat videos, will quickly move to regulate public clouds to oblivion. The party will be over when Amazon starts to look a lot more like AT&T.
There may just be life in that old data center yet.
Building marketplaces is hard, just ask Bill Gurley of Benchmark about the complexity of identifying good opportunities for marketplaces in the Internet economy. This was further evidenced today when I received a little email hand grenade from the CEO of Upwork, a leading online marketplace for freelancing.
In the wake of the ODesk-Elance merger, Upwork is inflicting a higher “Upwork tax” on customers and freelancers for the overwhelming majority of jobs, and is also levying a new financial processing fee. At the same time, Upwork is also lowering fees on higher dollar transactions, and introducing a processing fee.
Here is how Upwork CEO Stephane Kasriel’s explained these changes:
We help you build your business by acquiring clients, helping you connect with the right opportunities, and providing services like payment protection. On small projects, the costs we incur outweigh the fees charged; because they aren’t profitable, we haven’t been investing in growing the number of these projects. At the same time, client relationships that result in larger, repeat projects incur fewer of these costs because of the trust that’s been developed, and we want to pass those cost savings back to our users. – Upwork CEO Stephane Kasriel
I did have a good laugh at that. For those of you who not well versed in deciphering the charismatic double speak of startup CEOs, please allow me to translate: You need us. We see no reason whatsoever why Fiverr should be allowed to make 20% of client money while we make only 10% for same size transactions. Therefore we are increasing our prices because, frankly, we can. We are under pressure to make more money. This is a carefully calculated risk we think we can take because the market for freelance work is growing faster than we can count dollar bills.
What I found interesting is that this change highlights the fragility and complexity of marketplace revenue models. Marketplaces like Upwork are fundamentally middlemen. Middlemen exist in markets where they lower transaction costs for buyers and sellers. More specifically, middlemen exist when their fees are lower than the corresponding costs of buyers and sellers who transact with one another directly. These models begin to break down when their fees approach direct transaction costs. If you are a middleman, determining the true direct transaction costs between buyers and sellers is pretty important, since it determines in part whether or not buyers and sellers transact business with you instead of each other.
So what are the direct costs to freelancers and their clients, and how does Upwork know that the low end market will bear these higher fees? I am sure there is one hell of an Excel model at Upwork HQ, but one clue is the pricing of competitors. Upwork must believe it can charge as much as its competitors in the low end of the market. That might be a dubious assumption. Competitors like Fiverr have optimized the user experience for the low end transactions. These jobs are just harder to do on Upwork, period. By raising its prices in the low end of the market, Upwork opened itself to any competitor willing to step in and deliver a slightly better experience at Upwork’s old price. Why lower any barrier to entry at this stage of the game?
Which brings us to the really big strategic play. Upwork is betting that its future is to be the premier marketplace for large freelance projects. Obviously, the cries from larger clients that Upwork’s 10% fee was already too expensive did not fall on deaf ears. By lowering its fees to 5%, it believes that it can further disrupt traditional suppliers of short term talent. These include recruiters, temp agencies, and small consulting firms. Temp agencies have profit margins ranging from 1.4% to 8.8%, with the industry average being 5% according to the latest report from Barclays. Upwork is betting that it can disintermediate these existing middlemen by further reducing friction for buyers and sellers. If this strategy is successful, then Upwork probably secures its exit through one of these suffering middlemen. Pretty smart, and pretty risky. I like it.
I am a big fan of freelance marketplaces, having sourced many thousands of dollars of work on projects ranging from software development, writing, advertising, and even illustration for a children’s book. Given the usability friction for low end jobs on Upwork, I’ll probably just use something else for these jobs from now on. However, now I see no reason not to float much bigger contracts to Upwork in the near future now that the fee is 5%. I get the feeling that this is exactly what Stephane Kasriel would want me to do.
When I first started my business in the fall of 2008, I was faced with a classic founders dilemma. What type of entity should I incorporate, LLC, C-Corp, S-Corp? There are hundreds and thousands of articles with mediocre advice regarding the various tradeoffs of these entities. In this article I will share some real world experience with you creating my company 2008, and how our entity decision impacted us much further down the line. Before I go further, understand that I am not an attorney, and I am not licensed or qualified to give legal advice. Please consult with your attorney before making decisions like this one.
On day one, I chose a Delaware limited liability company. The incorporation was fast, inexpensive, and easy. As the business grew, I brought on a partner to be my CEO. We hired some of the best legal counsel in Silicon Valley, who quickly informed us that we would want to convert to a C-Corp to formalize our partnership and raise institutional money. It was explained to us that this was the preferred entity type for institutional investors for a lot of different reasons, but chief among them was the tax structure of the various VC funds. This was good advice, but I wish we had not followed it.
When we converted the LLC to a corporation, we promptly set about raising funds through a convertible note. Once we had money in the company, we quickly got down and dirty with building our new business. Ultimately we were acquired by another startup in 2011. This is where our choice of entity selection began to bite us. Our acquirer wanted to purchase our assets and not our liabilities. We entered into a purchase agreement where the acquirer gained our assets in exchange for preferred warrants. That meant that our company would go on in perpetuity with no assets other than the warrants.
One of the purported advantages of C-Corps is that their shareholder rights and governance are defined by statue, where as member rights and governance of an LLC are defined by an explicit agreement between the members. The problem with a C-Corp comes when, for whatever reason, you are unable to afford legal counsel to advise you on the statutes in question. Another problem is that C-Corps come with a lot of formalities with respect to keeping records, board and shareholder meetings, and other governance issues. These are all things that you need to stay on top regularly to ensure that your entity is in good standing and your liability is limited. But what happens when the music stops?
Now we had our C-Corp with no employees, no cash, and with no purpose except to serve as the holder of those preferred warrants for ourselves and our investors. We ended up making a subchapter S election to give pass-through treatment of our profits and losses. The accounting required for such subchapter S-Corps has many more complicated than that of an LLC. And, for two years we had to keep going right along with shareholder meetings, keeping minutes, and making resolutions. For two more years, I had to keep filing corporate tax returns, and keeping the entity in good standing with the State on the chance that we might need to liquidate the warrants and distribute the proceeds. Fortunately our acquirer was also acquired within two years. But what if I had to keep going with that process indefinitely?
What a nightmare.
In hindsight, we would have been much better off by keeping the LLC structure and amending our operating agreement as needed. The LLC afforded us a very simple and flexible structure, one that was easy to understand, and one that was very cheap to maintain. The costs and formalities for maintaining and governing LLCs are extremely limited. We could have delayed the C-Corp conversion to the point where we had a committed institutional investor who was demanding it. Or we might have been able to make an IRS election to be treated as s a corporation.
We went through the whole entity lifecycle, from LLC, to C-Corp, to subchapter S-Corp with one and the same business. In other words, we ended up back where we started, but with a lot of added complexity and expense. Seven years later investors and their counsels seem much more comfortable with the LLC entity. There are plenty of good boilerplate agreements you can modify suit your needs.
For this founder, the next entity decision will be easy.