In March 2014, most of the publicly listed high-growth technology and Software-as-a-Service (SaaS) companies started to experience large pricing corrections. Some stocks lost 25 percent in value in a matter of weeks. Some lost 50 percent. The media was reporting “Valuations Will Be Cut in Half”1 and calling it the “Twilight Zone of SaaS.” Mad Money and Squawk on the Street host Jim Cramer famously yelled, “The software-as-a-disservice to your portfolio days are upon us.”2 In Silicon Valley, people were drawing parallels to Sequoia Capital's famous “R.I.P. Good Times” milestone presentation from October 2008.3 All of the analyst's models were ripped apart. Things were messy for tech startups and SaaS companies.
At the exact same time we were in the final stages of preparing Zendesk for going public on the New York Stock Exchange. Once again, our timing was not the greatest.
But let's start with the start.
The IPO and everything that came with it were all vastly different from everything that had started less than seven years before, working in Alex's tiny loft in Copenhagen, arguing about everything, and trying to turn an idea into a reality. The fact is, most of the time my entrepreneurial adventures hadn't gone very well, or at least they hadn't gone according to plan.
I'm not just talking about Zendesk. I'm not a big fan of being called a serial entrepreneur. Call me old-fashioned, but nobody brags about how many broken marriages they have behind them. However, for better or for worse, I was no stranger to startups. For most of my working life, I never had a real corporate job. Although Copenhagen—a city of about a million people, in a country with the smallest private sector in Europe—isn't exactly the epicenter of the startup world, it was what I knew.
After I graduated from business school in the early 1990s, right when a recession hit, job opportunities were scarce. I hardly even looked for employment. Instead, I pursued something I was interested in—making things on computers—and started a small graphic desktop publishing company. At the time, I was fresh out of school, and 3D Magic Eye books, with illusions that allow you to see 3D images by focusing on 2D patterns, were the fad. I created an algorithm to create these types of 3D images or visual illusions called stereograms. I had serious headaches while working on this, but I loved being able to adjust my eyes to see the 3D element instantly. Then I turned that into a computer software program that made it easy for users to make these complex illusions.
This was my first formal foray into software, but the “software industry” was in such a different state then and had little in common with what we know today. Consumers and businesses had not begun to use the Internet, and software was kept on disks and used on desktop computers. There weren't any best practices for how to distribute software, and I was a one-man software shop figuring it out along the way. I took the packages to little computer shops to stock and sell. Customers sent me orders, and I shipped out the disks myself. I worked out of a teeny office space in downtown Copenhagen. It had low ceilings and crazy crooked floors. There were parts of the office where I couldn't stand upright. There was barely any room for me, let alone space to navigate large piles of inventory.
This effort didn't make me rich, but I didn't lose money. And although it was a terrible business in that it was very labor intensive, without much monetary payoff, I also found it was very satisfying to build a product. I loved creating something and having people use it. I was eager to read the reviews and talk to users. Most of all, I loved that I had taken something that was hard and made it easy for people.
This work led me to write a book that taught people how to create these complicated images.4 The book, which was released only in Danish, was sold in bookstores. People bought it and learned how to make these seemingly complicated images rather simply and quickly. I designed the cover for the book: it was a computer monitor with two eyeballs poking out (you know, 3D!). I rendered everything in a pirated version of Strata Studio Pro, and I scanned the iris from the cover of The Cure's Kiss Me, Kiss Me, Kiss Me album (I was a big fan of the album).
Although I was fascinated by computers and grew up using first a Lambda 8300, a ZX-81 clone with 2 KB of memory, and later an Amstrad CPC464 (with a built-in cassette tape!), no one would call Copenhagen a technology haven at that time. In the early nineties there were only metered, very expensive dial-up connections and only one internet provider. And there were very few people on the Internet.
This was the early days of Netscape Navigator, and the Web was just becoming searchable, but I certainly didn't know where the Internet was headed. However, a trip to San Francisco in 1995 made everything more clear—it was going to change everything.
In San Francisco, everything related to the Internet was being totally embraced. Everything was about the Web. Billboards advertised companies with “www” addresses. My San Franciscan friends ordered food and DVDs online. The Internet was a part of their daily lives. Everyone communicated saying, “Send me an email,” and everybody had an email address. I experienced emailing between Copenhagen and San Francisco and having these messages arrive instantly. I couldn't believe how much smaller the world was, now that it was connected by the Internet. It was completely fascinating. And I couldn't help but notice the significant gap between how technology was adopted in San Francisco and how it was integrated into our lives back at home. (It wasn't.)
Mostly what I took home from San Francisco was the sense of being at the start of something big. Today, technology is so extremely refined that it often seems to work by pure magic—you look at your iPhone and it's hard to distinguish from the supernatural—but back then it was possible to understand the web protocols and learn how to write HTML and have an idea of how to participate. You could see the path to being a creator in this new world. I felt a little bit like I had found oil on a nearby property, and I understood there was immense value in drilling it out of the ground.
There were so many things to be inspired by in the United States that I wanted to take everything I saw and bring it back with me. And that's exactly what a lot of us did.
When I returned to Denmark I was obsessed with what I had experienced in California, and I wanted to recreate it. I built a portal, a website that brought information together from multiple diverse sources—this was all the craze back then—and called it Forum.dk. Initially it was almost impossible to buy domain names for a new web service. The Danish system was extremely restrictive, requiring you to document that you owned the name already. It made everything slower and more complicated than I had anticipated. Further exacerbating matters was the fact that there were still so few people on the Internet in Denmark.
Still, we started toying around with the portal, researching how we could produce and gather content. We struck up a partnership with a source to distribute news for free. We also went outbound and created our own content: I had guys pick up flyers from dance parties and raves, and we put the events in a database so people could see where the best ones were happening. We added community elements and chat rooms (one of the first chat rooms in Denmark). The site gathered users, and although it was not our original intention, it became very popular with teenagers.
I kept a small office in the old part of Copenhagen, known for its cobblestone streets and little cafés and shops, and I worked with people who wanted to help out. They were not real employees. After all, there was no real way to pay them. It was hard to monetize this endeavor. Advertising hadn't really moved to the Internet in Denmark, so that wasn't an option.
Still, making money wasn't the main point; that was secondary. It was the potential that we were pursuing. (To deal with the more pedestrian matters of making money, I did a variety of unglamorous jobs, including waiting tables and cleaning houses. Living the dream!)
Though I had no clue how to make money off this project I loved, an opportunity came along and solved that problem for me. A newspaper offered to buy Forum. We sold off the actual service, and what we had left was the software. Now, instead of being a community portal that didn't charge anything for its services, we became a software company with products to sell.
Soon I discovered that there was a real market for this software. Companies wanted to have their own websites and provide community services for their users and visitors online. We rebuilt the entire product portfolio in J2EE, the most interesting and forward-looking software framework at the time, but also the most expensive in terms of software stack. We sold it on top of BEA WebLogic and Oracle RDBMS and primarily into large media companies or web properties, but all kinds of organizations including financial institutions became customers. Soon this remnant of our original service turned into a real business.
We named the company Caput, which means “the head” in Latin. And of course it also sounds like kaput—or “no longer working,” which was kind of cute in the beginning (kind of), but in retrospect of course it's a terrible, terrible name. But we were growing this in the halcyon days of the Internet. Things were both promising and exciting—everything was working.
At the same time, I was always working. I had no personal life whatsoever. From the time I woke up in the morning until I went to bed, all I did was work. The company was financed by a few angel investors, but it was funded mostly by selling the software. The dot-com boom that had swept the United States had now hit Copenhagen. We benefited from the hyper-frothy interest in all things Internet-related. I saw something that had the potential to become something much bigger. It was a lot of work, and it was also a lot of fun.
And then, quite suddenly, it wasn't. The dot-com crash of 2001 came as fast as the crest had. Our software was sold primarily via value-added resellers (VARs), and these companies were hit hard by the change in the tides. Media companies, a big customer base, were also brutally affected. They couldn't pay their bills, including ours. We struggled and suffered for a year as customers fell by the wayside.
Relying on sales to grow our business, and without a reserve, we had no way to cushion the impact. We had to let a lot of people go. It was really tough to have to do that to people and to go on the next day as if things were normal. I had to let go of people whom I considered very good friends—people whom I felt had built the company with me. And decisions about whom to let go relied more on where we could save most, rather than on what made the most sense for the business. Ironically, one of the über-talented guys I had to let go was David Heinemeier Hansson—who later became the father of Ruby on Rails, the framework we later built Zendesk on, and a cofounder of 37signals (now Basecamp), which became an inspiration for a whole generation of software startups, including Zendesk.
Letting people go wasn't even the worst part. There were many tough things about being in charge of the business. One time, one of our employees didn't show up for work for several days and we couldn't understand why. This young guy had struggled with rheumatoid arthritis since his early childhood, and I later learned that the disease had somehow led to a stroke and his far-too-early death. I was too young and too inexperienced to know how to properly deal with the situation. Walking into his home with the police was awful. His parents lived at the other end of Denmark, and the local police there would have to visit them to properly deliver the terrible news in person. They couldn't arrive until the evening, and in the meantime I couldn't take the parents' many, many calls. It was terrible.
We all went to his funeral. Or rather, what remained of the staff went, and it was good for all of us to get together and come together for this. But the very next day, I had to let more people go.
Things like this—terrible things—happened. But we had to go on, and we did.
There were abysmal lows, and epic highs, and ultimately unrecoverable blows. At one point, we won a huge deal in Germany that we believed could turn everything around again. We felt hope. We were so far along. The transaction was complete, but before the reseller could pay us, he went into Chapter 11—another victim of the dot-com crash.
We were out of money and out of luck.
And at this point, a year after the bubble burst, we found that the business could no longer go on. We first had to wind down our activities. Eventually, we had to shut down completely. Caput was fulfilling the prophesy of its name.
I was devastated. I had just turned thirty, and for almost four years this was the only thing that I had done. I was not depressed about losing my livelihood; I was depressed about losing the only thing I had.
In Silicon Valley there's a lot of talk about failure—there's almost a celebration of failure. People recite mantras about “failing fast,” and successful people are always ready to tell you what they learned from their failures, claiming they wouldn't be where they are today without their previous spectacular mess-ups. To me, having experienced the disappointment that comes with failure, all this cheer is a little odd.
The truth is, in my experience, failure is a terrible thing. Not being able to pay your bills is a terrible thing. Letting people go and disappointing them and their families is a terrible thing. Not delivering on your promises to customers who believed in you is a terrible thing. Sure, you learn from these ordeals, but there is nothing positive about the failure that led you there.
I learned there is an important distinction between promoting a culture that doesn't make people afraid of making and admitting mistakes, and having a culture that says failure is great. Failure is not something to be proud of.
But failure is something you can recover from.
My first failure was so tough because I had no perspective. I was ripped out of the reality I knew and thrown into a new reality. It was a reality I hadn't had a part in for many years.
But the human spirit, and its need to move forward, is amazing. After a little time, something interesting happened. I got involved in other projects. Soon I started to make money. And it was a lot easier than I had thought it would be. I had been so immersed in what I was doing that I was estranged from everything else, and anything other than what I already knew seemed scary. Now, with a few new opportunities, I slowly settled into my new post–Internet apocalypse life, and I found it was not actually the end of the world. In fact, it was the beginning of something else.
Shortly after closing Caput, I took a real job—my first real corporate job—as a general manager for a Scandinavian branch of a German consulting group called Materna. The company, which was named after its founder, had a great reputation, and one of its areas of business was help desks and customer service solutions. My job was to resell various customer service desk software products and deliver the consultants and services to properly implement the software, from a technical and business process perspective.
The German group had acquired a company in Denmark and then two additional companies, one in Finland and one in Sweden. When I joined I had to let a lot of people go and hire new people. It wasn't a terrific time for the local business.
We were at an interesting crossroads at Materna in Scandinavia. When you are trying to reboot an organization, you need certain kinds of people: athletes who can achieve many different feats and simultaneously wear multiple hats. They must have the deep intelligence necessary to solve deep problems by themselves, but they also need to be able to communicate effectively.
Morten Primdahl, once an intern and then a full-time software programmer at Caput, was a brilliant engineer who had demonstrated that he was also a great athlete, nimble and flexible. I needed his talent at Materna. The problem was that Morten wasn't so sure he wanted to do it.
“The whole setup was slightly boring,” he later told me.
I promised I'd make things happen. But I think ultimately it was something else that won him over. I offered him a company car.
I also asked Michael Hansen to join. Michael was an old friend, a big loud guy with a big heart who referred to himself as the king of Denmark. He had helped run sales at Caput, and I hired him to revive sales at Materna. Michael has a larger-than-life attitude that's hard not to fall in love with. People love to do business with Michael, and you can count on him to make things happen.
The problem was, we were selling something that was hard to make happen. The services we sold were expensive, the software we sold was expensive, and then it required expensive consultants to help deploy it and have people use it. Often the support and service reps were low on the organizational totem pole and not at all prepared to absorb or operate the software. Even with large organizations, we saw organ rejection and had to cut back on advanced functionality, one limb at a time. There was a lot more appetite to invest in technology than in organization—and service and support operations were considered first and foremost a cost center. And hardly anyone really thought about the customer experience. It was very backward.
I saw that all of this effort was more for the benefit of management—or rather, the institution—than for the people using it. In fact, no one really knew what the software did or had any expectations of what it could or should do. That's because of the backward way the whole traditional enterprise software system worked.
So, pull up to the fireplace and let grandpa tell you about the old days of enterprise software sales before Software-as-a-Service.
In the old days, people bought software without seeing it or testing it. The people in charge of purchasing, not the people who would be using the software, were in charge of deciding what to buy—and all they ever experienced was a PowerPoint pitch or a demo at a conference resort. It wasn't until software was purchased and installed and a large number of people were using it that you actually knew what it could do. And when people finally started using it, they found that it was heavy and cumbersome and often not helpful at all.
We knew the model we were pitching at Materna was incredibly ironic. Why would companies spend so much money and then ask the lowest-paid employees to deal with this giant unwieldy mess? How could you believe that putting your least sophisticated employees in front of maddeningly complicated screens, with almost no investment in user experience, could lead to great customer service?
It didn't make any sense. But at the time, customer support was a “check the box” exercise; companies spent money on it, but they didn't really mean it.
It wasn't just what we were selling; the entire industry was flawed. All of the systems on the market were terrible and complicated. It was impossible to do anything in less than six months, and usually it took as much as twelve to eighteen months. And it was very disruptive to the organization.
There was room and reason for a new generation of software. I knew this could be done in a way that was so much easier. Morten agreed. We both shared a healthy disdain for the traditional enterprise software sales model. The big platforms, the expensive salespeople with the expense accounts, the semireligious conferences, the PowerPoint software, the locked-in relationships—we despised the whole notion that an expensive sales process unnecessarily made the software both more expensive and more complicated. The whole notion that many people believed that complicated actually equals good was so disconcerting. We were motivated by the opportunity to disrupt the very core of the software model by making things easy, elegant, and inexpensive.
Morten, ever the engineer, saw it in concrete terms: “People pay a lot of money for this; we could give them 80 percent of that for 20 percent of the price and make it accessible to smaller companies, not just big ones.”
It made complete sense to us, but it was very evident that creating this kind of change within Materna would not be doable. I hadn't seen a corporate job as something I would pursue long term, and Morten didn't either. I left Materna in the summer of 2005 and started to do consulting work on my own, helping various organizations with their business and IT processes for customer service. Morten left to do security login work for banks. Later in 2005 we got together to sketch out what a new generation of customer service systems could be.
A simpler, less expensive system, delivered over the Web, could democratize software and make it easier for everyone to use and eliminate the software provider's need to hide behind complex mumbo jumbo. We were already experiencing that in the consumer world there was less distance between users and software—there people got to try the product and then decide if they wanted to buy it.
I reached out to another friend, Alexander Aghassipour, asking him to help us pursue this idea. I knew Alex from Copenhagen's startup scene, and we shared similar experiences. While I was running Caput, he had a web design agency called Araneum (which means “spider web” in Latin; we know how to pick catchy names in Denmark). When we ran our companies, we sometimes met for lunch, and we did a few projects together, but more important, we regularly hosted summer and Christmas parties together. The parties worked well because Araneum was much better funded than Caput, but Caput had a lot more women on staff.
Araneum did beautiful work and grew to be the biggest web agency in Denmark. Alex could have taken it far, but in 2001, when the bubble burst, everybody suffered. Araneum's customers no longer had the budget for its services. Although it was a healthy company, it was not very well-financed, because in Scandinavia there was not a lot of venture money or any risk-friendly money. That made it impossible to survive a market crash.
Alex, sensitive and cerebral and fragile in some ways, had to fire a bunch of people within a very short time. He took it very hard. It was difficult to have a company that was worth a lot on paper, and then suddenly was worth very little in real life, and then almost ceased to exist. In many ways his identity went with that company. Alex had a lot of anxiety about being a has-been. He worried he was never going to be successful again.
I don't think I realized any of this when I tried to win Alex over. All I knew was that Alex was super smart and had a modern approach to software and a high aesthetic. He had a great sense of details, of user interface and user experience. He was able to distill disparate ideas into something fundamental. His insight would be invaluable in building a product that people loved.
We wanted Alex's insight on this idea, but he wasn't interested, and not even because he was burned out on startups. He just didn't want to pursue this endeavor, calling it “the most boring thing ever.”
And there was that word again. The same one Morten had used. Of course, it's true that at the time making customer support and help desks better was not a very sexy idea. And although that's probably the best advice that Alex now has to offer to someone wanting to start a business—“Pick an area of an industry that is overlooked, pick something that might be perceived as boring”—it didn't make it easy to recruit him back then.
“It doesn't sound like the kind of thing I want to wake up every day and do,” said Alex, who had a pointed perspective on what was cool and what wasn't. Post-Araneum, he had created a system that allowed indie labels and musicians to sell music singles for 99 cents. (It was very much like iTunes, but pre-iTunes, and when iTunes came out they were toast.) Despite his initial complaining, cantankerous Alex soon recognized the beauty of the opportunity to revolutionize the industry with clean design—and in that he found motivation and meaning.
“I'll help keep the software beautifully simple,” he said. “We can take something terrible and have a field day and do something different and create something someone wants to use.”
Exactly. Alex started exploring the idea with us and brought an innovative concept and new approach to building a modern-day customer service desk. He understood that it rested on leveraging the Internet and cloud computing as a delivery model. But he also challenged every single assumption or paradigm that existed in the customer service desk world. He wanted every feature and capability translated into something that made sense outside of the core industry lingo. He pushed us to build not just a beautifully simple system, but ultimately a system that people enjoyed using, something they felt helped them with their customer relationships rather than obstructing them. He pushed us to build something that users would care for. Passionately.
We spent six months sketching out this product while we kept our consulting gigs on the side. It wasn't intense work. We were just toying with ideas. When we got too bogged down with work that paid, we didn't do any work on the help desk idea at all.
We initially worked in my apartment, sometimes together and sometimes apart. But we quickly moved our work to Alex's loft, simply because Alex felt more comfortable there and my budding family wasn't a distraction.
(Yes, that too had happened in 2005. I had met my future wife, Mie, at a wedding on a farm. We made out in the old cow barn. She already had a young boy from a previous relationship, and less than a year later we had our first beautiful baby girl.)
But back to work. As a team, we worked well. Morten and I were used to working a certain way—he was accustomed to being in a role in which he took instructions from me—and Alex came in and introduced a different viewpoint. It was a new dynamic. I think it was helpful for getting us to expand the way we saw things. By constantly asking questions and distilling capabilities down to their core, Alex not only helped us build a new-generation customer service system, but also ultimately helped guide us to our mission of helping organizations and their customers have a better relationship.
The three of us were very un-company-minded during this period. Our goal was to build not a big company, but a great product. This was partly influenced by our previous experiences. We didn't want to build something ambitious just to see it crumble before our eyes again. In an interview with a Danish technology blog, I proudly proclaimed that our ambition was to serve a global audience through the Internet, but to keep only a very small team on staff.
We didn't give much thought to titles, but when we had to incorporate Zendesk in Denmark, we had to appoint a CEO/president—the person formally representing the company. We sat at the table and looked at each other. None of us wanted to take on that role. We knew the responsibility and didn't really want it. I somehow ended up with the CEO title; Alex became the chief product officer, and Morten took chief technology officer. Initially it felt like I had ended up with the short end of the stick. But I'm probably also the kind of guy who likes to make stuff happen quickly, and therefore, on some level, I was not totally unhappy with the setup.
Alex was a PC guy, entrenched in the world of Microsoft. Morten was from the Apple and Java world. And they built the software entirely on Ruby on Rails, which was new, and which they taught themselves. It was a big learning experience. As Morten and Alex got more and more entrenched in coding, my role became more about being a sounding board and partner in the product management process and preparing for the product's launch and go-to-market strategy.
This is a difficult phase for many startups. And I think it was for us too. Founders' contributions can be very asymmetric in different phases of a startup's lifetime. This is especially true for bootstrapped startups like ours. Morten and Alex could work their asses off in periods when I could make very little difference. And even Alex and Morten had periods where one was way busier than the other.
The tension rarely showed its face in our case. But I have no doubt that it was latent the entire time. And we were probably lucky it didn't get to us. Appreciating and embracing the fact that people contribute different things at different times, and that everybody will have their opportunity to shine in a startup's lifetime, is easier said than done.
Working in this on-and-off manner, it took us a long time to build the product, but maybe that was a blessing. The software industry was shifting, and we felt more and more supported in our original beliefs. The provisioning of software was becoming simpler and less expensive. The trends in the consumer world were starting to sweep through enterprises. Salesforce.com was built on a new delivery model, and founder Marc Benioff was evangelizing “the End of Software” as we knew it.
37signals in Chicago was lean and small, another company blazing the trail. As mentioned earlier, it's ironic that one of its key people, Danish programmer David Heinemeier Hansson, had been let go at Caput. (It was assuredly not one of the most joyful times of his career; he had started working for us right before the issues and the problems.) Although we couldn't keep him, we knew he was incredibly talented, and it was exhilarating to see him create Ruby on Rails as a free web application framework and use it to create 37signals' first product, Basecamp (now the company name)—transforming the company from a web design firm to a software company. 37signals was almost like a religion; it established a new school focused on placing user experience above fussy features. We were disciples of this school and followed the company with great interest.
But we were in customer support—an industry in which these exciting new ideas were totally foreign. We were changing the existing offerings and industry on a philosophical level. We used terms like “love your help desk” and “tender loving customer care.” No one else was putting those things together. Customer support software was an option only for very large and very established companies with deep pockets. And it was supposed to be boring! They were all the same and marketed with stock images of people with headsets. (Stock image banks: where photographers go to die.) We were doing something different. But of course we had no idea whether it would work.
However here's the crazy thing. Despite the hours we put in, the money we invested, the expectations we had, we didn't set out to build a company. We were driven by a desire to build a product that was lean and self-explanatory and offered a new option. We wanted to create something people loved to use. That was our aspiration.
The one thing I learned from the little games I made on my home computer when I was twelve; or from building the stereogram software when I graduated business school; or from building Forum.dk; or from my time at Caput, building community sites, was that there was nothing so satisfying as building something that people used. We all have the urge to build and create something. And the amazing thing about the Web was that there was the potential for it to reach a global audience—for it to make a big difference.
In the earliest months—those days in my apartment, the initial time in Alex's loft—it was all about innovation and inspiration. Everything was a blank slate. Anything was possible. It was a pretty great time.
But soon the honeymoon ends. The idea becomes a reality, and you discover that even though you never really wanted to build a company, all of a sudden you are doing just that. And the reality is that building a company is hard—it's no longer about just pursuing an idea. The idea has taken off, and it becomes all about executing and scaling.
Before you know it, you have all these real customers who need real attention, and you need to start building the beginnings of an organization. With all the bureaucracy and management that comes with it. And the realities of your own life creep in. No one tells you in advance how little you get paid to get there. How in the early days you are constantly running out of money. How much credit card debt you accumulate. How many fights you have with your cofounders. How you have to live with their weaknesses and oddities. How there's the constant temptation of flashy objects and shiny new things. How much you keep from your spouse. How you overleverage your life. It quickly becomes very complicated.
We didn't know anything about this. We didn't know anything about anything.