What Upwork Got Right—and What Upwork Got Wrong

Many years ago, I met a former leader of oDesk, the company that would eventually merge with Elance and become Upwork. He offered to spend some time helping me understand the different players in the emerging online talent market. In exchange, he asked me to buy him some mountain bike tires. I bought the tires, and I still remember that interaction because he had already accomplished far more than I had and had no particular reason to spend his time helping me. 

Around the same period, I got to know some of the guys at Fiverr, who were also great. Looking back, I think oDesk, Elance, and Fiverr deserve a tremendous amount of credit for creating the market that companies like mine would eventually enter. Before Toptal, Andela, Arc, Lemon.io, Gun.io, and the rest of us spent our time thinking about how to find the best talent on the internet, these companies had to establish that hiring people through the internet made sense at all. Fiverr, for example, describes its original idea as making digital services purchasable in much the same way physical goods were purchased through e-commerce. Today, we take that idea almost completely for granted, but it was a genuinely different way of organizing professional labor.

For most of modern business history, geography placed a practical boundary around the labor market. If you wanted to hire people in another country, you generally needed some kind of infrastructure there, whether you built it yourself or paid someone else to provide it. Recruiting, contracts, payroll, compliance, payments, and management all became more complicated once you crossed a border. Large companies could solve those problems by becoming multinational organizations, while smaller companies generally operated within a much narrower labor market or relied on traditional staffing and professional services firms to extend their reach.

The internet made communication across borders practically free, but communication alone did not create a functioning labor market. Somebody still had to make it possible for two strangers in different parts of the world to find each other, establish enough trust to work together, exchange money, and build a reputation over time. That is what the second generation of services firms did. oDesk itself described its offering as allowing companies to hire, manage, and pay technical talent around the world, with a goal of work becoming independent of place. These companies gave a person sitting alone with a laptop some of the labor-market reach that had previously belonged to a multinational corporation. A founder in Nashville could find a designer in Eastern Europe, an engineer in India, or a specialist in South America and begin working with that person without opening an office there or building an international recruiting organization. It was pretty rad, and it changed the world.

It is easy to misunderstand what these companies originally accomplished because the market has evolved so much since then. Upwork today describes itself in terms of on-demand access to highly skilled global talent, but the historical breakthrough was broader access itself. When Elance and oDesk announced their merger in 2013, the combined companies already reported more than eight million freelancers and two million businesses across more than 180 countries. When the combined company relaunched as Upwork in 2015, it explicitly described the shift as opening businesses to an “earth-sized” talent pool unconstrained by traditional local hiring. The goal was access. Upwork made an enormous portion of the global labor market searchable and transact-able for companies that previously had no practical way to reach it.

That was an extraordinary accomplishment. Workers gained access to companies and opportunities they could never have reached locally, while companies gained access to people they would never have found through their existing networks or recruiting infrastructure. In a very real sense, these companies helped make it possible to be a multinational corporation with one employee, one laptop, and an internet connection. They also gave companies seeking talent considerably more control than the traditional services model. Rather than hiring a firm and allowing that firm to decide who would perform the work, you could increasingly compare alternatives, select the precise person you wanted, negotiate a price, and manage that person directly.

For these platforms, the trade-off was that all of this control came with responsibility. Once hundreds of people could respond to a job posting, finding the right person became a big challenge. You had to determine which people were credible, which were actually good, and which one you should trust with something important. The same abundance created a problem for talent. When workers from every labor market in the world compete next to one another, price becomes one of the easiest things for a buyer to compare even when capability is much harder to evaluate. Sometimes that is perfectly rational: there is no reason to pay an American engineer $150 an hour for work that an equally capable engineer somewhere else can do for $50. But not all engineering work is interchangeable, and a system that makes price immediately legible without making quality equally legible can make life harder for exceptional talent as well as buyers.

That was the interesting consequence of what Upwork got right. It effectively solved access, but the next problem was curation: determining who was actually good, matching the right person to the right work, and removing some of the complexity that the buyer had inherited along with all of that new control. By bypassing the old-school professional services or staffing firms with a broad global marketplace, companies found themselves suddenly missing the expertise inside those firms when it came to vetting, validating, and qualifying talent.

And if there is one constant in the internet age: when one market shift solves a problem while creating another, a new market will rise up to meet that challenge. You can see that response explicitly in the companies that followed in the latest generation of services firms, of which Gun.io is a member and leader: Andela evaluates technologists before matching them with companies; Arc vets for communication and domain expertise; and Lemon.io makes vetting central to its proposition.

For more reading on this Third Generation of staffing firms, read here.