Shortly, even the CEO will be outsourced to an online labor marketplace

Danny Crichton
Shortly, even the CEO will be outsourced to an online labor marketplace
Shortly, even the CEO will be outsourced to an online labor marketplace

Over the past decade, there has been a ferocious rise in the freelance economy in the United States. Millions of people today work on platforms ranging from Uber and Lyft to Taskrabbit and Fiverr, accepting what are usually short-term tasks that can be completed efficiently and repeatedly. While these casual jobs have been the focus of intense scrutiny about their pay structures and work security, their effects have been mostly limited to talent not engaged in business professions.

Times are quickly changing, though, and marketplaces are increasingly entering white collar territory with better product design.

Take for example Clora, which works with the highly-specialized talent required to produce and launch a new life sciences product. Or Paro, which connects businesses to bookkeepers and other financial professionals to manage a company’s finance department. Or Catalant (formerly HourlyNerd), which works with independent business consultants who can solve a range of complex problems like market sizing or product marketing.

It might be an exaggeration, but it is only a matter of time until rent-a-CEO options exist as well.

Part of this transformation of white collar work certainly comes from companies and workers desiring more flexible work arrangements. However, I would also argue though that renewed focus on product design has been critical for effectively building a marketplace for online talent.

Different tasks often have wildly different requirements and workflows, and the product needs to match the kinds of talent it hopes to attract. These newer marketplaces understand that professional work is often ambiguous and hard to judge for quality, and have built key product features to handle those challenges.

That’s very different from the early years of the consumer internet, when online labor marketplaces were designed as free-for-alls, with both sides of the market competing for transactions to occur. A customer might post a potential job to Craigslist, or a worker might post their availability to be hired. These were marketplaces built around serendipity, with almost no guidance from the platform on what to charge, how to charge, or how to find the best talent for a particular project.

Over time, it became clear that some tasks were much more popular than others, and they were quite repeatable as well. Uber takes a passenger from one origin to one destination, and Taskrabbit allows you to hire someone to install IKEA furniture. You don’t need to use paragraphs to describe what you are looking to do, nor should you have to negotiate a price every single time you want to get into a car or get a MALM bed frame installed.

That regularization became the product itself — suddenly the free-for-all marketplace became a tight menu of options with standardized pricing and rating systems. Even more importantly, the identities of the workers themselves are often shielded from the customer on these platforms. You are buying the company’s brand of quality, not the worker’s guarantee that they can do the job.

There is a marketplace today for pretty much every simple and easy-to-define task imaginable. The challenge has been how to design marketplaces to handle more complex forms of work and evaluate the quality of that labor, particularly in cases where quality can be in the eye of the beholder.

One answer has been to focus on hyper-specific (yet lucrative) verticals. Clora and Paro are good examples of this trend. Clora’s strength is knowing the hiring model for the pharmaceutical and life sciences industry. They understand what talent is needed, and even more importantly, how to evaluate that talent. By focusing on accounting, Paro has a relatively objective standard on what quality work looks like, and that makes evaluation a bit easier.

The key to each of these platforms is that pricing, ratings, quality, and product offerings are not standardized in the same way as the productized marketplaces that came before. Each of these platforms recognizes that, say, the launch of a cancer drug is going to be very different from a male-pattern baldness therapeutic. There is no secret algorithm that is going to be able to detect these nuanced differences without human judgment entering the equation.

So to compensate for that variability, these marketplaces put much more of the judgment around the quality of work on the shoulders of the professionals themselves. Unlike Uber, there is no GPS telling someone to move from this work to that work. As such, these newer marketplaces are something of a hybrid between the first-generation free-for-all Craigslist-style marketplaces and the second-generation productized ones like Uber we have seen more recently.

B12, a company I profiled recently, is trying to build a unified infrastructure for handling exactly these sorts of ambiguous jobs. Its open-source Orchestra platform is designed to handle not just the non-linear work patterns that arise in these contexts, but also how to judge quality on the platform (it’s answer is to have professionals evaluate other professionals). It’s an audacious idea, although I still believe we will see many verticalized marketplaces since sales and marketing into these each of these industries is challenging.

When we really evaluate what is happening around work, particularly professional work, there can be incredible complications that make machine learning and algorithms hard to execute. Product designers need to include the ambiguity, complexity, and social human factors into labor marketplaces to properly attack these industries. Thankfully, we are seeing a new crop of startups do exactly that, and that will not just provide opportunities for founders and VCs to make a bunch of money, but potentially millions of more workers as well.

What to read next

By using Yahoo you agree that Yahoo and partners may use Cookies for personalisation and other purposes