Finding the best freelancer: Openness helps

Some 34 million people in the U.S. — approximately one in three workers — now work freelance, according to one national survey conducted by research firm Edelman Berland in partnership with Upwork, the company that used to be Elance-oDesk.

Some 34 million people in the U.S. — approximately one in three workers — now work freelance, according to one national survey conducted by research firm Edelman Berland in partnership with Upwork, the company that used to be Elance-oDesk. Of those, 40 percent are independent contractors, 27 percent are moonlighters, 10 percent are temps and 18 percent hold multiple jobs. The other 5 percent see themselves as business owners, the kind of people who might pull together a crew of writers and designers to serve as a virtual ad agency on a project. All of these types of freelancers include people you might find online through hiring platforms like Upwork or Freelancer.com.

In such labor markets, you post your project with a brief description, your requirements, your budget and a call for bids. It’s a reverse auction: The winning freelancer gets the contract, while the employer puts out the dough. And, in such auctions, employers have the option of inviting sealed bids or open ones. It’s a choice you can make in the auction design. Open versus sealed bid auctions, which will bring you the best results? It is a classic question in economics and information systems, for which researchers aspire to seek an answer.

According to Yili (Kevin) Hong, a professor of Information Systems thinks you’ll probably do best with an open-bid approach for reverse auctions in online labor markets. His research “Comparing Open and Sealed Bid Auctions: Evidence from Online Labor Markets”, forthcoming in a premiere journal — Information Systems Research — shows that open-bid auctions will deliver fewer bidders but better deals and higher buyer satisfaction in the end by reducing freelancers’ valuation uncertainty.

Age of uncertainty

Hong has been studying the online labor marketplaces like Upwork and Freelancer.com since the early days of such employment venues: 2009. That’s when he made this phenomenon the subject of the award-winning dissertation he wrote to earn his Ph.D. Today, these markets are filled with promise for freelancers and those hiring alike.

Freelancer.com says it has nearly 17 million registered users and, according to Hong, the ratio of service sellers to employers is probably around three to one, so the sites give employers a great way to find low-cost help quickly. The site boasts that some 8.7 million projects have been awarded to lucky freelancers. At least, we hope those project winners were lucky, but maybe they weren’t. Things like IT projects can be trickier than they seem, in particular, cost estimation. “Even an experienced developer may not know how much work a project will take,” Hong says. This uncertainty is termed “valuation uncertainty,” because it reflects several questions:

• How much is the job worth?
• How much time will it take to complete?
• How much will the employer be willing to pay for the work?

Another area that prompts questions is competition: Do bidders know who else is bidding? Can they tell how much the other freelancers want for the work? Termed “competition uncertainty,” this ambiguity reigns when bidders are forced to participate in sealed-bid auctions versus those with an open-bid design. In light of these two forces — valuation uncertainty and competition uncertainty ­— few empirical research has been conducted prior to Hong’s to see which auction design would produce a higher “buyer surplus.” “Buyer surplus is mathematically measured as the price you’re willing to pay minus how much you actually paid,” Hong explains. In other words, it’s how good a deal the buyer got. And, while several researchers examined bid visibility as a factor in how buyers attain value, the findings were inconclusive.

So, Hong conducted an empirical study to see which auction approach would get employers the best possible outcome. Having helped one of the major online labor markets redesign its user interface, working with a leading online labor marketplace, Hong was able to acquire all the data on more than 71,000 open-bid auctions and nearly 7,500 sealed ones. These were posted by some 22,000 buyers, who paid an extra dollar to get the sealed-bid auction design. Results yielded a few surprises. While sealed-bid auctions averaged 18 percent more bids, they didn’t result in better results for buyers. Several measures indicate that open bids were actually more effective.

Why more isn’t better

Hong thinks that sealed-bid auctions inspired more people to bid because they eliminated the intimidation factor of knowing your competition. Bidders may decide not to bid at all if they see competitors with far more experience and knowledge bidding on the job, he explains. Along with eliminating competition uncertainty, open-bid auctions eliminate valuation uncertainty, too. And, there’s more to it than the transparency of each bidder’s bid. Uncertainty wans because bidders assume competitors are either drawing on experience or they’ve done some of their own research to decide on a price. “It becomes a collective price-discovery mechanism,” Hong says. When bidders feel comfortable with a value of the project (when valuation uncertainty is mitigated), they are more likely to offer a lower price to compensate that reduced valuation uncertainty.

On average, open bid auctions delivered an average of $11 higher surplus in Hong’s research. In addition, they were 55.3 percent more likely to result in a buyer selecting one of the bidders for the project. That choice, however, doesn’t guarantee a contract. Bidders can still walk away from a job. Nevertheless, the open-bid auctions were 22.1 percent more likely to result in a contract with a bidder than the sealed-bid auctions. And, along with getting employers better deals, the open-bid auctions also resulted in higher buyer satisfaction, as measured by the ratings service providers and employers gave each other after the project was completed.

Not only did the open-bid auctions yield more productive results and satisfaction, there’s also a problem with the one area in which sealed-bid auctions outperformed open ones: the number of bids received. It turns out that more bids are not necessarily better, particularly now that these online labor markets have matured. “During the period we observed, buyers were getting around 15 bids per project. Currently, because the market is developing so quickly, I would say companies are getting 30 to 50 bids on a project,” Hong notes.

While he didn’t precisely evaluate this angle in his research, he does point out that, “If you get many bidders, you have to spend more time evaluating those bids.” What’s more, sealed-bid auctions bring in a greater diversity of prices and higher prices to boot. Hong thinks this may be a reason such bids are less likely to lead to an actual engagement. If everyone pretty much tosses in the same bid, Hong still sees that as a perk for the buyer. “It’s good for the company because they also have less valuation uncertainty. Then the company can evaluate candidates on ratings, experience in the market and other quality signals,” he explains. One signal that comes into play is location.

Concurrent with this study, in another paper, Hong also looked at whether employers discriminate based on where bidders live and, it turns out, they do. “If the freelancers have the same prices, buyers are more likely to hire someone from rich countries than poor ones,” he says, adding that there seems to be an assumption that people from richer countries have better education and credentials than those in lower-income locales. Similarly, when prices are similar, buyers are more likely to hire people who speak the same native language that they speak, as well as those who live in the time zone closest to their own, suggesting that the Internet hasn’t completely succeeded in creating a flat world, one where commerce occurs on a level playing field. But, the Internet has opened up a wealth of opportunities for employers and workers alike.

Hong estimates that online labor markets are doubling in size each year. Contrast that with a projected 3.1 percent job growth in Arizona, the state credited with the fastest job growth in the U.S. That’s what recently Moody’s Analytics recently forecast Arizona job growth would be through 2019. According to research by Intuit, the contingent workforce in the U.S. has grown from 17 percent of workers in 1990 to 36 percent today. By 2020, it’s expected to reach 43 percent. So, there’s a good chance online labor markets are in your future. If you’re buying services, stick with open-bid auctions. It’s a superior auction design, Hong says, when “you want fewer bids with better prices.”