A feature allowing freelancers to put their own money behind a job bid made employers less likely to hire them on the online labor platform studied by researchers. The result shows how a trust-building feature can acquire a meaning its designers did not intend.
The research, published online in Information Systems Research on September 22 and discussed by Penn State on October 20, examined a voluntary worker-offered guarantee. It was a financial commitment by the freelancer, but it did not reimburse the customer if the work went wrong.
For businesses that buy freelance services or operate marketplaces, the question is whether an added promise provides new evidence of ability or simply draws attention to the need for reassurance.
The deposit did not protect the buyer
According to Penn State’s interview with co-author Chenhui Guo, the platform suggested a deposit equal to 10% of a bid. The money was returned if the freelancer was not hired or completed the job successfully.
If a hired worker failed to complete the job satisfactorily, the deposit went to charity. The employer did not receive it.
A promise that costs the seller money is not necessarily insurance for the buyer. In this arrangement, the employer could still lose time and face the expense of finding someone else to finish the work.
Guo reported that workers using the guarantee received lower ratings, were hired again less often and took longer to complete their work. Those comparisons concern the workers who selected the feature. They should not be read as proof that placing a deposit made a worker less capable.
Employers interpreted the extra promise
The journal abstract describes analysis of detailed bid-level data from a leading freelance platform. Qiang Gao, Mingfeng Lin, Yong Liu and Chenhui Guo conducted the study.
Workers with medium reputations were more likely to use the guarantee than either high- or low-reputation workers. Employers discounted bids carrying it, although the negative response was weaker for high-reputation workers, more complex projects and experienced employers.
The authors describe this response as counter-screening: buyers used the decision to offer a guarantee as another clue when judging a worker. The added commitment did not erase what they already knew from the worker’s record.
The research draws on signaling, the use of visible evidence to communicate qualities that another party cannot directly observe. In hiring, a completed project or a review can help a buyer assess a person they have never employed.
A new signal enters that existing system of expectations. Its meaning depends partly on who adopts it and what the buyer believes adoption says about the seller.
Marketplace design needs outcome measures
The result concerns one noncompensatory guarantee. A refund backed by an enforceable contract, an independently checked qualification and a voluntary deposit donated to charity have different economic consequences. The study cannot be applied automatically to all badges, guarantees or freelance platforms.
For a marketplace introducing a feature, uptake alone would be an incomplete measure of success. A feature can attract users while making their bids less competitive. A fuller evaluation would examine hiring, completion, repeat business and the effects on different groups of workers.
For buyers, it is worth separating a commitment from the evidence behind it. A promise can clarify a supplier’s responsibilities, but it does not establish that the supplier has performed comparable work.
Freelancers face the corresponding decision about which evidence to present. Our guide to writing a clear elevator pitch examines how relevant proof can make a proposal more credible.
The deposit in this study exposed the worker to a financial loss. The employer still had to decide whether that willingness to pay said anything reliable about the quality of the work.
