by Sameer Hasija, V. “Paddy” Padmanabhan, and Prashant Rampal
The term “gig economy” was coined by the former New Yorker editor Tina Brown in 2009. It described how workers in the knowledge economy increasingly were pursuing “a bunch of free-floating projects, consultancies, and part-time bits and pieces while they transacted in a digital marketplace.”
The received wisdom of the time was that the gig economy would redefine white-collar jobs and call into question the very existence of professional service firms: Why would you need to hire a data analytics firm for a project when you could have unrestricted access to a bunch of experts, connected by a digital platform with global reach, who could work together for you? For a time, it certainly looked like things were headed that way: the Netflix million dollar challenge in 2009 for developing the best recommendation algorithm was won by a team that didn’t belong to a single firm — or even geography.
But Brown turned out to be only half right. There has been tremendous growth in the gig economy, but most of it can be attributed to unskilled work such as driving (Lyft and Uber), delivering (food, parcels, etc. through DoorDash, Postmates), and doing simple errands (TaskRabbit). A vibrant gig economy for knowledge workers — engineers, consultants, management executives — has not really materialized.
What Went Wrong?
The work of Nobel laureate Ronald Coase on transaction costs provides an explanation. According to this theory, now almost a century old, firms won’t be necessary if there are low costs (money or time) to a customer (individual or business) in searching for alternative providers, assessing their quality, contracting with them, and overseeing and coordinating their work. Clearly, if the work is simple, repeatable, standardized, easily measurable, and controllable, these costs will be low, which explains the success of gig platforms concentrating on work such as ride-sharing, accommodation, and deliveries, largely at the expense of the firms that used to perform these services.
But imagine you’re a U.S. citizen living in Singapore looking for tax advice. If you want to get this through the gig economy, you have two options: Either find an accountant competent in both Singapore and U.S. taxation systems, which could be challenging, or use two freelance accountants, one specializing in Singapore tax law and the other in American law. If you choose option two, you’d need to make sure that the two coordinated properly with each other, which might not be easy. In either scenario, you’d have to find a way to figure out if they were as competent as they claimed to be, and you’d be responsible for drawing up the contract. All of these relatively high transaction costs (search, coordination, and contracting) would largely fall away if you hired KPMG instead, which is precisely why firms like KPMG are still very much with us.
Is the persistence of these costs for high-end, complex services simply because the technology isn’t quite there yet? We don’t believe so. New technologies have significantly lowered transaction costs across the board. The unlocking of information flows due to the advent of Web 2.0 has significantly lowered the cost of finding a freelance service provider. Digitization of knowledge work has allowed for more objective evaluation, which not only makes it easier to have more reliable customer feedback and ratings, but also makes it easier to create performance-based contracts. Rapidly developing AI algorithms have the capability to help in cost-effective matching of demand with appropriately skilled individuals. Products like Slack have the ability to significantly lower coordination costs. And technologies like blockchain that enable smart contracts can significantly lower the costs of contracting.
Given all this, in order to understand why knowledge-based gig economy hasn’t grown, we will need to look beyond technology and economics, and consider instead the role played by organization and culture.
The Culture Factor
Gig workers in the knowledge economy will have to work with and for firms that have pronounced values, incentives, practices, and preferences. But they do not assimilate easily into these organizations (unless they join them) as they often work at arms-length with them and are seen by people in the organizations as outsiders — or even threats —impeding effective cooperation and creating the potential for conflict. In this context, gig workers often struggle to understand, let alone accept, the larger organizational processes, people, and politics of many of the people they have to work with. Performance assessment may also be problematic, especially if the gig worker is hired by a firm to do a job that the traditional metrics of most organizations still cannot properly capture.
When you start listing these problems, it becomes less of a mystery why the firms still prefer to hire knowledge workers as full-time employees or other firms with knowledge workers rather than contract directly with gig workers, despite the ability of tech to reduce many of the more obvious costs.
This may, at last, be about to change. But not from the advent of any new technology — it’s from the global pandemic that is forcing the global economy to its knees. The organizational factors that act as barriers for knowledge-based gig work are the same ones that in the past have inhibited remote work by full-time employees. If these issues can be resolved, whether a remote worker is full-time or gig-based is simply a matter of contractual documentation. Clearly, the experience of working during the pandemic provides useful insights on how to successfully contract knowledge work to external contractors. But we need to approach these lessons carefully.
Focus on the Tasks, Not the Package
Knowledge work is not uniform and, to the extent that you can even talk this way, a given “unit” of knowledge work is itself highly complex. A university, for example, educates students for degrees. A unit, therefore, could be the degree that a student comes out with. But a lot of very different tasks go into creating that unit. So what does “gigification” mean in this context?
Universities could certainly consider using gig workers for graders, teaching assistants, or for pre-recorded online lectures. But it is unlikely that the majority of milestone classes (face-to-face or virtual) that need to be delivered live at specific moments will be delivered by gig workers. Since any degree will inevitably involve both kinds of classes, university teaching will always be hybrid between the two, at least at the course level, possibly even at the class level.
The lesson is that all knowledge-based work can be unpacked into a set of different tasks. To figure out the future of the gig economy for knowledge workers, therefore, we need to analyze things at the task level rather than at the work level. We have found the simple process chart shown below to be extremely useful in figuring out which kinds of tasks are amenable to gigification. It involves asking these three basic questions about each knowledge-intensive task involved in delivering a product or service.
1. Is the task codifiable?
We first distinguish between structured tasks that can be easily specified and measured more objectively, and unstructured tasks that can’t be. Codifiable tasks are definitely contractable to gig workers, and the organizational processes that involve such types of tasks are usually easy to reengineer. Gigification of non-codifiable tasks is not straightforward and understanding which such tasks can enter the gig economy will involve answering the second question.
2. Is there a delay between value creation and value consumption?
In some tasks, value creation and consumption need to be simultaneous, such as when a physician conducts a patient’s physical exam. If such a task is customer facing, it is a big risk to “gigify” it, as these tasks have no possibility of quality checks and re-work. And if that customer is internal, a further layer of complication is added because dealing with internal customers usually requires a high degree of familiarity with the culture of an organization.
But for many tasks there is — or can be — a gap between creation and consumption of value. For example, auditing a firm (value creation) and sharing the results with the board (value consumption) can happen at distinct points in time. In fact, a delay between the two instances is useful, as it provides a window of opportunity to insert a quality check process. Moreover, having such a delay makes it possible for the workflow to follow a more modular design, reducing the need for collaboration, and with it the need for a worker to understand the power and politics of the organization. All this, of course, means that the task will need to be reconfigured, which poses no mean challenge and brings us to the third question.
3. Can the task be done remotely?
Before the pandemic, the firms most comfortable with remote working were software companies like GitLab, which has more than 1,200 employees working remotely. GitLab has put together what it calls a “remote manifesto”, which identifies where remote practices differ from workplace ones. According to this document, remote working favors “flexible working hours over set working hours,” “writing down and recording knowledge over verbal explanations,” “asynchronous communication over synchronous communication.” Note that all these practices would be difficult to implement if there were no gap between the creation and consumption of value.
Before the pandemic, outside of the software industry, firms like GitLab were few and far between, which meant that there was a certain amount of risk to non-software firms in adopting approaches like GitLab’s. But the Covid-19 crisis has forced businesses in industries previously impervious to remote working to reengineer their work processes and bolster their technology support systems, which have been the traditional barriers to alternate work arrangements. This provides a wide variety of natural experiments, many of which are more relevant to a given firm’s need than the experience of software firms and will provide a good starting point to firms contemplating a switch to the gig economy model.
The Covid-19 epidemic could well prove to be a pivotal point in the gigification of knowledge work, and many firms will be attracted by the prospects of the direct and indirect cost savings that the gig economy model seems to offer. But given the complexities of knowledge work there’s also a risk of overreach and wasted investment. The simple task-based categorization we propose will help managers make smarter choices about how just what tasks should be contracted to gig workers.
This article was originally posted in Harvard Business Review.