To outsource or to automate? When it comes to repetitive, manual work, that is the question everyone keeps coming back to. And, as artificial intelligence (AI) and robotic process automation (RPA) have matured, it’s one that’s become increasingly relevant.
Large chunks of many functions can now be automated. These functions span industries and departments, and include major cost centers like invoice processing, customer service, data management, and product testing. And with each step forward in automation technology, consultants and business writers have predicted the eventual demise of business process outsourcing at the hands of automation.
(“Will cognitive automation spell the end of outsourcing?” asks a Deloitte advisor; The rise of AI will lead to the “obsolescence of offshoring,” say Gartner analysts.)
The problem is that automation and outsourcing are a false binary — one does not exclude the other.
And if businesses continue to think of them as mutually exclusive, they’ll miss out on opportunities to improve performance, reduce costs, and beat their competitors to market with big ideas.
Can it be automated?
Debates over whether a particular function can (or should) be automated generally hinge on three factors:
- Whether RPA and AI are advanced enough to handle it
- Whether third-party tools already exist or custom development will be necessary
- Whether the costs and operational challenges of pursuing such a project are worth it
On one end of the spectrum, you have common, ongoing functions like billing, employee onboarding, and compliance. There are many proven third-party solutions that automate these processes, and the cost of automation is typically much lower than outsourcing. Automation of these functions is a no-brainer.
But on the other end of the spectrum, you have functions that are specific, complex, or both, and for which existing solutions are few and far between, if they exist at all.
Let’s use an example: A construction services firm is piloting a SaaS product that automates safety checks based on photos taken in the field. It would be far too costly to hire in-house employees to annotate each photo, so should they automate or outsource?
While AI image recognition technologies are available, they have to be trained for each use case, and their work must be validated. The firm might decide that, due to the high-stakes nature of the results, the technology just isn’t ready yet. They would also need to custom-build the rest of the platform around this technology that enables the delivery, analysis, validation, and storage of the photos.
The other option is to outsource the entire process to another company. But outsourcing also has its drawbacks, namely high relative cost, potential quality issues, long time-to-value, and lack of scalability. These drawbacks might make it untenable for this project.
A third choice
If we think of automation and outsourcing as a binary, then the firm in our example has no good options. The whole project might be dead on arrival. But if we change our mindset and start thinking about how automation and outsourcing can support each other, a third, better alternative becomes clear.
By zooming in, we can understand every repetitive business process as a series of individual tasks connected by logic. From this vantage point, the question isn’t “can this function be automated” or even “can this process be automated,” but “can this task be automated?” Some portion of the tasks in a given process will be automatable using existing tools. But there’s usually another subset of tasks which require human input, whether to perform the task itself or check the work of an AI.
At Stak, we’re tapping into the potential of human-machine cooperation at the intersection of automation and outsourcing. With the Stak platform, businesses can easily build workflows out of both automated and human-powered pieces. Stak allows businesses to tap into a burstable ecosystem of cloudworkers from all over the world who get paid on a per-task basis.
Stak facilitates the payment of these workers in Bitcoin via the Lightning Network, enabling them to work from anywhere and get paid on their own terms, and giving businesses access to a 24/7 workforce.
No more waiting
Taking up our construction safety example again, we can see that with a combination of automation and outsourcing, the future of the project looks a lot brighter.
With a platform like Stak, the firm can use a global workforce of real human beings to either annotate the photos themselves, validate the work of an AI solution, or both. The Stak platform is easily integrated into existing systems and cloud environments, allowing for seamless inputs of photographs and outputs of validated results.
The bottom line: Businesses no longer have to wait for automation technologies to match human abilities, or for third-party tools to be developed that support their specific business needs.
Instead, they can take advantage of human skills while still reaping the benefits of automation, and easily build their own solutions on a powerful pre-existing platform.
The answer to the automation vs. outsourcing debate? It’s both.