You might think this is another article about machines stealing jobs in what is called the 4th Industrial Revolution. It is actually about the age-old need to get the most from employees.
Companies were created to co-ordinate the activities required to produce a product or supply a service in an environment that was complex and competitive.
By bringing the supply and production inside an organisation they could exploit efficiencies that staff doing a specific task could master, increasing production and reducing waste and errors.
For some of the tasks, a machine was used because it could perform the tasks of shaping metal or filling a bottle better than any human. Had a machine existed for overseeing quality or managing the the staff needed to fulfill orders it would have been the first choice.
Humans were better at the job and so they became managers.
That is changing, nowhere more so than the gig economy where micro jobs are assigned based on demand to a decentralised work force.
The growing number of companies like Uber and more recently UberEats manage the demand and supply, not through human managers talking to contractors but through an algorithm communicating to vetted service providers via an app.
The case for algorithmic management
The principle advantage is that a digital boss can look after thousands of staff members and never take leave or have bad day or have favourites or any of the situations that could arise from human interactions in relationships.
The quality of the algorithm determines the quality of the manager and, while it can certainly evolve, it will be very consistent.
It saves on all of the costs associated with a human manager from salaries to leave and challenges of promotion and them leaving the company.
They are precise; able to evaluate huge volumes of data over a period of time and to respond with praise or enforce rules and penalties every time there is a transgression.
They could be so good that they complete the final piece of the company puzzle to create the perfect organisation.
The case against algorithmic management
The term algorithmic management was only coined in 2015. The skill of the digital manager is determined by the quality of the instructions it is given. For now that would suggest it is quite basic.
Algorithms don’t have bad days, but employees do and right now the machines can’t really account for it.
Humans are not machines, some people will need a little feedback, some will need a lot. Giving lots of feedback to someone that does not appreciate it might have the opposite effect to the one intended.
A human manager will be able to explain decisions taken by the business, machines simply inform you of them. This could lead to exploitation.
A sense of transparency is important to gaining staff commitment. The transparency though could result in staff “gaming” the system to their advantage.
Slowly at first then suddenly
Changes like this don’t happen overnight, but the improvements tend to be exponential.
The group that is most vulnerable are middle managers. They tend to manage scale in a business without having specific outputs themselves.
A machine-based system could replace that and see all staff managed and organised via algorithms and a human capital department.
Willing and able
There are two elements that a manager typically has to oversee. An employee's ability to to do a task and their willingness to do it.
Promotion in companies typically rewards those that were good performers by making them managers. Their performance may have been based on their skill at the task, so while they could be very good at training staff to perform as well, they may be less equipped to motivate them to do so.
If a supervisory role is managed by an algorithm, companies can continue to reward high performance staff members without requiring them to move up the corporate ladder in order to do so.
Most staff members don’t need training on what to do, they benefit from on-going reinforcement to motivate them to do more and work better.
While training and supervision can be done by a machine, motivation is still something a human is much more suited to do.
The best scenario is a combination with human and machine working together.
The machine does the monitoring and flags those in need of praise and training, while the human can intervene or work specifically on the motivation of staff.