Sign up

Rethinking the management structure of machine-assisted companies

June 8, 2018

By looking for — and learning from — patterns in the data, machine learning algorithms bring quantitatively better decision-making to an organization. This shift demands a reshaping of an organization’s management structure; a shift from a single downward-branching command structure to a set of semi-independent command hierarchies, operating in parallel.

Is a pyramid the ideal management structure for an organization that relies heavily on machine learning?

The traditional pyramid is shaped by experience. Those at the top are the ones who know the company best — those whose understanding of the industry’s rhythms and trends has enabled them to make informed decisions about the organization’s strategy, and to accurately forecast the results of those decisions.

When you introduce machine learning into an organization, this management structure becomes redundant. Because the machine learning software serves up smart, actionable insights to everyone who can access the data, those at the top no longer need to enforce such a strict, pyramid shaped hierarchy.

They no longer need so many delegates checking that their message from the top trickles down to the base.

And because of the accuracy and efficacy of insights the machine learning software guarantees, they don’t need so many branches of line managers reporting on the efficacy of its implementation as it resurfaces back at the top.

For machine-assisted companies to thrive, they need to embrace a new management structure.

This is where the pillared structure comes in. By drawing on their data and acting upon the insights the machine learning software produces, they provide the quantitative foundations for the qualitative decision-making above.

In a machine-assisted company, decision-making is still guided by human intuition, but it now has the data to defend it. Less a pyramid, the management structure evolves to become a Parthenon.  Tweet This

Why is this the way forward? Because an inevitable outcome of machine learning is that it elevates the importance of the employees using the data.

Once subordinates at the bottom of a pyramid-shaped hierarchy, they become incorporated into one of the company’s pillars.

Each pillar, each team, is independent; its people collaborating to produce analyses that shape decisions at the top.

This pillar management structure only works if machine-assisted companies democratize their data.

For people in a machine-assisted company to get the most out of each others insights, everybody must be able to see and access both the raw data and the software’s actionable insights.

More importantly, after accessing the data, every team must be able to act independently on the insights it provides. Having to report up the chain of command means wasting time and allowing valuable insights to stagnate and become irrelevant. So having a team leader who understands how to read the data is vital.

Independence is crucial. But this independence doesn’t mean doing away with hierarchy altogether.

While democratized data is essential, democracies still have leaders.

A machine-assisted company still needs industry veterans at the top of the management structure. Their role is to make experience- and intuition-based qualitative decisions about the organization’s values and overall direction.

These decisions determine how people in the organization use the data. They decide which data to feed the machine learning software, which patterns the software should look for, and which insights they want out of it.

The big difference in a machine-assisted company is that instead of just dictating their decisions downwards, leaders use data-driven analyses to shape their qualitative decision-making.

Thus, being able to read the data, and understand how it provides the quantitative brickwork to their qualitative mortar, becomes a key aspect of leadership.

This is where qualitative, value-driven leadership comes into play.

Now that machine learning software can gain decades of industry expertise in a matter of days, different kinds of experience become valuable in different ways.

For executive management, the intuition to distinguish between relevant and irrelevant machine insights becomes the defining quality of leadership.

For the team leaders managing the individual supporting pillars, the defining qualities of leadership are an instinct for trends and the ability to ensure cohesion between the individuals in the team and the other pillars in the organization — cohesion that sees everybody directing the machine insights towards the same goal.

Machine-assisted companies have to replace the pyramid with a more columnar management structure.

They need a management structure in which pillars of command are each free to act independently on the insights they surface from their machine learning tools — while at the top, a group of veteran experts may still make the qualitative decisions.

As an unprecedented type of technology with almost limitless potential, machine learning is already revolutionising not just your company, but all aspects of our lives. If you want to keep up, putting the right management structure in place to meet it is not a choice, but a necessity.

About the author

Alexander Meddings is Evo’s content expert on artificial intelligence, machine learning, and related topics.

He is an experienced journalist who covers branding, social media, marketing, and technology, with degrees from the University of Exeter and University of Oxford.

Hey! Was this page helpful?

We’re always looking to make our docs better, please let us know if you have any
suggestions or advice about what’s working and what’s not!

Send Feedback