How Many Direct Reports Is Too Many?

Episode #396 How Many Direct Reports Is Too Many?

OVERHEADS? OR ENABLERS?

There’s a seductive idea that’s quietly been taking hold in the corporate world. That layers of management are nothing but overhead; that self-managing teams are the future; and that the fewer bosses you have, the more agile and innovative your team becomes.

Well, it sounds like it should be true, but it’s not. It’s bullsh!t.

In this newsletter, I start with the fundamentals: why organisations need teams; why teams need structure; and why leaders are essential to making both work.

This leads into the discussion on the main tool of indiscriminate cost-cutting: span of control. What really happens when companies blow past their traditional limit of 6 to 12 direct reports, in the name of efficiency? And who pays the price?

Is Meta’s recent decision to run an AI team at a 50:1 ratio flirting with disaster? I’m going to examine the real costs of de-layering: the ones that don’t show up on a balance sheet until it’s too late. I also give you four practical factors that you can use right now to assess whether your own structure is working for you or against you.

CAN SMART PEOPLE SELF-MANAGE?

Many years ago, a colleague of mine told me about an experiment he ran. He was in the tech industry and he decided to harness the brain power of the best and brightest people he could find.

He went out and hired a couple of dozen PhDs: all brilliant people with high intellect and deep knowledge in their area of specialisation. He figured that if he put these geniuses all in a room together and gave them some broad objectives, they’d find a way to maximise value.

You may not be surprised to hear that it was an unmitigated disaster. After a little over a year, he had to shut the experiment down. He told me it was like herding cats: everyone went off in a thousand different directions; massive amounts of activity, exploration, and theorising, yet no commercial value whatsoever was harvested from that unit.

This is just one story that reinforces to me why organisations need teams… and why teams need structure… and why leaders are needed to bring this structure together.

Organisations are called organisations for a reason. They enable resources to be organised into work groups that perform specific functions. This principle is called Division of Labour. The theory goes all the way back to the Greek philosopher Plato, around 400 BC, but the modern roots are anchored in the work of Adam Smith. His publication from the late 18th century, The Wealth of Nations, is still relevant today: it’s a classic.

Division of Labour has served us well for centuries and it’s going to continue to do so long into the future. It has become part of how we think. Even in the age of AI, we’re going to be thinking about what tasks can be given to AI, and what tasks have to be divvied up amongst the humans.

FIVE REASONS WHY LEADERS ARE ESSENTIAL

If you assume that you need division of labour and teams, then you also have to assume that you need someone to lead those teams. Here are the top five reasons why, in my estimation, leaders are necessary to achieve any sort of team performance.

1. People need direction.

Someone has to feed the work to the team, and it has to be coordinated with other teams. Individual contributors are typically pretty poor at this. They just want to do sh!t.

Leaders have access to planning information and other data. They make decisions about resource deployment. And the larger the organisation, the greater the complexity of the strategy at the top (and the more difficult it is to break it down into manageable chunks).

2. Organisations need visibility.

There has to be some feedback mechanism for results. Leaders are there to ensure that results match objectives; they have to report on progress; they have to solve problems; and they have to make adjustments on the way through.

The feedback loop from the bottom upwards is almost impossible without leaders.

3. Talent needs to be allocated deliberately.

The highest value work should be allocated to the most capable people. Leaders are in a position to match capability to opportunities. This also feeds into talent management and workforce planning. This stuff doesn’t just happen by itself (not to mention the fact that your talent requires development and nurturing… if you want to keep it, that is).

4. The will of management needs to be executed.

At the very top of the organisation, the leaders know what has to happen in order to create the most value for the business. You don’t have to study the psychology of incentives for long to realise that leaders are more likely to execute the will from the top than are the lower-level workers.

Even if you could find a way to incentivise frontline employees sufficiently, it’s almost impossible to differentiate their performance in any meaningful way, without a leader making that assessment.

5. Communication needs to be disseminated.

The messages from the top have to make it all the way to the bottom: email has shown itself to be a very poor method of achieving this.

People see the world through the eyes of their direct manager. They don’t want something explained by someone four levels up the hierarchy. They want it explained by the person they report to day-to-day, who they at least have some relationship with.

When it comes to managing teams, there’s a whole bunch of things that simply won’t happen unless a leader makes them happen.

One might even go so far as to say that the self-managing, self-regulating team is an A-grade bullsh!t story.

It was cooked up by executives who were trying to reduce management costs so that they could improve their bonus potential… it was supported by employees who were much happier in the absence of any oversight… and it was perpetuated by leaders who couldn’t articulate why they were critical to the functioning of the business.

Quietly and gradually, the commitment to managers and leaders seemed to evaporate.

COULD YOU MANAGE 50 DIRECT REPORTS?

The most recent evidence of not being valued came in the form of an article that I read recently in Fortune Magazine, titled Meta’s new AI team has 50

engineers per boss. What could go wrong?

This brings in the concept called span of control. Traditionally, the accepted span of control was between 6 to 12 people. In other words, anyone in a leadership position would have between 6 and 12 direct reports.

This is a pretty good rule of thumb, but there are also some good reasons to break it from time to time.

For example, some smaller functions require higher-level expertise and strategic skills at the top, even though the number of subordinate functions may be limited. Take the general counsel (the chief lawyer) of a large company. They may be at the senior vice president level, very senior in the hierarchy, but possibly only two or three direct reports looking after key service areas or technical disciplines.

By the same token, a regional manager might have direct reports in quite a few different locations, demanding a span of control even greater than 12.

To give you an idea how it should play out, when I was running CS Energy, the spans of control varied greatly from level to level. The company was generating over $1 billion in revenue each year; the balance sheet was worth over $2 billion; but there were only about 500 full-time equivalent staff.

As CEO, I had five direct reports:

1. Operations (which included asset management);
2. Finance;
3. Energy Market Trading;
4. Commercial and Strategy; and
5. Corporate services (including legal, HR, risk etc).

These five portfolios gave me line of sight to the big value drivers of the business.

Underneath me, my executives had a range of spans, varying from a low of 3 to a high of 9 (and I was pushing the exec with 9 to slim it down, as he couldn’t cover the breadth of that easily).

Knowing that I ran a multi-billion dollar company with only 5 direct reports should give you a little bit of perspective.

If you have too few direct reports, the purpose of each level can become blurred. And if you have too many, you can’t give each of them the attention they need.

SOME INDUSTRIES PREFER FLAT STRUCTURES

Tech companies have always been outliers. They tend to have very large spans of control. They’ll tell you that this is because tech companies lend themselves to flatter structures; theoretically, it’s a function of the nature of the work, because collaboration and Professional standards are supposedly high.

I find this statement fascinating, particularly given the fact that so many technology projects either fail outright or, in the best case, don’t deliver the promise benefits.

It’s worth going back (just for sh!ts and giggles) and having a listen to a podcast episode I did last year where I outlined the reasons for this. It was Ep.363: Why Your IT project Will Fail. But I digress…

Historically, even the tech sector had considered a span of 25 to be the outside limit. But, apparently Meta has gone to an AI team with a 50:1 ratio.

The theory says that companies are more agile when they have flatter structures: it streamlines decision making, and it positions management closer to the frontline workers and the customer experience.

Well, that’s plausible, but the downside consequences are massive. According to the latest survey from Gallup, in the last year, average team size in the US has increased from just under 11 to a little bit over 12. This represents a 50% increase in team size since Gallup first measured it in 2013.

But it’s still the case that two thirds of managers have less than 10 direct reports. So, the numbers are skewed by a small number of companies that choose to go wild with de-layering. There’s no doubt that flatter structures save direct costs, but there are a whole lot of indirect costs too. And just because you can’t see them doesn’t mean they don’t exist. You can’t always recognise them immediately, but they become more and more obvious over time.

When you think of it in the context of the five reasons why we have organisations and teams in the first place, it’s really easy to see the number of ways that this could go wrong. You’d have a situation where:

• Communication between layers actually deteriorates because the leaders who are left don’t have the bandwidth to understand everything that’s going on below them;
• Junior employees get overlooked, and don’t receive the necessary development;
• Line managers burn out because they’re so overworked;
• Good people and bad people alike are left to their own devices;
• Projects in distress aren’t identified until it’s too late; and
• Decision quality deteriorates because there’s no higher level input.

Personally, I have no doubt in the value of management, leadership, and organisational structure. It’s never perfect, but it’s better than the alternative, which is predictable chaos.

Companies prove to be highly resilient to the sh!t decisions that their boards and management teams make, but anyone who thinks that removing layers of leaders doesn’t have an impact is living on borrowed time.

FOUR KEY FACTORS IN STRUCTURE DECISIONS

I want to give you four key factors that you can lay up in determining how your structure should look and how many direct reports you should have:

1. The physical constraints.

Are you running teams in different locations? If so, having a direct report from each location often makes sense.

But, once you start to get up to 8 to 10 locations, you should probably consider putting in a span breaker. What does that mean? Let’s say, I’m an operational manager in Sydney… I might have 15 store locations that sit underneath me. I could potentially use a couple of area managers to oversee some of the smaller stores.

In this way, I could turn 15 reports into, say, 7: two area managers, each with five smaller stores underneath them, and the five big flagship stores still reporting directly to me.

2. The experience of your direct reports.

It matters what level you’re operating at. Using my CS Energy example, my five direct reports were all capable, experienced senior leaders: they were independently competent and they required less oversight for me as their direct manager.

However, this also demanded that I spend significant time with each one to make sure they were getting the direction, support, and guidance they needed from me to keep their teams moving.

Less experienced reports at lower levels will need even more handholding on the basics.

3. The uniqueness of the role itself.

In organisational design, there are a few rules of thumb:

• Each layer must have a unique purpose;
• Each layer must produce a unique set of deliverables; and
• Each layer must focus on a unique time horizon.

If you apply these principles, you’ll find out pretty quickly what’s wrong with your structure.

It doesn’t necessarily mean you shouldn’t use workarounds like the span breaker area managers I mentioned before, but it certainly forces you to consider why your structure is the way it is.

In the Meta example, I suspect the amount of ground that needs to be covered by only two layers (that’s 50 technology gurus and 1 manager), is unworkable.

In cases like this, what typically happens is that the results become more dependent on individual performance. Some people are going to find a way to make it work and others are simply going to flounder. But with such a large span of control, you’ve got no chance of catching an imminent catastrophe before it hits.

4. The dominant culture.

The culture of the company is going to largely determine what’s acceptable in terms of structure. In Meta, I suspect you wouldn’t get very far if you suggested that each of your leaders should have a maximum of five direct reports: it may even be a career limiting move.

There are other cultural factors as well, of course, like:

• How much a leader is expected to dip down and work on the tools?
• To what extent is decision making either centralised or decentralised?
• Are decisions pushed down to the lowest practical level at which they can be made?
• How strong are the single-point accountability structures?
• Is a bias for action and individual independence valued inside the culture?

Understanding the barriers and expectations is the first step.

The second step is to push your span of control as far as the culture will let you… and then, make the case for going even further.

KNOW YOUR VALUE

If you don’t have the right number of direct reports, you’re not going to be able to give each of them the attention they need to bring out their best performance. This is why you’ve got to be really clear on how you’re going to make that happen as a leader.

Vague notions of why you think you need either fewer or more direct reports simply aren’t going to cut it: you’ve got to be able to clearly articulate the value that might come from any structural change.

To make any structure work, you need strong leadership and a constructive high-performance culture; and you can’t achieve that unless you’ve got the bandwidth to spend with your direct reports.