The Mathematics of Leadership: How to overcome a Productivity Collapse

As your business grows, it will, inevitably and periodically, experience severe degradations in productivity and output. To get past such points, you'll need to change just about every aspect of how the organization works in one way or another. But how do we know that we're not just making arbitrary changes that could make things even worse? What guides our choices during such times?

In the previous article in this series, we explored the mathematical phenomenon of Catastrophe Theory as it applies to organizations. Specifically, we examined the nature of, and reasons for, periodic, sudden collapses in productivity that occur as teams grow in size. In summary of that, and with reference to the chart below, we would hope that hiring more people would result in a corresponding rise in business outcomes (the solid line) or better (the dotted line):

But, in practice, our outcomes more closely resemble the chart below, which shows that we'll regularly experience productivity plateaus, or inflection points, as we grow the size of the team:

We previously discussed why this phenomenon occurs, and we built a simple model to predict approximately when it will happen. This, then, leaves the question of what to do about it when it does happen. I previously stated that you have to change almost everything about how your organization works to successfully navigate such an inflection point and that, if you don't, your business will instead likely take one of the paths depicted by the dotted lines above. In other words, if you are not prepared to make radical changes to the business, you will never exit the inflection point (unless you count business failure as an exit).

But what specifically are we trying to achieve through such changes? What guides them? It's clear that arbitrary, wholesale re-structuring, process changes, senior hiring, etc. are likely to only make things even worse if they are not directed towards some clear purpose. Unfortunately, it is just such arbitrary actions as these that are often undertaken during the panic conditions of an inflection point, under investor or shareholder pressure.

So, what goal should we be aiming for, to direct the changes that we need to make?

Navigating an organizational inflection point

The primary goal in any organizational design exercise - and especially when we are trying to recover business productivity after hitting an inflection point - should be to optimize for the Agency of individual employees within your business, particularly those on its front-line (i.e. the people who actually do all the concrete work of the business). Specifically, in the case of having to navigate an organizational inflection point, your task is to restore lost agency.

Agency for an individual employee feels like this:

• I am sufficiently competent and skilled to do my job.
• It is clear to me and others what I own and am responsible for.
• I'm well aligned to and understand, the wider goals of the business.
• I am able to take appropriate decisions about the things for which I'm responsible, without unnecessary referrals for approval. Where appropriate approvals are required, decision latency is very short.
• I have the necessary resources and tools to carry out my responsibilities.

Note that installing Agency in our business holds us to a much higher standard than creating Autonomy. With autonomy, I might not be competent to do my job or be aligned with company goals though I still have the freedom to make consequential decisions (does that sound like anyone you know?). Or, I might know what I'm responsible for without having the resources to carry out those responsibilities. Agency requires all of the elements listed above to be in place, without which there is no agency, only the temporary illusion of agency.

The Trouble with Agency

The trouble with agency is that, left alone, it declines as a business grows, even in the absence of organizational inflection points.

If you're not sure about this, ask yourself whether employee number 104 in your business has the same agency as employee number 4. How about employee 204, or 504? Let's briefly explore the reasons for this decline by comparing employee 4 with employee 204 in a typical, growing business.

Employee 4 communicates easily and closely with the other three staff members - they sit at adjacent desks, after all. So she knows the goals of the business, doesn't need authorization (or can obtain it very quickly) and owns most of the delivery tasks required to carry out her job. In fact, she probably owns an entire function of the business, as a co-founder, or first employee.

But that's not how it feels for employee 204, She wasn't inducted very thoroughly, so she doesn't quite understand - neither in detail, nor at a fundamental level - what the goals of the business are; the strategy slide-deck is out of date, her manager is too busy to really explain things properly and, these days, no-one has enough time to properly re-invest in making the onboarding process better. Employee 204 isn't entirely clear what she is responsible for (several other people in the business also seem to believe that they own at least part of what she thought she owned). It's not clear to her how to get authorization for certain decisions, nor is she sure what decisions she can make and which ones require that authorization (ownership is not well defined, and is somewhat implicit).

It's easy to expand on this example, of course, and, if you've ever been that "employee 204", I'm sure you'll recognise the scenario. It's obvious what this type of experience does to any sense of personal agency. Over time, the above issues each elongate and combine with one another, further reducing agency until a tipping point is reached.

At this tipping point, collective agency suddenly collapses. This, in turn, precipitates the organizational inflection point, and overall company productivity markedly declines:

How to tell when Agency is about to collapse - The Frustration Index

When the tipping point comes, it comes suddenly, but we can still receive some prior warning that agency is about to collapse before it actually happens. The most effective and committed people in your business, when experiencing an initial decline in agency, quickly become very frustrated.

Frustration rises when agency falls and reduces again when agency recovers, provided agency is recovered soon enough (within a few months of the initial collapse of agency). If agency hasn't recovered by then, frustration still recovers eventually, but only because all of the people in your business, that cared enough to be frustrated, have left.

Suppose that your business was able to track a Frustration Index; this index would be your early warning that agency was declining.

In practice, we can approximate this index to a useful degree. It's well worth the effort to do so because frustration is easier to detect than agency. If you are paying attention to your organization, you can detect relative changes in the Frustration Index anecdotally, for example, in conversations with your colleagues or changes of tone in their written communications. These provide a form of early warning that something is amiss, in time to mitigate a major organizational inflection point. Better still is to create a culture that monitors and brings rising frustration levels to the surface early and reliably. There are various ways that early detection can be built into the organization, for example:

• Meet front-line staff regularly, and listen to them. Actively encourage them to explore what annoys and frustrates them about the business.
• Encourage uncensored (but not inappropriate), open employee discussion on your intranet and related forums.
• Have a means by which employees can raise concerns anonymously, and can submit questions anonymously to townhall meetings.
• Staff surveys are useful here too, but they are not enough to track frustration by themselves - they're typically run annually and the questions are set by senior management, which naturally brings a certain biasing to the survey. So the questions asked may not be the right ones to detect frustration. Some organizations, therefore, run a partial survey every month, along similar lines to the Net Promoter Score concept in product management, to overcome the frequency issue and to reduce the tendency towards sentiment-biasing.

Restoring Agency

Individual-agency would, in an ideal world, remain constant as a business grows and it's useful to keep that in mind as an aspiration in everything that you do, organizationally (i.e. never deliberately sacrifice agency in favour of other organizational imperatives).

But, in practice, because agency tends to reduce with time, our task in organizational design is always to continually work to restore lost agency. How can we achieve this? In the previous article, we discussed that there are three components that constitute your organization; how these work individually and with each other dictates the performance of your business. Therefore, it follows that restoring agency usually requires action in each of these areas.

Structure (Statics)

Favour structures that assign clear, distinct and non-overlapping ownership. If everybody owns a thing then nobody really owns it. Eliminate overlay-leadership models. Don't run with too many facilitation roles, which often obscure ownership, beyond a certain level. Equip teams appropriately to reduce hand-offs to and dependencies on other teams as much as possible. People need to have ownership and they need to feel that they do, to have agency. If the people with ownership roles in your business don't profess to a slight nervousness about that ownership that's probably because they probably don't really perceive themselves to have ownership; maybe they are right.

Process (Dynamics)

Configure your decision-making model to make decisions as close to the front-line of the business as possible. Decisions that can be made on the front-line of the business have very low latency, which allows more decisions to be made in a given unit of time, thereby increasing velocity and agency.

On the other hand, decisions that require executive approval have a much higher latency - the vector distance between the front-line of the business (where the real work gets done) and the most senior people in the business is, relatively, very large. Only decisions required by the front-line that absolutely must be made at the executive layer should be made there. All other decisions should be made as close to the front-line as possible.

When the decision model is misconfigured, such that decisions that should be made on the front-line of the business are referred to deeper levels in the leadership hierarchy, there are serious consequences.

Firstly, decisions made far from the natural point of decision are of generally poorer quality because the decision-maker lacks context. And they are late, due to the increased latency. Worse still, when decisions are made far from the point at which they should be made, the front-line of the business becomes passive in organizational terms - if you send people a signal that they don't have agency or that they don't own the evolution of the business, they begin to act accordingly. The combination of these issues - poorer quality, high latency decision-making and a passive front-line - heavily impacts both the performance of the business and its development.

People (Capability)

Finally, a key element of agency is Individual Competence. Without this, setting up structures and processes that put decision-making close to the front-line of your business is extraordinarily dangerous.

The critical point here is that what constitutes competence at each level of scale is different. Therefore, you must continually re-invest in the capability of your people as the business grows, just to maintain the same relative level of individual competence as existed previously. For example, managing a team, designing a release process, influencing peers, communicating etc., all of these must be performed differently at each level of business scale.

This isn't to suggest that those working in larger businesses are somehow better than those in smaller ones but, rather, that an employee's capability needs to match the configuration and scale of the business. Senior managers in very large businesses often screw-up spectacularly when joining small start-ups through exactly the same principle. For example, they've often only operated an existing organizational model, whereas smaller, growing businesses need leaders who can create or upgrade the model.

If you are regularly increasing the number of people in your business, then you know that an inflection point is coming.

This People corner of the People-Structure-Process triangle depicted above is the most important of the three; it is frequently the cause to the others' effect. It's most often a failure of the business or its people to continually re-invest in individual competence that, in turn, leads to its people not being able to engineer structures and processes appropriate to the next level of scale, thereby precipitating an organizational inflection point.

The best defence to this outcome is anticipation. If you are regularly increasing the number of people in your business, then you know that an inflection point is coming. Therefore, long before they need to operate at the next level of scale, you should actively encourage the people in your business to learn about how larger organizations work. They - and you - should be regularly reading biographies, how-to's and blogs from industry peers on organizational design, and on best-practice, scalable processes and structures.

Your people should be reaching out to larger businesses, and asking for their time to share their learnings and their mistakes (which, perhaps surprisingly, they will usually be willing to do). They should be networking actively with people from companies that have moved beyond the inflection point that yours is currently approaching. And a continuous and intensive leadership development regime that prepares your people for later and greater leadership responsibilities is also critically important.

Optimizing for individual agency is not easy. It challenges leaders to "give away" control while still retaining accountability. And it places real ownership pressure on everyone in the company. But focussing on agency is essential if we are to scale a business productively. And, should you find yourself mired in an organizational inflection point where productivity has severely degraded, then agency is your north star, by which you can navigate to better waters.

-----------------

Other, selected articles:

Mathematics of Leadership: Markov Organizations

Mathematics of Leadership: Catastrophe Theory and Team Size

"This was the good town once" - what makes and breaks a great company culture?

If your company's strategy was a song, would anyone buy it?

Don't ask me why I'm leaving: why exit interviews don't work

Ending the unconscious sexism of the tech-industry

How to stop hiring the wrong senior people

The best leaders know when to be led