Let’s imagine you sit at the very top of a very tall, multi-layered organizational tree. You have a question about something that can probably only be figured out by those at the bottom of the tree. You package up this “ask” like the unique snowflake that it is. You send this snowflake down the tree to get sorted, expecting an answer back in a short time frame.
Now let’s follow that snowflake on its journey down the organizational tree. As it goes down, it hits successive layers of management. At each layer, management has the following choices to make: add their own ask to the snowflake, keep the snowflake exactly as is, or subtract something from the snowflake. Most managers are going to add something. A few might just let it pass through exactly as is. Very few will subtract something.
And so as the snowflake starts to progress down the tree, it starts to transform. On the whole, it will start accumulating snow because of how many managers down the tree will choose to add something. Eventually, by the time it finally falls to the base of the tree, the snowflake has grown into a huge snowball. The people at the base of the tree can only shake their heads at how “leadership” could expect so much in so little time. They crank out the work anyway and send the snowball back up.
Now a few days later, you get an update that the snowflake you sent down has been answered. But instead of the snowflake you sent down, you’re presented with a giant snowball. This isn’t what you wanted at all. Your intent has been lost.
Stacking Snow
Let’s evaluate the outcome of this little vignette:
The high-level leader ended up dissatisfied
The lower-level individuals ended up doing a lot of work for the wrong outcome
Both the leader and the lower-level individuals bore the cost of middle management stacking snow. So why do they do it?
Because they seek to impress.
It’s only human. We all want to look good to our bosses. But if taken to the extreme, we lose sight of what was really important: the intent of the snowflake to begin with. And in the context of the organization, achieving the intent was the real goal. The desire to impress in such a way that it overwhelms the leader’s intent cost the organization its ability to achieve that intent.
You might wonder, why not just get rid of middle management? Couldn’t this entire thing be avoided if the leader had the ability to go straight down to the individual workers? But this is fundamentally impractical because of the amount of time and coordination cost that would take. It would be an inappropriate and low leverage use of time as the high-level leader to spend time on this effort rather than delegating it.
It’s important to understand that at a certain scale, middle management is a feature, not a bug. Its existence is to stand up the culture, propagate right intent, enrich things with context, and coordinate across the organization. It’s not practical nor desirable to assume that middle management is useless. In the history of humans coordinating to do large-scale things, we have eventually all gravitated to organizational structures that create middle management functions to scale.
What this vignette is highlighting is not the uselessness of middle management but rather the importance of doing it right.
Filtering Properly
Middle management generally gets a bad reputation. To the individual contributor, the filter from middle management is all they can see on higher-level intent. If the middle manager cannot extract the intent properly or contextualize it appropriately, the IC gets confused and frustrated. There’s also a larger number of middle managers than there are high-level leaders. By definition, you’re going to have a larger population and you’re going to see greater diversity in competency at doing this kind of filtering.
For the middle manager at large, the job isn’t easy. By being in the middle, you are beholden to those above you and to those below you. You have to manage both ways. Thus, your ability to seek true intent, modulate how you will try to impress, and how you contextualize things is essential.
Let’s revisit the three possible actions a manager could take in our vignette of the falling snowflake: add snow, let it pass through untouched, and remove snow. A competent middle manager would evaluate what the intent of the ask was and then consider carefully what is required. None of the actions are necessarily wrong in and of themselves. They are wrong when poorly considered without any critical filtering or evaluation. For each action, let’s consider the set of critical evaluation questions we should ask:
Add snow. Would the intent of the ask be better served with an addition that only you understand? Or are you just trying to add something so you come off prepared and extra diligent? Only add if you truly believe it would better serve the intent of the ask.
Pass through. Can the level below you fully understand the intent? Do you need to contextualize it? Do you need to give guidance on its direction? Are there parts of the accumulated snow that you disagree with and would even cut out? Only pass through if you believe the intent is clear enough for the next level to adequately determine how to carry forward.
Remove snow. This is perhaps the most unpopular action, but often one of the most essential. People do not take this action out of fear. So ask yourself do you agree with everything that is being asked? Do you think it really answers the original intent? Do you know it will be extremely costly? If so, it’s worth the double check and at least argumentation for removal. This might sound scarier than it is; there are plenty of professional and polite ways to get additional clarity from your management chain and to pick and choose your arguments for removing things that do not answer the original intent.
Trickle-Down Trust
The argument I’m making here requires a lot of intentionality from every single layer. It’s costly to each individual involved, whether that’s management or not. It’s inefficient at the scale of the individual in exchange for efficiency at the scale of the organization.
But there is a way to eventually reduce that cost curve: and that’s through trust. Things become more fluid and easy when eventually each layer is able to trust those above it and those below it. Trust is the precondition for candor and feedback. If you trust that your management chain will not view your questioning as insubordination or unwillingness but rather genuine desire to obtain clarity, then you will be much more inclined to be intentional and to be honest in your assessment.
That trust must start from the top and it must be earned in order to effectively trickle down. The most important decision a leader can make then is to effectively always ensure that the next immediate layer below them is composed of people they can trust and who share a common definition of high performance.