Whom does (or should) a good leader serve?
Exploring a seemingly trivial question with a non-trivial answer
A few weeks back, I had an interesting discussion with a friend — our conversation centered around the question of whom does (or perhaps should) a good manager / leader serve. This seemingly trivial question led to an interesting (& heated!) debate, by the end of which I was surprised to realize that my own perception of this topic had apparently shifted quite a lot without me realizing it, and what I would have considered to be the correct (& obvious) answer a few years back suddenly didn’t feel quite right.
That’s why I thought it would be good to put some of my thoughts on the topic on paper & share it here, in the hopes that it would also be useful to others.
Also, before we get started, let me say this: some of the things I’ll state might appear quite controversial — if you find yourself disagreeing with any of it, that’s totally ok — in fact, I would be very interested to hear from you what your thoughts on the topic are!
World as an optimization problem
Over the last couple of years, I got to work in a couple product areas that deal with the optimization problem.
From Wikipedia:
Optimization problem is the problem of finding the best solution from all feasible solutions.
Many business problems are ultimately optimization problems: for example, Uber trying to match drivers & riders in a particular area at any given time in the best possible way is one, trying to construct an optimal investment portfolio within certain constraints is also an optimization problem, and so on.
Typically, solving an optimization problem involves constructing an objective function that takes in all the necessary parameters & constraints and outputs the utility that can be achieved given these parameters & constraints, and then searching for a solution that maximizes said utility, within the set of possible solutions.
The reason why I’m bringing it up here is that a ton of questions we face in the real world can actually be represented as optimization problems. And, as with things such as matching riders & drivers in the most efficient way for Uber, the approach to answer those questions shouldn’t be any different. This means:
First, we need to define what the right objective function is
Next, we also need to figure out what the set of possible solutions looks like (e.g., what the constraints are)
Finally, we need to figure out how to solve the optimization problem, e.g., how to find the right solution maximizing utility (this also involves ensuring that we are indeed finding the right solution, and not e.g., the local maximum)
Now, going back to the question of ‘whom should a good manager / leader serve’, it once again is all about solving the optimization problem (trying to figure out what’s the right answer that helps maximize utility, that can be defined as net present value of future profits, or some other similar metric) — and, as with any optimization problem, if you don’t make sure that you have the right objective function & understand the constraints correctly, chances are you’ll never find the utility maximizing solution (or, even worse, you might be optimizing for exactly the wrong things).
The most intuitive answer to the question of whom does a manager / leader serve seems almost laughably obvious: of course, your goal is to serve “the business” (which would also be consistent with using something like net present value of future profits — or other similar things, like average LTV, free cash flow, etc. — as the utility you’d be trying to maximize).
This was the position my friend was advocating for during our discussion. Again, this appears to be a very reasonable & in a way natural answer — the argument being, serving the business ensures that you are optimizing for the one thing that matters the most in the long run.
The tale of stakeholder capitalism
A few years ago, I would have wholeheartedly agreed with this assessment (with perhaps one minor adjustment: I’d substitute “business” for “shareholders” — but this is mostly semantics).
I still remember learning about stakeholder capitalism (stakeholder capitalism is the idea that businesses have a responsibility that extends beyond their shareholders) back in business school & vehemently disagreeing with it, since it felt like an attempt to replace one clear ‘master’ (shareholders) with many masters whose interests are often in conflict and can’t be easily untangled / prioritized, making the problem incredibly & unnecessary complex.
To be honest, I am still not a believer in the theory of stakeholder capitalism when it comes to deciding how to run companies. Shareholders are, and should be, the only ultimate master the company’s management should serve. Other stakeholders (employees, customers, etc.) are of course important, in the sense that without them, there would be no business — but that seems to be exactly the reason why they don’t have to be viewed & optimized for separately: focusing on the interests of the shareholders should ensure that the interests of other stakeholders are also accounted for.
That, of course, rests on the assumption that the company in question operates in a competitive market with robust regulations, so that if the shareholders decided that the best strategy was to be exploitative towards either their workers or their partners/customers, then other players operating in the same space would over time eat away at that company’s business, until the company in question chose to reverse course (or went out of business). Now, in case of monopolies / poorly regulated markets, the situation would be quite different — but then, instead of proposing embracing the concept of ‘stakeholder capitalism’ to ‘fill in’ for the government, the government (and, by extension, the voters) should work to break up true monopolies & beef up regulations where necessary.
This way, all companies operating in the space would face the same rules — whereas calling on individual companies to embrace ‘stakeholder capitalism’ feels a bit like asking companies not to leverage legal tax loopholes: even if you personally chose to do that, some of your competitors surely wouldn’t, and as a result, it would ultimately just help shift the balance away from the responsible market participants and towards the exploitative ones, which won’t be good for anyone.
The reason why I am bringing the concept of stakeholder capitalism up here is that while I don’t believe it’s a very good idea in the context of running companies, I feel the ideas that it advocates for become a lot more useful in the context of being a manager / leader (but not the CEO!), and trying to figure out what you should be optimizing for.
To elaborate on this, when you are a CEO running a company operating in a large & competitive space, acting in the best interest of shareholders should be sufficient, since there is the market to take care of the rest, including situations where the shareholders’ demands for some reason become irrational — worst case scenario, the company goes out of business, thus punishing stupid shareholders & rewarding smarter shareholders with stakes in the competing firms; and if the CEO disagrees with the shareholders, they should really quit & go work elsewhere, which should be another way to maintain balance in the market.
Now, however, let’s consider a case where you’re a manager for a particular team, e.g., at a big tech company (I’ll focus on tech, since this is the industry I work in & am most familiar with — most of what I say below should be generalizable to other industries too though). It can be a small team (e.g., you’re a line manager, also known as M1), or perhaps it’s a rather large team (e.g., you’re M2 or M3 — in other words, there are other managers who report to you) — but ultimately, you’re still a small-ish cog in a very big machine, so there is plenty of management layers above you, and also tons of other teams & products in other parts of the company.
In this case, your immediate stakeholders include:
Your team (both those who report directly to you, and their reports)
A few partner teams that you work most closely with / have dependencies on
Your org (e.g., your product division, etc.)
Your customers
You could perhaps also argue that the ‘business’ (i.e., the company and/or the shareholders) should also be counted as your stakeholders, which brings us to the crux of the debate: is the ‘business’ really your stakeholder in such a case, and can you truly serve it as a manager / leader?
As you might have guessed by now, I would argue that the answer is ‘No’. That’s not to say that you shouldn’t think about the business (quite the opposite) — but for a small cog in a large machine, the ‘business’ in a traditional sense isn’t necessarily something you could possibly serve directly, and trying to do so would often do more harm than good.
Let me illustrate this with a few practical examples, which would hopefully help make this claim a bit less controversial.
”I lead, my team follows” fallacy
A version of this use case is what actually originally prompted the debate I had with my friend.
Without focusing on specifics, my friend had quite a lot of success building/scaling a couple of tech products/businesses in the past (not as an independent founder, but rather as a manager/leader in a large corporation — this part is important). They (just using a gender-neutral pronoun here - it’s still one person I’m talking about) also did it in a market that was markedly different from the U.S., both culturally, and in terms of the number of opportunities available to talent there. This in turn lead them to acquire a management philosophy where, for the lack of a better way of putting it, the leader knew best (optimizing to serve the business), and the team gladly followed, as long as the leader served the interest of the business.
On the surface, this doesn’t sound particularly problematic. In reality though, I am strongly convinced that if you were a manager/leader at a big tech firm in the U.S., following this mentality could be disastrous, for a fairly simple reason: you would likely lose all your top employees very quickly with this approach.
Why would this approach, that clearly works, albeit in a different geography, fail here in the U.S.? The answer mostly has to do with the culture of big tech in the U.S., as well as the differences in the job market’s structure & liquidity.
Culturally, most of big tech (and also many of the top-tier startups too) value the ability to think independently & be capable of solving ambiguous problems on your own. You do, of course, still have managers & some of the strategy inevitably has to be set top down — but ultimately, there is an expectation your manager will mostly coach you, rather than dictate you what to do. I’m sure there are some exceptions to this rule, but all the places / teams I worked on so far were like that, esp. at higher levels of seniority.
Second, the U.S. job market in tech is large & highly liquid (well, perhaps not so much at the moment — but surely in the long run). This means that people don’t have to tie their future to the wellbeing of a single company, and can instead focus more on their own goals & careers. The result is, people might often prefer to work at a place that provides them with flexibility to experiment, make mistakes & iterate (and ultimately learn from that experience), versus having to follow an authoritarian leader who might very well know better, but would stall their development.
This is also the reason why companies might sometimes keep producing great talent long after their heyday (Yahoo comes to mind as the most obvious example) — and vice versa, why other firms that are very successful as a business aren’t always going to be known as a great place to acquire useful skills/experience from the employees’ perspective.
The obvious question to pose here is: if the leader in question is truly capable & knows what’s best for the business, won’t it be worth for the employees to stick with them for the ride, and eventually be compensated in other ways (e.g., through promotions / higher pay / etc.)? Remember the famous phrase credited to Eric Schmidt, “If you’re offered a seat on a rocket ship, don’t ask what seat. Just get on“? This idea seems to fit nicely with the narrative of a visionary manager / leader who truly does know what’s best.
The nuance here, of course, is that, as mentioned before, we are talking about a manager/leader here who is still just a part of a much larger organization. This is a critical point: if we were talking about a startup founder and/or the CEO of a large firm, the situation would be different, since in that case, they would have the necessary autonomy to determine the right compensation model/culture/etc. As a manager at a large established firm though, you don’t have leeway to change those things: leveling & pay are typically set for the entire firm and can’t easily be changed for a single part of it, and so on.
So, as a leader in this context, you won’t have the necessary levers to compensate your team for what would likely be perceived as a negative change in culture (due to the loss of autonomy, if nothing else) — and as a result, if you choose to go ahead, hoping that your team will come to realize that you do indeed know better, you will likely fail & you will do so for a very simple reason: in your attempt to maximize utility (‘serve the business’), you didn’t correctly account for the constraints.
Don’t underestimate the power of constraints!
In the same vein, you are very likely to face similar issues if you choose to ignore other types of constraints that tend to exist in such environments, such as organizational culture, the existing alignment of various teams within the company, the political nuances of the place, and so on.
For instance, the priorities of your business might matter very little to the partner teams you depend on (and sometimes for good reasons too); trying to change the culture too fast & too drastically can in fact damage the morale of the employees beyond what can be justified by any business gains; long-term benefits of maintaining a less than ideal org structure (e.g., having two teams with somewhat overlapping charters) for a while might outweigh any gains you stand to reap from rushing ahead to streamline it (since you’d risk losing good employees who get frustrated in the process) — the list goes on.
In other words, as a manager/leader, you truly do have to resort to the ideas & principles of the ‘stakeholder capitalism’ here to maximize your gains, even if on a micro, rather than macro, scale.
Make sure you have both the right objective function & the constraints
Hopefully, the above examples help illustrate my point to a sufficient degree here. Does a good leader has to care about, and in a way, serve the business? Of course! But at the same time, a manager who tries to ignore the broader context of the org & the constraints they have to operate in, and instead just focuses on trying to build & scale their business first, is quite unlikely to succeed — and in a way, I’d argue that such a leader isn’t actually doing a job they were hiring to do very well, even if they somehow manage to do right by the business for a while.
That last point might be worth spending a bit more time on, and I’d like to use another example to illustrate this.
At one point, it was quite popular to frame the job of Product Managers as being “mini-CEO” or “CEO of the product”. I feel that framing is a lot less popular these days, and for a good reason — while it might not be a bad idea for a PM to assume the “CEO of the product” mindset if it helps them think about the product & the business holistically, in many other ways such framing is going to be by definition flawed: similar to what we discussed above, PMs operate under a large number of organizational constraints that are completely beyond their control (culture & pay, broader organizational priorities, the structure of the sales org, and so on).
Failing to understand those constraints, or worse, trying to ignore them, can be disastrous & certainly won’t lead one to success, and the same applies to most types of management jobs out there, including those that have nothing to do with Product Management (which just served as a convenient example here).
Conclusion
To sum it up, upon some reflection, the question of whom does (or should) a good leader serve seems a lot more multifaceted to me from where I stand today than I would have expected.
It also provides a great illustration of some of the benefits (and also dangers) of trying to turn real-world problems into mathematical abstractions: do it right, and you have an extremely powerful instrument to analyze the world with — but do it wrong, and your view of the world risks becoming dangerously one-dimensional.
Hopefully, this article has provided for some entertaining & thought-provoking reading! As always, if you have any feedback you’d like to share, or more thoughts on the topic, please let me know in the comments section below!