Toward the end of my tenure at one tech company early in my career, my director expressed the following regarding my leadership potential:
“In order to be a manager here, you’d need to be more of a, uh…company woman.”
I understood what he meant. On most disagreements, I predictably stood on the side of the person with less power. I would speak up on behalf of the designer frequently asked to play secretary in meetings. I would speak up about the vast majority of leadership coming from one demographic. I would speak up on behalf of an experienced product manager struggling to provide input in the PM hiring process. My propensity to do this didn’t make my director’s day easier.
His reticence to put me in a position to be heard makes sense. More than a few studies have demonstrated that we prefer to hear from those who confirm what we already think. (If you’re interested, Noreena Hertz presents a few of them in this piece for the Harvard Business Review).
So it’s no surprise that people in leadership positions, in general, disproportionately collect input from people who tend to agree with them. If somebody frequently challenges what would otherwise be consensus, they slowly find themselves consulted on fewer and fewer decisions. Well, that’s no good. So when company people disagree with the apparatus, they don’t say so.
That creates undesirable downstream effects. You’ll hear about those downstream effects if you sit down to lunch with the employees. This is because all of those concerns that company people don’t push up to leadership have to go somewhere, so instead they get pushed down onto employees. That upsets employees, and people talk—especially upset people.
And that’s why, when you sit at that lunch table, you’re going to get the real story.
The developer on your left had ethical concerns about working on a project. Her manager responded with ‘the director knows what he’s doing and you have to trust him.’
The product manager on your right raised a concern about inclusion, and her manager responded with placating words.
The designer across the table filed a harassment report. Her manager, and HR, have responded with silence.
This happens because company people protect the easiest thing for the company in the short term. That’s usually the status quo. There are some notable exceptions: for example, when the company is making an offer to someone they really want or when someone the company wants to keep gets an offer elsewhere, the status quo is the company not getting the resource they need. In these cases, changing that status quo is the easiest thing for the company in the short term, so company people reach to change that status quo. This is why company people respond to inquiries about raises and promotions with ‘not the right level’ and ‘wrong time of year’ and ‘not in the place right now’…right up until those people find an offer somewhere else. Suddenly this is absolutely the right time and place, and the whole thing gets approved in a matter of hours.
If these things are happening at your company you can rest assured that everyone knows: no corporate response is going to stop the grapevine. But clearly it doesn’t make your leadership look good, and it’s not great for your culture either. Let’s talk about why managers become ‘company people’ in the first place.
Managing at Most Tech Companies
When managers keep a lid on their reports’ concerns, their bosses don’t have to address the concerns. The managers’ bosses are usually responsible for choosing, retaining, and promoting managers, and they’re looking for people who make their jobs easier.
So, to be chosen, retained, or promoted as a manager, you gotta make your boss’s job easier: that is, keep a lid on your reports’ concerns. If you don’t do it, you don’t become/stay a manager. Thus, managers have to exhibit company people behavior.
Additionally, a lot of literature about management (written, predictably, by bosses of managers) advocates for managers to strike a balance between ‘managing down’ (representing the interests of the company to reports) and ‘managing up’ (representing the interests of reports to the company). That is not an even balance. The company is the power-heavy side, so the ‘balance’ seesaw falls hard in the ‘manage down’ direction. To even begin to ‘manage up,’ managers must start out with social and political capital, and ‘managing up’ uses up that capital—so managers might run out and not be effective advocates for their reports, or they might be saving up that capital for their own next promotion.
That’s not the managers’ faults. They’re stuck in a company structure that does not equip, train, or encourage managers to advocate for their reports to the company.
Now, let me suggest a different approach to managership and why I think it would benefit a tech company.
Managers as Advocates
Lawyers represent their clients. In court and in mediation, the lawyer makes the case for the client’s interests.
What if management worked this way? What if, instead of asking a manager to strike an ambiguous balance in making their choices, we made explicit the different sets of interests associated with a decision by assigning each set of interests to a separate person?
So in a manager meeting, for example, a manager would advocate for their reports and the boss would advocate for the company. The boss-manager relationship stays intact for the same reason that lawyers go out to drinks after they’re done arguing opposite sides of the same case; they’re out here doing their job and their relationship is independent of the content of their work.
When this is how managers are taught to operate and how bosses are taught to receive managers, managers don’t need social or political capital to advocate for their reports. Instead, it is explicitly their job.
Why is this good? A couple of reasons.
1. Saves the company a pile of money related to turnover.
In tech, turnover is high, and replacing skilled professionals is expensive. This summary gives us a range that might help us estimate just how expensive it is. People in the United States with the title of software engineer make an average annual salary of $79,357. So for that job, we’re looking at a number somewhere between $15,800 and $168,000 to hire each employee.
That’s a big range; we don’t have an exact number. But we can look to tech industry veteran Michael Lopp for some guidance. Michael has served as senior manager/director/head/VP of Engineering at Apple, Palantir, Pinterest, and Slack, respectively. And in his blog post called “Shields Down,” he explains what makes him cranky about engineers who leave:
The reason I’m cranky is I’m doing the math. I’m placing a cost on the departure of a wanted human leaving and comparing that cost with whatever usually minor situation existed in the past that led to [this] situation. The departure cost is always exponentially higher.
Why so much higher? In order to replace a person who is leaving, a company must source, recruit, screen, and interview candidates—usually several candidates. That takes time and resources from several different departments. Then the company must negotiate an offer and onboard a new person. Companies must also orient and train that person, and they must deal with lower productivity from that person as the person ramps up. This all adds up to a large expenditure.
Given the immense cost of replacing an employee, retaining employees has an unbelievably attractive return on investment value.
Here’s something else that you should know: when skilled employees leave a company, chances are they brought up their reasons for leaving to their managers long before they left. Had those managers been equipped, trained, and encouraged to manage up, they might have had these issues addressed so the person could stay. If you’re a director, a wanted human leaves your company, and you are completely shocked by the reasons they are leaving, one of two things happened:
1) That employee expressed their concerns to their manager, and instead of bringing those concerns to you, their manager elected to ‘manage down’ without resolving the concern until the employee hit their limit.
2) That employee had been ‘managed down’ enough in the past that they no longer trusted that their manager would do anything productive with their complaint, so instead of saying something, they started sending out their resume.
If you want to avoid this situation, it’s worth your time to examine why those things are happening. Are you rewarding managers for not bringing problems to you (“make my job easier”), or are you supporting and believing them less and less each additional time they come to you (charging them social and political capital)? If so, these managers are implicitly trained to manage down, and that’s going to cost you and the company a lot of resources in the future.
An extreme example: This is a true story about an engineer we’ll call Sue. Sue’s manager was one of several candidates up for promotion to associate director. One evening after work, Sue found a box of moldy, wrinkled blueberries in the company fridge; some of the blueberries had plants growing out of them. Sue knew the building had no compost system, so she took the box home with her to compost the blueberries. The office director, who was in charge of choosing the associate director, saw Sue leave with the box, and he asked her manager to deal with the situation of employees taking shared company food out of the office. The next morning, Sue’s manager pulled Sue out of her team’s workflow to have a sit down meeting with Sue and the office HRBP about the seriousness of “stealing company property” and “abusing the office’s trust.” Yes, really.
So what went wrong? The manager’s boss, who would decide the promotion, had asked this manager (a promotion candidate) to take care of something. And by god, he was gonna take care of it. Because the manager needed all his social and political capital to have the best chance at promotion, he couldn’t spare any to say “Mr. Director, maybe that box of blueberries isn’t a big deal and we should just ask her about it.”
That same day, a company Sue had been talking to on and off for a while sent her an email asking if she was open to new opportunities yet. If you think this is a contrived situation, you should know that your engineers receive an email like this, at a minimum, on a weekly basis. You hope they’re not too upset when that email comes in, so they’ll ignore it or shoot off a ‘very happy here, but thanks for thinking of me!’.
Well, Sue was very upset when that email came in. So Sue responded and agreed to meet for coffee. A month later she gave her two weeks’ notice. Can you imagine Michael Lopp sitting in his blood-red office, comparing the costs of Sue’s departure to this box of moldy blueberries? I suspect he’d be livid.
This part of my argument relies on the assumption that reports at your company are wanted humans who you would prefer to keep. So before we go on, let’s talk about the situations where that is not the case. There are three of them: unacceptable conduct leading to immediate termination, unprofessional conduct leading to strategic termination, and unprofessional conduct leading to managing out.
a) Unacceptable conduct leading to immediate termination: your company suddenly finds out that this report did something that violated company policy, and they must be fired now. This one happens so fast that a manager doesn’t need to change their behavior with the report to deal with it. One minute the manager advocates for the report like a wanted human, and the next minute, meeting with HR. So this isn’t really a special case.
b) Unprofessional conduct leading to strategic termination: In this case a company wants to get rid of a certain employee, but they can’t justify it with a violation of company policy. Reasons for this are usually ‘unprofessional behavior’ or ‘performance issues,’ indictments that gets disproportionately levied against employees from marginalized groups. Because the person the company wants to axe disproportionately comes from a marginalized group, the company takes steps to protect itself from a discrimination lawsuit. So they start with what’s called a Letter of Concern to put the employee on notice to stop doing whatever it is the company hates (frequently it’s speaking out about things that are unjust or wrong, unfortunately). If the person still refuses to pipe down, they get the PIP. PIP stands for performance improvement plan, and it contains the company’s order list of what behaviors they want ‘improved’ by some deadline. Usually the PIP isn’t really intended to change the employee. Instead, it’s a paper trail to use if the employee sues to prove that there were reasons to get rid of them and the company did try to address them. Often the directives in the PIP will be pretty vague to allow the company room to subjectively decide, over the course of a few weeks of progress meetings, that the employee is not fulfilling the directives. Then comes the axe.
I’m going to address this situation together with the next one, which is…
c) Unprofessional conduct leading to managing out: Managing out happens when a company wants a certain employee to leave, but they want to avoid inciting a scandal/paying the severance package/feeling guilty. So instead of firing the employee, they make that employee’s experience at the company so undesirable that they leave of their own volition. If managing out fails and the employee hangs on, sometimes companies will haul out the LoC/PIP/axe routine as a plan B.
Let me address b and c together.
First of all, I think the idea of ‘unprofessional behavior’ gets frequently misappropriated to shut people up.
- Making unflattering statements about your old employer is ‘unprofessional’ (how convenient for employers that treat people unfairly).
- Asking how your compensation compares to that of other people doing your job is ‘unprofessional’ (how convenient for employers that treat people unfairly).
- ‘Gossip’ (otherwise known as the grapevine) is ‘unprofessional’ (how convenient for employers that treat people unfairly)
- Being angry or sad at work is ‘unprofessional’ (how convenient for employers that treat people unfairly).
- Calling out bad behavior by someone up the hierarchy is ‘unprofessional’ (how convenient for…well, you get the idea).
If managers were equipped to deal with reports’ concerns in the way I’m describing, though, employees would see traction in their 1 on 1s and get their issues addressed. It’s when employees don’t see a benefit in expressing concerns to their manager that they have to choose between two options:
- Escalate issues of their own volition to try to get the situation fixed. That escalation might mean speaking to people up the chain or, if that doesn’t work, grassroots organizing with other employees to get their grievances addressed. These activities are often how the ‘unprofessional behavior’ gets pegged.
- Paste on a smile as best they can, but mentally or emotionally disengage from their work to cope with their compromised situation. Since their grievance was not addressed, now they create distance to protect their mental and emotional state. This is often how ‘performance issues’ get pegged.
What I’m saying is that we can look at the idea of managers advocating for employees and say ‘well, that only works if the employee is well-meaning, trying their best, and wants the best for the company.’ The thing is, the vast majority of employees start out being exactly this kind of employee. Most people do not sign on to spend the majority of their waking lives undermining an operation to their own detriment.
But when the employee initiates an uncomfortable conversation that is good for the company to address in the long term and they get ‘managed down’ instead of having their issue addressed, they become jaded. So they escalate the issue to try to get it fixed (unprofessional) or they disengage from their work to avoid despair (performance issues).
In other words, managing down doesn’t address or solve situations of unprofessional or poorly performing employees. It creates them.
That’s why it’s important for managers to make a habit of checking in with their employees. Are they bored with their work? Are they feeling isolated from coworkers? Are they working closely with someone who keeps smothering them with microaggressions? Did they figure out that the person they’re mentoring makes $15k more than they do? Were they hired with the intention of ‘upskilling’ them, but then thrown onto a project that’s way above their heads and their senior colleagues not making time to mentor them? If a manager identifies an issue, they need to have support from their directors to address the issue. The more that this happens, the more trust your managers will build with their reports. That’s the next benefit to tasking managers, specifically and exclusively, with advocating on behalf of their reports.
2. Preserves trust up the management chain.
Lawyers encourage their clients to tell them everything—even the incriminating stuff—because knowing the whole story makes it easier for the lawyer to advocate for the client most effectively. This is helpful for the manager-report relationship as well. It’s an especially helpful shortcut in the modern workplace where things like anger, sadness, envy, feeling sick, not being as productive as usual, and other unideal circumstances are ‘incriminating’ for employees. Your company is not a year away from changing the corporate culture that makes employees nervous to share these circumstances with the company. But your company might be a year away from employees trusting their managers with those circumstances in spite of corporate culture. If employees trust that their manager’s job is not to penalize them, but rather to help them get the support they need to fix underlying issues, companies protect their employees and their bottom lines.
Additionally, this arrangement relieves pressure on managers. Currently, a lot of managers are already advocating for their reports, but it’s expensive for them because they have to spend their social and political capital to do it. That sucks for them, and it means the higher ups won’t hear about problems until managers have to share them—at which point, it might be too late. Additionally, even when managers do advocate for their reports, they cannot talk about their upward communications with their reports because they’re expected to sound like they’re toeing the company line. Even as they’re advocating for their reports’ interests, their reports aren’t hearing updates so they think their manager doesn’t care. This is a great way to burn out your managers. It would be much more gratifying for them if they could honestly share with reports ‘I brought it up in the manager meeting, and they need to see if we have room for it in the director level budget.’ When managers can be honest about their advocacy, they can engender trust with their reports.
I am aware that this point, in particular, brings up a possible internal validity concern in my proposed model. So far in our examples, we have assumed a three-tiered organization: director, managers, reports. But with more levels of hierarchy, managers who are also reports would end up in a position of not only representing their reports to their manager, but also representing the broader company to their reports. So even though they’re advocating for their reports in the background, their assigned role when talking to their reports forces them to argue on behalf of the company. The reports might not know that the managers are advocating for them and might think their managers don’t care, which would be sad for both parties.
In my experience, this already happens a lot. So here’s my proposed solution, which I caveat with the fact that I have not seen this solution in action: in this situation, the manager should express the company’s concerns as a representative of the company, but should reassure the report that, as an individual, the manager is on their side. Keep in mind what we talked about before: the vast majority of employees are well-meaning, trying their best, and want the best for the company. So whatever concern it is they have, it’s not a prosecution in the same way a court case is, and they should not be met with a stonewall defense. The report is trying to get things fixed—not bring the company to its knees.
This is true all the way up. A report brings up an issue regarding unrealistic deadlines, for example, to their manager. It’s in the company’s interest to set the realistic deadlines so clients, employees, and project managers can believe the projections. The manager takes that to their director. The manager advocates for their report’s idea—to have more realistic deadlines—so that everyone can believe the projections and so the report doesn’t feel like they’re spinning their wheels at work. The director responds that they know the higher ups are trying to push out more work faster and they think that people allow work to fill up the amount of time that they have, but they can take the evidence from the manager and send it up the chain, and they’ll report back. Now the director goes to the vice presidents and explains the situation. The VPs can say, look, we need this much dealflow in this amount of time to hit these revenue numbers, but given this evidence from below, maybe it’s time to consider ways to do that that don’t squeeze employees. Together they decide to hire some more people to make the deadlines easier to meet. Issue addressed, nobody stonewalled anyone.
Managers as advocates has third important benefit for companies.
3. Helps the company make better decisions.
You don’t have to take my word for it. I’d recommend checking out a copy of Originals: How Non-Conformists Move the World by organizational psychologist Adam Grant. I cannot offer a unreserved recommendation for the entire book, but I can suggest a skim of the chapter called ‘Rethinking Groupthink.’
The mainstream conclusion among modern management literature coalesces around this idea that homogeneous groups of people mostly agree with each other and miss alternative approaches to decision-making—and, as a result, make a poorer basket of decisions than a diverse group. This makes sense: you need different perspectives in the room to see all the sides of an issue clearly.
Thing is, diversity is necessary but not sufficient, because all the perspectives in the world don’t matter if the leadership isn’t hearing a chunk of them, either because the leadership isn’t listening or because people don’t feel safe to express disagreement with leadership. In ‘Rethinking Groupthink’ Grant talks about this, then goes on to praise the extreme approach of billionaire investor Ray Dalio who, evidently, fires employees who don’t challenge his ideas (my reservation recommending this book centers on my hesitation to condone extreme approaches like this one).
You don’t have to axe employees who don’t start beef with you. But in order to optimize your decision-making strategy, you do have to solicit all the perspectives available to you in your company. Particularly if your company is big, you can’t do that alone. But you do have some soliciting leverage—your managers. Your managers can collect, share, and advocate for the perspectives of your entire company, if you make it safe for them to do so without losing their own social and political capital, and if your employees trust their managers enough to share dissenting opinions with them.
Hopefully you can see how structuring the roles of managers as advocates might benefit your company’s trajectory.
This is not a new approach
And also, lest you think I am a loony IC hawking unproven ideas about my pie-in-the-sky management structure, I should mention that I (and you) have seen this approach in other contexts.
First of all, it is how the American legal system works (as well as legal systems in the UK, Wales, and Canada, and in France for criminal allegations). These legal systems aren’t perfect, mind you, but there’s a reason that they don’t ask lawyers to strike a ‘balance’ between defending and indicting their clients. Instead, for example, Americans’ federally granted Miranda rights include “You have the right to an attorney. If you cannot afford an attorney, one will be provided for you.” That’s pretty explicit. To make it even more explicit, law enforcement is required to say the above to everyone they arrest, in case they didn’t know. It is done this way to give the court the best possible chance of issuing a just ruling. It is also done this way to prompt appropriate changes to the law. The law, and its interpretation, change all the time in response to precedents set by cases where a ruling went one way or the other. Having advocates on both sides is a key safeguard in making sure that the precedents have weight and are worth consideration.
Second of all, this is also how representative governments are meant to work. Each representative in the legislature/parliament advocates for the interests of their constituents with the idea that, together, they’ll find the solution that works best for all of the constituents. Alaskan senator Lisa Murkowski isn’t in the senate to strike a ‘balance’ between her constituents and the needs of the White House. She’s there to advocate for Alaskans, period. The idea is that all the state advocates together will collectively strike the right balance for the country. Again, I’m not saying that the system is perfect. I’m saying that it’s set up the way it is for very well thought out reasons. Your management strategy can benefit from the thought and planning that goes into the running of things far larger than your company—like countries.
But it’s not like tech’s current management model is working…
The industry is notorious for high annual turnover, the highest echelons of leadership regularly make jaw-droppingly bad decisions, and half of the women get driven out in five years. We’re not #winning, y’all.
I’m not saying there’s a silver bullet. But I am saying that, since an employee’s relationship with their direct manager is the most important thing to them besides the actual work, adopting a structure that maximizes that relationship’s chance of success is the closest thing to a silver bullet that we’re gonna get. It might address the turnover and the industry toxicity. Also, if managers are funneling employee perspectives upward, maybe some of the how-did-you-seriously-not-see-this-coming moments on behalf of senior leadership would get seen before they come.
So we might as well try it.
We ask managers to independently strike a balance between advocating for their reports and advocating for company interests. We do this despite the fact that the balance is uneven, and we penalize managers’ social and political capital when they do advocate for reports. This structure is fair to neither managers nor reports, and it contributes to high turnover and the propagation of poorly-considered decisions.
If we empower managers to advocate on behalf of their reports, we can externalize the process of finding a balance that better reflects competing interests, better represents the needs of the workforce, and better capitalizes on the diversity of perspectives available for making decisions. We can save the company a lot of money related to turnover, improve trust relationships through the ranks, and end up with better decisions from leadership. We have seen this model in other places, and it’s not like what we have in tech is working. Why not give it a try?