Ask HN: How do big companies ensure the execs get the right info from below?
I am looking to find out more information on various ways large companies (>500 engineers) ensure the right information flows from junior members to executives.
For example, usually every junior person knows a lot about the project they are in charge of/involved in (and maybe how it aligns with the company’s strategy) and very little about other projects within the engineering organisation.
On the other hand, for an executive, it is important to be aware of what projects employees work on (with much less detail) in order to assess which areas are working and which areas need changes.
I am curious what methods larger companies employ to ensure the right level of information flows up the chain.
The short and true answer is that THEY DON'T. Why is this? Because of the SNAFU Principle: "True communication is possible only between equals, because inferiors are more consistently rewarded for telling their superiors pleasant lies than for telling the truth."
Junior people almost never give senior people the unedited truth because doing so is a good way to end up unemployed. Instead, junior people figure out what senior people want to hear and tell them that instead.
https://foldoc.org/SNAFU%20principle
Ha, this is absolutely true, especially in more traditional companies dominated by a patrician management.
I still wonder what happened to Randall Stephenson when AT&T decided to buy DirecTV for $67B in 2014. What could only be described as thinking completely in an isolated bubble. It seems that the survival of traditional companies are based on the sheer incompetence of competitors to continue allow the swallowing up of disaster after disaster committed by the management.
Or more likely they are so big that even decent sized fuckups every now and then don’t make much difference.
For the mega corps, they could shave half their workforce and still be making the same amount of profits. Some companies just want to grow for the sake of growing and invent new problems to fill time.
I have a fortunate circumstance where supplying informations to direct reportee is acceptable.
We value effective problem identifications, especially those that come with solutions.
> it is important to be aware of what projects employees work on (with much less detail)
In my experience, employees get told which project to work on by management, usually through a formal budget / roadmap process which allocates resources to the initiatives being pushed by different stakeholders such as marketing, sales, IT (e.g. for maintenance projects), etc.
So the need is not so much to know what the employees are working on but rather how the projects are going. This is usually handled by steering committees or similar governance bodies, where each successive hierarchical level gets a broader (but necessarily more concise) overview of all the projects they are involved in.
What makes it through these different sieves depends on the company culture and what is considered the "right" level of information. In the worst cases, bad news will be edulcorated, then removed, until every line on the dashboard is green. In more result-driven companies, what makes it to the top is what is susceptible to impact business, so it's going to be the project delays and other bad news. But this process is like the "auto-summarize" feature in Word: a lot of nuance gets lost in the different layers and there's alsways a risk of ending up with nonsense. There's also a lot of mediation being done by middle management.
The closest I've come to direct feedback from the junior employees to the execs at large (>10k FTE) companies is when execs tour the office and come take a look at the grunts in the trenches (this will happen once or twice a year). Inevitably there's an awkward roundtable with a chosen team (usually the one working on the most visible project) where the exec asks each person in the room "what is blocking you and what can I do to help you" while that person's boss and grand-boss glare at them from the back of the room, silently promising retribution to anyone who raises any actual issue.
>So the need is not so much to know what the employees are working on but rather how the projects are going. This is usually handled by steering committees or similar governance bodies, where each successive hierarchical level gets a broader (but necessarily more concise) overview of all the projects they are involved in.
I agree this seems to be a common pattern. You do run into multiple issues with it (though there is no perfect solution) - some of which already described by pg in one of his essays (http://paulgraham.com/boss.html).
I tend to think the quality of information is maybe inversely proportional to the levels it has to go through to reach the intended recipient as each layer will inevitably add noise/distortion.
> the quality of information is maybe inversely proportional to the levels it has to go through
I think that each level distorts the message to bring it more in line with company culture (e.g. make it look better, or highlight blocking points, or focus on financial topics, etc.). It follows that the quality of the information doesn’t necessarily have to decrease - in theory, the right company culture will ensure that only the most relevant information is kept (and I’m pretty sure you can tie this to successful company traits)
More generally, I think that in this context you can’t treat information quality as a purely objective measure. Here quality is (at least partly) in the eye of the beholder, so you could argue that making the reporting more in line with the expectations of you higher-ups actually increase quality.
Different large companies do things in different ways. Some companies do things very top-down to the point where "execs getting the right info" is merely "some report down the chain seeing something or other has to be delayed". Other places are a bit more open. Depending on the exec in particular, they may spend a lot of time reaching out to random individuals/reading planning + update docs/etc.
In general, I'd say it's part of the exec's job to ensure that the right sort of culture is set where individuals feel empowered to surface relevant information. This includes trying to make sure that people feel safe enough to surface issues, getting rid of people that aren't good about responding to feedback, etc.
As a corollary of this, you can tell when an exec is bad since not only do people snark, but the same topic of snark lasts for years and years.
There's also a part of their job to have the right organization structures in place such that if a piece of information should really get resolved at a lower level, it does. If a lot of information is having to be resolved by an exec, it probably means an organizational refactoring should probably be happening in the first place.
The “right” info is the challenge. A company with >500 engineers will also have at keast that many sales, marketing, bizdev, legal, facilities, HR all put together. At least 1,500 people. And esch executive needs to handle communications laterally and from above. So “right” info is challenging to define from all the possible info. Good lower management will be able to prioritize necessary and useful information. Bad lower management will too, just with different ends in mind.
I don’t think I’ve seen good examples, except in stating that there’s a culture of transparency and inclusion and then not shooting the messenger that points out things that need improvement. Just describing that it’s not easy and likely not possible to get all of the right info, only some of it.
I think what you say makes sense.
I do think good lower management helps, though I do wonder if there are certain processes or practices that a company can adopt to maximize the chances of 'right' information making its way to the relevant parties.
These days, it seems that people's terms of choice are OKRs, though I think these are useful if clearly defined and updated regularly. However, if they aren't defined well, they have the potential of getting people to focus on the wrong thing.
Charles Duhigg's The Power of Habit covered how an aluminum company did it.
They made safety absolute highest priority. It was a goal that the unions and managers didn't disagree on. The processes were designed to empower lower ranked people to override when things were getting dangerous, and the CEO had to know within a few hours whether an accident happened. Long story short, this put in the processes to let information flow up the chain rapidly.
Again, the key is the right information. It's possible to have all kinds of junk go up when everyone can talk to the CEO. Managers act as abstraction layers for information for a good reason.
The story also says that once the process was in place, other information like optimizations went up the chain and profited them millions. But you want to set a standard that only really important information should go up this path.
It can only be as accurate as the employees reporting to you. Trust in your subordinates to collect and synthesize the right information that will produce the desired results. This strategy relies on the leader being transparent with their team around company and org goals, short comings, etc so they can properly report and frame data.
It seems they do it completley in the dark. They think they have information, but they don't. As such, after many "we are doing great!" top level presentations, they just announced layoffs and selling of some divisions and so on. That's how "great" we were doing.
Execs on high level may get this information by weekly or biweekly one-by-one meeting with junior employees. The higher you get the harder to get this information...
I think one main task of executives is to build up the formal and informal information channels to gater intel for decisions.
I’ve had success by attending executive customer briefings and providing feedback to the execs afterwards. Of course you won’t get invited to such meetings as a junior. Happens more at a senior technical level.
Not coming from a big company.
However, to mitigate these kind of problem, pick a junior having a more or less same vision as you, plant the person into the team you wish to spy, and have them report to you.
I'm guessing that they don't.
https://www.youtube.com/watch?v=gfWpYwHlwtI