CEO Shadow Program
about.gitlab.com
about.gitlab.com
The experience really opened my eyes about the requirements for the job. The amount of code-switching required to communicate well with ownership, the other C-levels, employees, customers, and potential acquirers could make my head swim. The breadth and depth of technical, financial, and human expertise necessary to verify all the parts of the business and smell out where issues might be hiding. All of that before even getting to (in this case) integrating several businesses into a whole, identifying up-and-coming leaders, and keeping the employees happy; let alone actually having a strategy, getting buy-in, and implementing it. And then spending most evenings networking with the elite class that you need to know to have access to capital. You can never have a bad day. Ever.
This was for an org that was comfortably under 1000 employees, all located in the US, where none of the VPs were the kind of political operators you get in meatier companies.
I walked away thinking that there can't be even 100k people in the country with the ability to do even a non-laughable job as the CEO of a 500ish employee company. Not even do a good job, just do a job that doesn't look like one of those 80s child-parent-bodyswap movies where the kid tries to work in an office.
I still think CEO pay is out of control, but I understand better why boards end up paying what they do. Any weakness in any of these skills is an existential threat, and you don't know for sure what will hit you this year.
Feigned/false strength might be much more disastrous.
I can think of a few household name companies whose CEO was widely lauded during the time, but might have fared better had a tree trunk was there instead.
"The New York Times business columnist James B. Stewart wrote in 2017, “Hardly anyone considers Mr. Welch a management role model anymore, and the conglomerate model he championed at G.E. — that with strict discipline, you could successfully manage any business as long as your market share was first or second — has been thoroughly discredited, at least in the United States.”
The financial crisis of 2008 delivered a blow to G.E.’s fortunes. In the years before the crisis, the company had built sprawling lending operations that helped drive its growth. But the finance businesses became a crippling liability when, during the crisis, credit markets froze and borrowers struggled to pay back their loans.
G.E., like many large banks, tapped emergency government loans to help it get through the upheaval. In the years following the crisis, G.E. sold off most of its lending businesses, but other problems emerged, some of which were in its large power unit.
G.E.’s stock price now trades roughly 80 percent below the high it hit in 2000. The company has significantly lower revenue than it did that year. Last year, G.E. reported a loss of $5.4 billion."
See https://www.nytimes.com/2020/03/02/business/jack-welch-died....
Whether it's rank and yank, diversification into the hottest financial markets, throwing away manufacturing because those weren't making a lot of money, killing of R&D because they are a cost center. You'll see so many "capable" CEOs who could handle the day to day gracefully but manage like this, and lead the companies to their doom.
It's not about appearances.
> Any weakness in any of these skills is an existential threat, and you don't know for sure what will hit you this year.
Hard disagree with this conclusion. I regularly work with businesses from 100~1000 employees and at a certain scale (number of employees might not be a great proxy), the engine of the company simply takes over - good or bad CEO.
There are great CEOs that can do ALL of the things you mention, but I've also sat in board meetings with some pretty terrible CEOs of $100M revenue businesses. To the point where I would seriously question how the company still was operating, but alas, it chugged along.
It's actually incredible to me how many businesses that do have poor leadership that continually survive. Once you've built a brand, a reputation, a network of clients and a somewhat steady book of business, it's surprisingly hard for most businesses to fail.
Note - tech companies on a VC track are a bit of an exception to this, but thats usually because they are spending more than they earn, so they literally strangle themselves. It's important to remember in the grand scheme the number of VC backed companies is greatly dwarfed by the number of businesses that exist.
I suspect this is the case when there are competent people at other levels in the org. It seems like the rest of the company is often aware when leadership is bad and each push back against poor decisions in their own little way to try and keep their parts of the business running smoothly.
P.S: I'm one of those individuals.
One exec in particular:
- Made power grabs for other execs' areas, forcing those execs out to other organisations.
- Continually shifted goal posts for key technical staff. In the words of one of these people, this was the primary cause of their mental breakdown.
- Filled positions with his own cohort of sycophants and yes-men from previous organisations.
- Defunded major projects, turning them from promising potential businesses into graveyards.
After just a couple of years, the organisation was a shell of its former self.
What does that mean? Sounds to me as if you were thinking that the CEO might in fact have been working for a competitor (I didn't downvote, just curious)
In my experience the CEO does three things uniquely that other roles don't:
1. They are the ultimate authority in tradeoffs between internal interests; 2. they are the interface between ownership and the company; and 3. they are the driver of strategic change.
When the whole company is aligned and rolling downhill with product-market fit around a single offering, those functions aren't as necessary. It just works. When you start branching out, you need a final referee that can determine the exchange rate between the desires of internal kingdoms.
But it's the second and third functions that IMO drive the executive compensation bubble. Our economy is kind of hourglass-shaped, with an economy of wealthy asset owners barely joined to an economy of consumers. The thing that is supposed to tie everything together is that the value of the assets owned in one of the bulbs is tied to DCFs of the money circulating in the other bulb. The CEO is at the narrow opening between the bulbs, mediating between the investors in the equity-class-relationships economy and the actual operating business in the actual-humans-buying-things economy.
There has always been a "strategic" layer that insists upon its own inevitability, the BCG/Bain mindset that says that things can't just run on autopilot and someone needs to be looking Towards The Future. For any given company there's going to be someone out there who has an investment thesis for how that company could make more money with some changes, using whatever that decade's version of the Cash Cows/Dogs/Stars matrix is.
What's different (IMO) is the massive amount of inflation we had only amongst the asset owners, that hasn't until recently been matched by inflation in consumer sectors. Valuations have gone up simply because there is so much money to invest, and investors have been given basically two options: keep your money in cash because equity is over-valued, or believe someone with a thesis of how a company with a NPV of $300M can actually be worth $1B with a few changes. The CEO's job is to prevent the uninflated DCF from fully decoupling from the inflated company valuation.
So we get all sorts of businesses that are doing fine for what they are, but not doing fine enough to justify their new valuation. At a macro level it's driven by QE2 and other Fed schemes, but for this one particular company it's an intriguing and reasonable idea. Now the investors put the CEO to work implementing the thesis that will will justify the valuation premium. I think that's the driver behind a lot of the exec comp and the flailing about.
Thanks.
I was once told once a CEO does three things: (1) sets the vision of the company, and (2) manages the energy to get there and (3) coaches subordinates to execute on the agreed upon vision.
(Corollary: I also have a ton of respect for technical people who become successful CEOs, like Jay Kreps, Peter Zaitsev, Aaron Skonnard. I have a high opinion of my own abilities and I don't think I could do it.)
- Their calendar is often broken up into 15 minute meetings
- Lots of external meetings because a big part of the job is "being the face of the company"
- The 15 minute meetings are often just a final "sign off" on some important decision; CEO bring the company-level perspective and can highlight key issues identified by the board
- What a department might spend 3 months working on, the CEO often has to understand and make a decision on in 10 min (see prior HN post "how to communicate to executives")
- What people presenting to the CEO thought was important was often very different than what CEO thought was important (e.g. team thought product launch price was critical while CEO was worried about a $500M supplier contract that was in the final stages of negotiation and if this product launch could be leveraged to close the deal and save the company millions)
- The political side was very important; HN likes to think raw intellectual horsepower matters, but ability to communicate and "get people on your side" is at least 50% of a CEOs job; your idea might be brilliant but if you can't convince the sales team of it, your idea is worthless
It's gotta be 50% of every job. Heck, even on open source projects, you need to be lucid and explain what you are doing.
Look no further than for example, the ANTS project, which has done some very cool work in the plan9 space but hasn't always gotten a good reception due to the author being somewhere between Terry A. Davis and a normal programmer. And it's in the communication that's key.
And many of these aren’t people who will shy away from telling you “no, if you do that I’ll just undermine your idea so my idea prevails”.
Can you share a link?
Many CEOs are, of course, very well compensated. But, whether or not you think it's extraordinarily difficult, it's certainly very demanding in terms of time and schedule.
Now imagine some large company and I'm not that shocked that the pool of people who can effectively lead large (>10,000 employee) organizations is relatively small. It's basically a small subset of those who can lead small organizations.
And not to say unqualified people don't end up in those roles, they do. But generally I'd argue the qualifications of a CEO are likely as scarce as say the qualifications of a neurosurgeon or some other hyper-specialized technical role.
Large companies can accomplish things that smaller companies cannot but keeping the ship reasonably on course is a really hard organizational problem. When I was an analyst and spent a fair bit of time with senior execs at tech companies, I often saw how much fragmentation and lack of coordination there was among different groups. But my wonder wasn't that this existed but that the companies were able to function at all.
But each of the C Suite members can't spend all their time staying up-to-date and coordinated on everything each other is doing, and they'll also often disagree.
So you need somebody to manage the C Suite, understand and integrate and make decisions... and that's the CEO.
At the end of the day, somebody has to be in charge, or else it's either a) anarchy or b) incredibly inefficient as too much time is taken up in NxN communication and lengthy group debating and decision-making.
And somebody who's capable of managing every C Suite member effectively, whose areas span the entire company... that's going to be an inherently fiendishly difficult and complex job.
maybe simple supply and demand?
I think you're right about the number of people who could effectively do those things. I think you're missing a key skill that a CEO can have that can make them effective if they're lacking in one or two key areas if the company is big enough: delegation.
I'm a generalist and have some leadership skills (nothing like what would be needed to CEO a company of a thousand employees, but I could do a decent job with a dozen or two), and one of the first things I do in team situations is figure out who in the group is good at things I am weakest in and whether or not I can trust their judgement/work ethic/etc. For example, if I'm working on/helping to manage a tech project, it's probably not a good use of my time to do get into the technical weeds even though I can do so, because that is something that a good, trustworthy CTO can do for me, and it's easier to find someone with technical acumen whose judgement I trust than to find someone with communication acumen whose judgment I trust.
Team and bench-building is a key component to managing big companies well, particularly given part of reducing risk is reducing single points of failure, including making the company overly reliant on one person to function.
This is also why somebody's network/reputation is a big deal - it signals what 'bench' they might have available if you choose them as a leader.
So while it's very true few people can solo lead, there are a decent amount of people with ~80% of the skills needed and a network that can cover the 20%, so it's not completely dire.
https://www.lesswrong.com/posts/CKpByWmsZ8WmpHtYa/competent-...
> Next we met with Vinod, Bruce, and investment partners Brian Byun and Sven Strohband for a company brief and brainstorm session. Sid began with an update on the financials, and detailed our massive growth and expansion both in terms of people and product. There were no slides presented. Instead, Sid used our website as a visual aid. Nearly every question or discussion point was first addressed with a Google search to pull up the appropriate GitLab web page to reference.
[...]
> I’ve been working at GitLab for two and a half years as a content writer on our marketing team and at times have been extremely frustrated with our marketing website—the content on it, how it’s organized, what we’re presenting, etc. It doesn’t look or operate in a way I’m familiar with so my instinct was to not trust it. But nothing we do at GitLab is “normal,” and witnessing our CEO use the website as a single source of public truth to inform our investors is just one example of what it means to be a transparent company.
[1] https://about.gitlab.com/blog/2019/04/08/khosla-ventures-git...
Because I find it faster to switch by just spotlight + app name, in many cases even faster than command + tab, because it's kind of an O(1) switch, vs an O(n) switch to cycle through many open apps.
This is also the reason why both mouse-clicking and touch-tapping are slow compared to keyboard typing: you likewise need to look at what is where on the screen, then aim with the mouse or the thumb, instead of just jerking the fingers to keys that are in muscle memory.
To make things worse, I tried switching back to Alfred but it it seems to just piggyback off spotlight for finding apps.
Otherwise I would switch back to Alfred if it just had a “my index doesn’t get corrupted like spotlight” as a feature.
Otherwise, apps can conceivably be added as keywords or something. Rather annoying since it won't pick up new apps automatically, but oh well. Myself, I prefer Alfred anyway since Spotlight doesn't do too much aside from launching apps, converting currencies and occasionally looking for files (the latter is not so useful when there are thousands of third-party code files on the machine, and Finder is not my IDE anyway).
P.S. Also btw, at least with Alfred, I don't need even a full word for most apps—just two to four letters: i.e. ‘fi’ is Firefox, and ‘do’ is Double Commander. Alfred quickly ranks results higher if they were invoked previously for the same input, so the desired app pops up in the results just after a few letters, and this behavior sticks around without perturbation from re-indexing. This also means that at first it pays to enter those couple letters and select the app with the arrow keys, since Alfred distinguishes between inputs of various length—i.e. might learn for ‘firefox’ but not for ‘fi’.
I could well imagine that avoiding this one pitfall raises productivity enough to make up for any potential advantage that competitors have.
And in quarter of the meetings they do join they are asked to leave towards the end so that there can be a short closed session.
We also have extra confidentiality restrictions for the shadows.
when i approach life with truthfulness and transparency, i feel light as a feather.
I will have to remember that phrase. I once worked for someone like that. It was awful.
• The leader was the sole and absolute authority on everything the team did.
• If a team member came up with an idea of our own, we were not allowed to write a ticket to open it up for discussion. We were expected to have a private discussion with the leader first, and they would decide if it was worth pursuing.
• Heaven help you if you ever wrote a balanced comment citing the pros and cons of any particular approach to a problem. The leader would quote you out of context in a team meeting, reading aloud only the parts of your comment that supported their position.
• Secrecy was important. For major new projects, the leader would convene a private working group with hand-picked team members who would follow their lead, excluding anyone with more experience who might have insights or even disagreement on the development approach.
• If you ever made a mistake that caused problems in production, you would get called out for it in front of the team. If the leader made a mistake and you tried to respectfully help shore up the problem, they would dig in their heels and argue against your fix. The leader does not make mistakes.
• In that latter case where the leader made a mistake that caused problems in production, they would never acknowledge it or apologize for it. Apologies are a sign of weakness.
Never going back there again.
My previous boss did exactly that plus a bunch more psychopathic stuff like:
- Praise someone in public before they leave, then block them from coming back - Promote the incompetent to further solidify their power position - Lie to customer, stakeholders, etc. Everything must present a good "image". The work or actual results were barely relevant - Have scapegoat at every level in case of failure. - Refuse to join status update meetings so that he could claim not to be involved with things that weren't going well and sacrifice a scapegoat
> CEO shadows label the handbook MRs they create with the ceo-shadow label. It's a point of competition between CEO shadows to try to best the previous shadows' number of merge requests.
I am no longer surprised by the length of this section of the handbook.
I see them often on slack updating various random pages:
https://gitlab.com/gitlab-com/www-gitlab-com/-/merge_request...
However, its "open source internal practices" model is the gold plated standard for a 21st Business.
If you're any kind of manager of anything, do this.
See: http://blog.idonethis.com/managers-write/ for more.
In terms of the actually forming a business product market fit, they definitely have.
However, they're seen as a cut price second supplier to GitHub Enterprise + Atlassian. Not an ideal situation.
In terms of "maybe we should become more user friendly instead of going all in on enterprise features," not so much. And that's important.
Citation needed? Perhaps you see them that way, I don't...
Atlassian is seen as a true nightmare.
I hope some small national or regional government takes the lead and also adopts similar openness practices, whenever possible.
It's pretty darn cynical bc market fit is pre-established. Must be frustrating for folks coming from the technical innovation side of their market.
Ultimately, product market fit isn't lawfully protected intellectual property.
Samsung Austin Semiconductor used to have an "eat lunch with the president" program where groups of four or five from each department would each lunch with the president at least once a year. Obviously nowhere near as immersive as this program, but still having the opportunity to ask several off-the-cuff questions like, "Why did Samsung choose Austin?" definitely added value for me.
From my perspective, the real value of the lunches was strictly building rapport across the company. At some places high level employees (doesn't necessarily have to be management) often feel unapproachable and they feel more abstract concepts than people. Having lunch with them re-enforces they are also people, people you may even get along with.
Other companies I've been at only do Q&A sessions with large groups of employees, and this is still useful, but the format doesn't help the culture. The answers you get are often the company line, and you cannot dig deeper. The people answering the questions are still abstract identities that appear to tout that company line, with no real personality.
However, being a TA also involves active advising and commenting first on document reviews (a lot of executive communication at Amazon is via written documents and review meetings that begin with a doc read). Many of our executives were former TA's.
CEO shadows an employee all day. Get a feel for how the peons live.
Understand why it is so hard to get things done in the company.
If you're not a CEO, you can learn a lot from shadowing your CEO, without impacting the viability of your company. But if you're the CEO, all you can learn is "how something seems to work at a level you are not supposed to have control over." Unless your company has an incredibly flat org, without a board that the CEO is accountable to, and where the CEO is actually makes direct hiring, product, etc decisions, realistically having the CEO shadow employees means that they're not being the CEO. The larger the org, the more people they'd need to shadow to get meaningful insight, and the more the board will go "we need you to be the CEO of this multi-hundred-million dollar company, so if you want insights: you should already have a team in place for that."
The CEO and his favorite executives and VPs salesmen live inside their own little bubble with little real knowledge of the problems and challenges that the company is facing internally.
This is why I made good money.
It is easy to see this when management sends out decrees on how to improve performance or throughput .. etc that does not contain any mention of, nor solutions to the problems that is keeping the performance / throughput /(or whatever fitting metrics are for the company)
I might have a bias in that I have mostly worked with companies that hired us to help them. It is possible that the pool I have been exposed to is different than the majority but I do not think so.
I know of companies that are run in a more sane way but they are usually smaller, at least in number of employees and thus leadership is more aware of how things work.
This is broken or non-existent at many companies.
Couple of years ago the idea of CEO Shadowing was mentioned on Kernal. https://kern.al/idea/shadow-a-ceo-lessons-courses-programs Since then no one picked it up, not sure why. But I'm interested and want to make a real living application for CEO and who want to be CEO Shadow.
Is it worth it?
Since then no one picked it up, not sure why. But I'm interested and want to make a real living application for CEO and who want to be CEO Shadow. Is it worth it?
I think this role is good though. We have similar staff that been included in all the things and they got up to speed fast.
And a bit from the shadow updating the online docs, but maybe that's not for the CEO.
Edit: So you're downvoting this comment? Some feedback? How is it not surveillance with extra steps?
Just because someone is watching you doesn't mean they're surveilling you. When I mentor coworkers by letting them watch me do a task, it's not chilling me or oppressing me.
People are downing you because you're making a rather absurd conflation between "they aren't working in total secrecy" and "now it's an oppressive police state"
You made a snide comment without enough content in it to fuel a decent discussion. If you want a discussion then it's on you to initiate it properly.