The 35% reduction refers to the number of managers who oversee fewer than three people, according to a person familiar with the matter.
If you oversee 0-2 people, in most cases that’s probably not an efficient ratio. How did Google get so many folks in that position in the first place? And I assume the other 65% take up the slack to fluff their teams? Or what? Leave the other 65% managing 0-2 people?
Around 5 is the correct number for a first line manager of a technical team. Go to 10 and it’s impossible to keep track of things. The day has only so many hours. Managing takes time.
For bigger teams (10+) you either need individuals who are very independent and driven, or have dependable line managers.
That's been the mantra in the tech world for at least a decade now, and I've grown to really disagree. Managers should have 10 at minimum. Most should have 15-20. They should specialize in being a manager and shouldn't be solving problems except pure management problems. (Good) teachers easily manage classrooms of 10-30 kids, even in private schools. The low direct report count ends up creating managers that have too much power, too much say, and too much time to mess around.
This was called the TLM role at google. Technical Lead/Manager. You were expected to code and manage a couple of more junior engineers.
It’s part of an effort to have dedicated managers and dedicated engineers instead of hybrid roles.
This is being sold as an efficiency win for the sake of the stock price but it’s really just moved a few people around with the TLMs now 100% focused on programming.
Lol I went through a few FAANG Pm interviews for the lols.
Ah trust me - its actually insane how every firm manages to pull this off - the secret sauce that made the company what it was evaporates and the management optimise for process.
I’ve never worked anywhere where managers added value, in fact the best places I’ve worked are where the product people have very little power over what the technical team do and instead of specifying what they often specify why, giving the team the opportunity to suggest much simpler solutions.
I met a long time Google employee this week interventions that most of the senior management were ex oracle people.
It’s nearly 20 years since Google had a category defining product - they haven’t built or acquired a single thing that dominates in the same way that android, maps, search, docs, etc. has since about 2006. It figures.
I never really understood the concept of small teams. Managing a small team really does not provide the scale and benefit that a medium to large team does. Lost bandwidth of manager of said small team or extra salary of the same manager seem like something the company could use in other places.
But often such teams in faang come up as a by product of someone’s empire building and that is unfortunate for others involved in it.
Note to the publisher: when I was about 2/3 through the article the article disappeared and I was scrolled to the footer. When I scrolled back to the top there was only the title, key takeaways, and about 3000px of Taboola. Bad form.
can someone with experience doing this shine some light? i have been offered this type of role from engineer to 50/50 (as i feel it) or 80/20 (as they say) IC and managing. in a series C startup. i feel like it’s never good to context switch. i never seen a tech lead or manager who did well both roles at once. am i crazy to think that the tech lead or manager role should be 100%? either go the IC track or the manager track. but i lack evidence to substantiate this idea of mine.
Im convinced the big tech firms are so overly bloated simply because they do not possess high-quality leadership at the top who are able to clearly distill a vision of where they want their sub-ordinates to go.
That's not to say it's easy - its absolutely not. But Apple is living off of Steve Job's visionary prowess and continues to do so.
Fewer managers with fewer direct reports seems impossible. If you eliminate managers, then the remaining ones will naturally absorb the direct reports. So one of these cannot be true. Unless people don't report to anyone.
I'm curious, what did the management part of the TLM role entail?
As someone who used to be an engineering manager, I was always surprised at how inefficient the division of responsibilities seemed to me. I mean, when I was an eng. manager, a substantial portion of my time was just taken up by logistics - like when we did a move to a new office building, a huge time sink was stuff like seating charts. Perhaps a company as big as Google has more folks taking care of stuff like this, but I still think the following breakdown makes sense:
1. A "technical mentorship" role: someone who codes, but is also explicitly responsible for skills growth and technical feedback of ~5 junior engineers. This person would not be responsible for stuff like salary/raise negotiations, promotion decisions (but would obviously feed into that, more on that below), logistics questions, etc.
2. A "directing manager" (obviously that name kind of sucks, but I didn't want to confuse this with other "director" or "manager" terms). This person explicitly does not need any technical skills. They are responsible for all logistics/salary negotiations, etc. They would be responsible for around ~5 technical mentors, so then up to 25 people under that. Promotion decisions, for example, would be made amongst the 5 technical mentors, deciding who on the team is most deserving to move up. But then the actual salary decision would be made by this "directing manager".
I'm sure this could be tweaked, but the overall idea is to separate technical vs. non-technical skill sets more efficiently.
> The 35% reduction refers to the number of managers who oversee fewer than three people, according to a person familiar with the matter.
I wonder why these people are made managers in the first place. That too this kind of title seems quite prevalent in the company given they found 35% of all managers are like this. Either the statistic is just plain false or google is really dysfunctional
Good riddance. 7 years too late. I worked as a Google Cloud consultant for a major part of my career. We had this really large client, we were bleeding money because one of our ex-employees promised the client a cost reduction strategy that was impossible and I took over his role - literally a hot seat. We walked into Google's office, it was fantastic, they had something like 12 cuisines in the dining area, multiple game rooms with Table tennis and all. Sleeping rooms, you name it. It was more of a 5 star luxury resort experience than anything.
We walk into the meeting room, past the employee desks, over 50% of it was empty. I asked one of the colleagues showing us around where they all were - "they just work when they like!". Wow, what a dream! Google has/had? this rule where you can take some 20% of your time in a given day for yourself. But, mostly people used something like 40% of it in reality. I thought this was the perfect working culture with great work-life balance. I had very high expectations for the employee quality until I met the TLM and his team we were supposed to meet.
What a bunch of clowns. They suggested we use Cloud Spanner - one of the most expensive offerings at the time and our bottleneck (and bleeding) came from legacy MySql. He didn't even know that Spanner was more expensive than their Cloud SQL offerings, not to mention completely different offerings (it's No-SQL). None of them even knew when to use Spanner for and when to stick with SQL for. It was their own product line-up and we had more knowledge than them! I didn't pass any GCP exams even, at the time. That was the funniest part. We just needed help with some re-architecting of the client's application on a legacy PHP-MySQL combo. They didn't even have suggestions on what stack to migrate to! No idea whatsoever and they tried to talk their way out of it. In the middle of the meeting, my boss leaned over to me and told me "Neya, I think you probably know more than these guys, let's leave" and we wasted no time. To Google's credit, they did offer us some credits to use, but it was far too less than what were bleeding.
I went back to office and out of curiosity searched for how much these guys make - It was easily 120k - 250k (back then) depending on their experience. That's when I literally stopped trusting Google with anything at all. I could never be that laxed when I take that much money from anyone. Later, I learned a lot of these people were just there because depending on the country, you had to maintain a quota (ratio) of certain demographics within Google, most of the TLMs and their teams just conveniently fit into that. Why fix something that's not broke, right?
The Google we knew with the original startup culture was long dead and this story is atleast 7 years old now. For over 5 years since then, they did nothing but make money by simply increasing prices for GCP offerings. I remember our costs rising up suddenly after they just pulled the rug on us with Cloud storage price increase at random. No innovation, nothing. The AI mode they launched now after ChatGPT took the market by storm was what should've been 5 years ago. They are following the pattern exactly as described in the book we learned in our Masters' - "How the mighty fall" (really good book) and well deserved. You reap what you sow.
I have a similar story, many years ago I worked in Brazil as an "IBM consulting partner". Meaning my company implemented IBM solutions on clients, but as a separate company from IBM. Usually as a more of a budget option to IBM internal consulting work.
(I didn't actually work with IBM stuff, I did development of our internal product that we sold in conjunction to IBM solutions to our clients)
I went to their office in São Paulo once and it was the most surreal office experience I ever had. The lobby was all promo stuff and self-aggrandizing marketing. Then after some walking came auditoriums full of people on laptops, not rooms, auditoriums. 200 people-big rooms (although they seem only 70% full) with maybe 30cm between each seat on all sides. I thought at the time that I would rather die than have to work in one of those rooms every day.
The building is actually quite cool looking (from the outside):
Then they took us upstairs where it looked like a concrete maze full of meeting rooms without windows. The discussions were not very fruitful.
My manager told me once that IBM consulting was considerably worse than most of their certified partners (even though they were implementing IBM own products) because they rely on the brand name. They pay slightly more than the partner does for their employees but they treat them like crap.
From that experience I understood how this world of consulting/client-interaction works. Unlike software, consulting scales linearly. So the companies want to squeeze as much as they can from these labor-linear income flows. So it is usually not the best people doing that kind of work.
These companies much rather you buy their solutions and do it yourself, the consulting is just a growth strategy for the product. That is why the external partners were better with the tools than the people inside the company.
60 comments
[ 3.5 ms ] story [ 74.0 ms ] threadIf you oversee 0-2 people, in most cases that’s probably not an efficient ratio. How did Google get so many folks in that position in the first place? And I assume the other 65% take up the slack to fluff their teams? Or what? Leave the other 65% managing 0-2 people?
It is awful, send help.
I don't think mature enterprise companies should have managers of 0-2 people though.
For bigger teams (10+) you either need individuals who are very independent and driven, or have dependable line managers.
It’s part of an effort to have dedicated managers and dedicated engineers instead of hybrid roles.
This is being sold as an efficiency win for the sake of the stock price but it’s really just moved a few people around with the TLMs now 100% focused on programming.
https://www.youtube.com/watch?v=bTuLbiIVDpk
Ah trust me - its actually insane how every firm manages to pull this off - the secret sauce that made the company what it was evaporates and the management optimise for process.
It’s nearly 20 years since Google had a category defining product - they haven’t built or acquired a single thing that dominates in the same way that android, maps, search, docs, etc. has since about 2006. It figures.
I met a long time Google employee this week that mentioned. Sorry I was using voice recognition.
But often such teams in faang come up as a by product of someone’s empire building and that is unfortunate for others involved in it.
That's not to say it's easy - its absolutely not. But Apple is living off of Steve Job's visionary prowess and continues to do so.
As someone who used to be an engineering manager, I was always surprised at how inefficient the division of responsibilities seemed to me. I mean, when I was an eng. manager, a substantial portion of my time was just taken up by logistics - like when we did a move to a new office building, a huge time sink was stuff like seating charts. Perhaps a company as big as Google has more folks taking care of stuff like this, but I still think the following breakdown makes sense:
1. A "technical mentorship" role: someone who codes, but is also explicitly responsible for skills growth and technical feedback of ~5 junior engineers. This person would not be responsible for stuff like salary/raise negotiations, promotion decisions (but would obviously feed into that, more on that below), logistics questions, etc.
2. A "directing manager" (obviously that name kind of sucks, but I didn't want to confuse this with other "director" or "manager" terms). This person explicitly does not need any technical skills. They are responsible for all logistics/salary negotiations, etc. They would be responsible for around ~5 technical mentors, so then up to 25 people under that. Promotion decisions, for example, would be made amongst the 5 technical mentors, deciding who on the team is most deserving to move up. But then the actual salary decision would be made by this "directing manager".
I'm sure this could be tweaked, but the overall idea is to separate technical vs. non-technical skill sets more efficiently.
I wonder why these people are made managers in the first place. That too this kind of title seems quite prevalent in the company given they found 35% of all managers are like this. Either the statistic is just plain false or google is really dysfunctional
Not that I'm trying to gain favor or hoping my future AI managers will remember my kind words and promote me when they take over my company.
We walk into the meeting room, past the employee desks, over 50% of it was empty. I asked one of the colleagues showing us around where they all were - "they just work when they like!". Wow, what a dream! Google has/had? this rule where you can take some 20% of your time in a given day for yourself. But, mostly people used something like 40% of it in reality. I thought this was the perfect working culture with great work-life balance. I had very high expectations for the employee quality until I met the TLM and his team we were supposed to meet.
What a bunch of clowns. They suggested we use Cloud Spanner - one of the most expensive offerings at the time and our bottleneck (and bleeding) came from legacy MySql. He didn't even know that Spanner was more expensive than their Cloud SQL offerings, not to mention completely different offerings (it's No-SQL). None of them even knew when to use Spanner for and when to stick with SQL for. It was their own product line-up and we had more knowledge than them! I didn't pass any GCP exams even, at the time. That was the funniest part. We just needed help with some re-architecting of the client's application on a legacy PHP-MySQL combo. They didn't even have suggestions on what stack to migrate to! No idea whatsoever and they tried to talk their way out of it. In the middle of the meeting, my boss leaned over to me and told me "Neya, I think you probably know more than these guys, let's leave" and we wasted no time. To Google's credit, they did offer us some credits to use, but it was far too less than what were bleeding.
I went back to office and out of curiosity searched for how much these guys make - It was easily 120k - 250k (back then) depending on their experience. That's when I literally stopped trusting Google with anything at all. I could never be that laxed when I take that much money from anyone. Later, I learned a lot of these people were just there because depending on the country, you had to maintain a quota (ratio) of certain demographics within Google, most of the TLMs and their teams just conveniently fit into that. Why fix something that's not broke, right?
The Google we knew with the original startup culture was long dead and this story is atleast 7 years old now. For over 5 years since then, they did nothing but make money by simply increasing prices for GCP offerings. I remember our costs rising up suddenly after they just pulled the rug on us with Cloud storage price increase at random. No innovation, nothing. The AI mode they launched now after ChatGPT took the market by storm was what should've been 5 years ago. They are following the pattern exactly as described in the book we learned in our Masters' - "How the mighty fall" (really good book) and well deserved. You reap what you sow.
/endrant (thanks for reading!)
(I didn't actually work with IBM stuff, I did development of our internal product that we sold in conjunction to IBM solutions to our clients)
I went to their office in São Paulo once and it was the most surreal office experience I ever had. The lobby was all promo stuff and self-aggrandizing marketing. Then after some walking came auditoriums full of people on laptops, not rooms, auditoriums. 200 people-big rooms (although they seem only 70% full) with maybe 30cm between each seat on all sides. I thought at the time that I would rather die than have to work in one of those rooms every day. The building is actually quite cool looking (from the outside):
https://arquivo.arq.br/projetos/edificio-sede-da-ibm
Then they took us upstairs where it looked like a concrete maze full of meeting rooms without windows. The discussions were not very fruitful.
My manager told me once that IBM consulting was considerably worse than most of their certified partners (even though they were implementing IBM own products) because they rely on the brand name. They pay slightly more than the partner does for their employees but they treat them like crap.
From that experience I understood how this world of consulting/client-interaction works. Unlike software, consulting scales linearly. So the companies want to squeeze as much as they can from these labor-linear income flows. So it is usually not the best people doing that kind of work.
These companies much rather you buy their solutions and do it yourself, the consulting is just a growth strategy for the product. That is why the external partners were better with the tools than the people inside the company.
No manager with more then 7 reports No IC with more then 7 managers between them and CEO