Hacker News new | past | comments | ask | show | jobs | submit login

Any company that has been around for as long as Microsoft, has a huge amount of dead weight in its mid-level management. I mean people whose main skill is mastery of the internal political process.

Lest people think I'm singling out Microsoft specifically, this same problem is present at Google as well, but is less of a problem because Google has not been around long enough to pick up the same amount of dead weight personnel.

The first generation of people at a company were fighting for market, and they rose due to their ability to deliver value to colleagues and customers. But when a company becomes successful, the resources available to teams within the company becomes somewhat decoupled from company revenue. Soon, political skill becomes as important or more important than impacting the company's bottom line.

Microsoft and Intel alumni in the late 90s used to tell me that they knew managers whose main goal was to grow the headcount under them, in order to grow their own prestige.

Today at Google you can see some of this same behavior in its nascency. Most of the deadweight is outside Engineering teams because its hard to bullshit when you have to deliver a product. Still, there are peripheral functions like business development, marketing, intellectual property, privacy, policy, etc., where you can always find a few deadbeats that talk slick, but don't seem to have much to show for their time except hiring more people and making a few high impact appearances at meetings. You can go a long way if you can talk the talk, look the part, and kiss the right asses.

Microsoft has been picking up useless deadbeats for nigh on 25 years now. They really need to shed these people.

One company you have to admire is Facebook. I don't know if its true, but I'm told you basically get fired at Facebook in the first year if you don't know how to deliver something of tangible value within that time. That is hardcore.

If that metric was applied at Microsoft and Intel, you'd see the companies shedding 25% of their workforce immediately.

I suspect even Google would shed 10-15% of its people.

Its good in theory, but in practice it doesn't work. Those guys who mastered the political process. They don't get fired. Not getting fired is what they do. Thats the thing they're really good at.

Exactly. And it's not just tech. Most large companies have mid-level workers whose sole job is to not to get laid off (fired). I pity anyone under them cause they are usually the hardest working and first to go. Companies should be looking top-down instead of the usual bottom-up.

And man, it gets cut-throat. You know, when I traded, I used to think that was the most cut-throat work environment. Then the firm I was trading for got acquired by a bigger (much bigger) "traditional" bank. I quickly realized the trading floor was a picnic compared to mid-level management (at a bank they are called VP's) politics. At least a trader, you know the score. Make a profit you win. Don't you leave. Working at this bank; I had to learn to watch my back.

It was at that point I decided to leave banking, and NYC.


> And it's not just tech.

It also applies to countries. Look at the increasing resources that go to people who produce little or nothing of value to others eg lawyers, lobbyists, politicians, financial engineers, LBO operators etc. Not to mention many of the people on welfare.

> Not to mention many of the people on welfare.

Spoken like someone that has never used welfare, doesn't know anyone on welfare, and gets all their ideas from talk radio and Fox News.

> Not to mention many of the people on welfare.

Woah, slow down tiger :)

More seriously, please elaborate. Or maybe I am just a bleeding heart.

You're saying that lawyers and politicians provide little or no value to others? Don't democracy and a fair legal system require good lawyers and politicians?

Sure, and companies might need some middle mangers. That doesn't mean none of the middle managers, lawyers, and politicians are dead weight. This latest session of Congress has been historically unproductive, passing fewer than half as many bills as the previous record set in 2013.

With apologies (and a tip of the hat) to Mr Gates, I posit that measuring the progress of Congress by the quantities of bills they pass is roughly analogous to measuring the progress of airplane design by its weight.

> You're saying that lawyers and politicians provide little or no value to others?

Looking at the state of many governments around the world he's probably being kind in his characterization.

I should have elaborated.

You are right, hiring and firing is a political process, and so you are right, its vulnerable to manipulation by the shrewdest political players.

My assumption, however, is that the layoffs are managed by senior management, who are the shrewdest of all the political players. If they do layoffs right, they put the power of firing into the hands of people in a separate power structure. I mean people like external consultants, or a special committee.

If that isn't the case; and if mid-level management is providing the data to direct layoffs, then yes, its a colossal disaster.

One example of right layoff I have seen is at the rail-road company in my town. They hired consultants to help with the process. My friend explained me how it went. Each person was given a questionnaire asking whom they are reporting to and who are reporting to them. The organization was so bad that many managers did not know who are all the people reporting to them. These people and others many people who were working from home most of the time were the first set of people to go.

This is why the problem inevitably gets worse over time. If you think it's bad at Microsoft, visit IBM. It's mostly insiders, and everyone has a way of looking like they hit their metrics, while the company falls apart.

As someone currently with the company, this description of some mid-level management really resonates with me.

I am incredibly skeptical, however, that any round of layoffs attempting to get rid of these people will have a good hit rate. Plenty of really talented engineers and engineering managers will end up losing their jobs too.

I a lot of my older peers from university went to work for IBM for a couple years right after school, where layoffs occurred pretty regularly. The culture they've described to me, that was present as a result of constant layoffs, isn't something I think many current Microsoft employees will want to be a part of.

I can't believe I'm about to do this, but this is a situation where Michael O'Church hit the nail on the head.


The reason engineers and other tech staff will lose their jobs is because 1) politics above them, 2) lack of adequate technical understanding amongst mgmt to recognize how to usefully allocate technical staff, and 3) lack of political capital amongst the technical staff themselves.

That's a wonderfully over the top and appropriate article on the subject. And it resonates with me because the solution I've generally taken as an engineer is to annoy everyone around me with facts and benchmarks until I find someone in the management caste that has the same goals as I do and then jump ship to their team.

Considering the author, it's amusing that this blew up in my face at Google. Facts were irrelevant there, only status, and this was pointedly explained to me in an effort to get me to conform. So as a powerless new senior engineer blind allocated into something random and unsuitable, I found someone at another company with similar goals and jumped ship shortly after starting when it became clear that was my only other option.

In retrospect, had I endured the tedium of my initial assignment, everything I was saying ended up relevant once they hired people with status making use of my skillset to improve search.


When whom to layoff is a political decision, layoffs will just compound the problem.

That always happens. The problem is the CEO needs an internal partner to get a list of people who need to be let go. And that's generally delegated to the immediate exec staff, they delegate to some layers below and when the list is built, it bubbles up back to the CEO.

In the process, most people think performance is the key to decide who stays and who leaves. While the fact is first ones to go are political enemies, followed by people who pose threat to the middle management(Kind of people who managers think will eventually take their job, if they stay) these kind of people are worst sufferers, because they would have performed well only to be perceived as threats. Next come people who don't fit well in the organization's pay parity. The last one's if they ever go are weak people who can be sacrificed to save manager's yes men and manger's inner circle.

What generally remains are people who are yes men, political alliances and generally people and partners needed to keep political cartels running.

Which is why if a company is laying off people, its generally an indication they will perform worse in the time to come. Because a round of good people have been laid off, many good ones are leaving. And worst stay on, and are more entrenched in the new set up.

That's because when the performance evaluations are top-down and forced to a curve, "performance" itself is a resource to be allocated, and resource allocation decisions are always political.

It's the "evaporative cooling" of groups.

"The problem is the CEO needs an internal partner to get a list of people who need to be let go."

Prepare three envelopes...

I wonder how they'll apply this plan in France, where the spirit of the law is, you may massively fire people only if it's part of "plan social", and the only argument is ofthen it's-this-or-we-go-bankrupt. Especially:

In those kind of schemes, you must ask who wants to leave, give them a package, avoid those with children, the older ones go last, etc. So I doubt they'd let go of the worst people first.

I get what you are saying, but you do make it sound like the (older) people with children are the worst ones, by virtue of having children or being older. Also there is a reason why these measures are in place, being that people with children have more risk (more expenses and responsibility) and are less mobile. And older people typically have a harder time in the job market.

That's correct, I didn't mean to talk about who should be fired first. I meant to say that the company doesn't really get to choose who should leave or stay. There are rules so we don't harm the economic tissue when firing people.

I'm yet to be convinced of how this economy can thrive in the first place. I'm persuaded most French people work in really Dilbertesque comapnies with unpassionnate workmates because of those rules, which correlates with the fact that French people are the first consumers of anti-depressants in the world [1]. I'm not saying it's linked for everyone.

[1] http://www.france24.com/en/20110802-france-world-most-depres...

good point. but there is a very small engineering team in france (the remainder of the musiwave acquisition) and the rest is sales and marketing that spans EMEA (europe and middle east). if there are cuts, likely not broadly hit.

They can always open a new company and hire only those they want from the existing one, can't they?

Those games are pretty obvious, and they've often been used. We judge by the spirit of the law more often than in US. I know a company of 5-8000 people who's firing 9 managers per month [1] so they don't fall into the "group layout" category. They have to persuade them with a resignation package. They plan a "plan social" in a few months. Here's documentation about it: [3]

It's a job. See this agency, at random: [2]. My godfather was a HR manager specialized in "plans sociaux" (=closing factories when famous tech brands moved to Asia), and his job was pretty much explaining CEO of other continents that they have to put money on the table to be allowed to dump a whole factory, which they couldn't really get, obviously. The mother company has to help refurbish the local economy so workers find new jobs, or hire a job agency to find jobs and prove workers aren't just ditched. When the job is done, he writes the last two resignations: his, and the local director. The length of the process is around 2 years for 500 people. Needless to say, mafia jobs are probably safer.

[1] See "New procedures for implementing mass lay-offs" in http://www.proskauer.com/publications/client-alert/new-labor...

[2] http://www.triplet.com/50-10_employment/50-12_laysoff.asp

[3] http://uk.practicallaw.com/0-503-0054#a362756

I always knew France had insane labor laws, but they idea of having to ask permission to close a factory is being insane to me.

Also while I doubt I will ever have to, I wouldn't start a business in France under those rules.

The military is especially prone to this problem. I’d hesitate to refer to people as dead weight, but the retirement structure encourages mid-level management to hang-on for twenty. Even if people are burnt out after 10 years, which the military is very capable of doing, many stay even when they and the Service would be better off if they moved on.

The problem with the military is how benefits are set up. There is a big jump in benefits at twenty years. If you put in 10-12 years your benefits aren't significantly better than someone who just did minimal time. If there was a small jump around the ten year point they'd probably see people willing to stop there instead of holding out 8-10 more years.

Well the personnel problems with the military go far beyond the timing of benefits alone, unfortunately.

As it stands the military actually has great difficulty retaining people past 10 years. Many military communities are chronically undermanned at the senior enlisted and officer leadership levels.

So while it's fair to say that the military has held onto lower-skilled workers, it's not exactly the case that it's leading to a glut of excess deadweight just hanging on. Rather, we use the "idiots" to fill the desks that would simply be left empty were it not for these guys biding their time to 20 years.

Whether that's a net positive or not depends on the given billet being filled, but we'd have to do away with billet-based manpower and fine-tuned "year group" tracking entirely in order to truly see benefit from dropping deadweight in the 10-18 YOS ranks. But the way it stands now, many billets require a certain rank or grade to fill them, no matter what the skill level might be of excess junior (or senior) personnel.

I know several people who left the U.S. Navy after 6-8 years rather than taking unpalatable billets (one was a master-at-arms (IIRC), serving on several of the hospital ships who was given the choice of leaving or moving to Diego Garcia; several others were Sea-Bees) or who took the latter of the up-or-out option after being unable to advance under new requirements.

It always seemed like the Navy was doing its best to get rid of skilled workers.

It's also a feeling of being stuck. I mean, you a 13B, where you gonna get a job? At Burger King? And you're (were) an E6; a platoon leader with responsibility for personnel and artillery equipment worth millions and now you're flipping burgers?

I've always thought the Army needed a job training problem (maybe through the VA, if nothing else) where you could "un-learn" your military skills in favor of civilian skills.

I live in an area that had a relatively high number of returning vets about a year ago. We instituted a job training program at our college. Many institutions, especially community colleges, provide similar programs.

We have been surprised to see that some of the skills making these young people successful while serving have almost no translation into the civilian world, outside of being able to follow orders really, really well.

It's easy to forget that many members of the services do jobs that don't exist in the civilian world. It's easy to forget that many returning vets enlisted at 18 and have had no practical training to ease their transition back to civilian life.

Could you elaborate on those kinds of jobs? Obviously relative few companies hire people to plan artillery strikes, etc, but I suppose those aren't what you mean. Does a purchasing department at Big Corp really differ that much from logistics and supply work?

It's actually less to do with military as in fighting, and more in the direct application of the skills they're supposed to have. For example:

Most requests for our returning vets come from transportation and trade (we're a river town relatively close to major north/south and east/west highways and a major railway line). The largest portion of our returning vets were in the army or army national guard, in the local transportation groups. We don't have infantry in our area, we have truck drivers and mechanics. Anyway, the requests and the expertise seemed a custom fit - the soldiers ran big trucks and trains, and there is no better organization than the United States Army when it comes to large-scale logistics.

Our returning vets, though, are unable to branch outside of their assigned mode of operation.

The mechanics are able to work on, say, heavy machines and diesel engines so they're able to work on the trains and barges really well, but are unable to translate those skills to tractor trailers or small(ish) engines like pick-ups or delivery vans.

The Army transport drivers are able to drive tractor trailers or operate barges. BUT, they do not have the certifications necessary to legally do that work stateside, AND they lack skills in the classroom to be able to sit and pass the test for the licenses. Usually it's because they have never been able to do math well, and are unable to pass a college level mathematics course (it's applied math, so it has practical applications in their jobs - not college algebra or statistics or anything like that). Along with that, they're unable to translate their existing skills into heavy machinery for construction, demolition or excavation.

The leaders/officers are unable to transition out of the military 'because I said so' mentality and into the corporate 'because it's in your best interest' mentality. Our retraining in this area has been the biggest failure; they don't see the need to change their mode of people management, and we don't have a reason for them to either (other than long-term employment).

The paperwork individuals, like you use in your example, transition easily. If the person worked in an office, they will come back and find employment almost immediately.

This could all be helped if the military provided training ON THEIR END before they come back to the world. Believe it or not, the transition from the military to a college environment is incredibly difficult for most folks.

I have worked in two industries that hire a lot of ex-military with essentially a high school level of education.

The first is nuclear power. A lot of the people working in the control rooms are from the military, and it is a very high paying and comfy job for someone without a college degree. The other is high skill technician and assemblers in manufacturing.

The things these jobs have in common is they need a strict adherence to rules and procedures, which is something military training makes people well suited to.

I'll note with the nuclear power example that they don't simply hire ex-military, they hire from a very specific subset of Navy nuclear propulsion workers.

Even being good at following orders isn't going to help you get your civilian SRO license, but sailors who have been trained on the Navy's various reactor types have easily-transferable knowledge for civilian PWR and BWR plants, and so they get picked up quite easily in the civilian sector.

"I've always thought the Army needed a job training problem (maybe through the VA, if nothing else) where you could "un-learn" your military skills in favor of civilian skills."

British Army has various qualifications of its own some of which can be 'mapped' over to civilian equivalents. Mainly technical areas plus management/supervisory. I can't find a clear single source of information but I know from ex-students that the process exists.

Nothing similar in US? Is this not one for the military to consider in terms of recruitment?


The U.S. military always advertises the skills it provides, and they are certainly present in many cases. (I know a fair number of Sea-Bees; carpenters, mechanics, electricians all have marketable skills.) On the other hand, there are limited civilian opportunities in door-kicking-in or explosive ordinance disposal.

(Weirdly, I just got a job notice from my employer about an EOD position. Not it!)

The Navy has something called "COOL" which tries to map civilian certifications to equivalent Navy ratings and NECs, https://www.cool.navy.mil/

But I'm not sure how well it helps in practice, nor on whether other U.S. armed services have something similar.

"If there was a small jump around the ten year point they'd probably see people willing to stop there instead of holding out 8-10 more years."

Or if there was less of a jump at 20.

The military has the up or out program. It varies depending on the branch, but if you don't get promoted past e5 by ten years you get separated. Generally if you get to e6 you will do 15+ and get retirement benefits.

The problem is that if the culture is already bad, it's very easy to shed the false positives, i.e. the good people, and keep the false negatives.

i think you mean the other way around.

when looking for crap, setting off the crap sensor would be a positive result... you don't want to accidentally fire the good ones.

In the last layoffs, many of the people who were let go were back working for Microsoft again in a year or so (often as full-time employees, more often as contractors).

I believe that Microsoft has more contractors than it does employees. That is just crazy.

Also, it's nearly impossible to fire even the most grossly incompetent people. One person I worked with was a utter flatline and doing negative work (very bad Q/A, argumentative, and the most useless person I have ever had as a coworker. It took six months to fire the guy, and it was no secret to the third and fourth layer management that the guy was a huge mistake.

MS needs to get rid of the right people.

> I believe that Microsoft has more contractors than it does employees. That is just crazy.

No, that's smart business if you can do it. You basically remove the burden of tax-minimization and continuity of employment for these people from the corporation. Overall cost might be slightly higher in cashflow terms, but it removes a lot of risk and liabilities.

From a worker-rights perspective it's terrible behaviour (back to the XIX century, choo-choo!), but it can be a very good strategy for a company.

Contractors can be cheaper than employees, despite paying a higher salary. The reason is that they don't need to do so much legal work or pay their insurances and etc, etc. When a project is finished, they can just end the contract and avoid all of the legality. I've known many people that leave their job as employee and get contracted to the same company for those reasons (and more).

Or it's about appearances.

You know the government does this all the time. Congress mandates a department cuts their budget. So ok.. fine, the director lays off the IT department. Then all the "deferred" IT folk go to work at"some consulting company" who get the contract to provide IT services to the department. At twice the price the original payroll was.

it happens all the time in government. And not just in IT.

And contractors don't get stock options. Microsoft has gotten in trouble before about their use of "permanent contractors".

The stock isn't really a big deal anymore now that the company is stable, more or less. Back in the '90s, however, some preferred stock could mean retiring at age 35. And when the orange-badges started seeing their officemates buying mansions and supercars, despite not having worked all that much harder, they felt the need to get the lawyers involved.

Is this the case? I mean, sure options won't be as lucrative, but RSUs are basically an extra way for companies to compensate employees, and still count as "velvet handcuffs".

If the stock rises, that's magnified, but even while relatively static, RSUs count up.

> Contractors can be cheaper than employees, despite paying a higher salary. The reason is that they don't need to do so much legal work or pay their insurances and etc, etc.

Unless they're independent contractors, I've never understood how this can be true. Sure, the company paying the contracting company doesn't have to directly pay for that kind of thing, but the contracting company does, and on top of that, has to earn a profit, and that should all be included in the billing rate.

Having worked for several subcontractors, I can tell you there's a vast difference between the cost of Microsoft doing HR and a much smaller company doing HR.

Microsoft probably keeps good records and actually has good HR staff.

A lot of subcontractors don't keep very good records, outsource their HR staff to India and aren't against closing up shop and reincorporating with a new name.

Edit: Out of 5 of these companies that I actually needed employment verification from after leaving, only 1 of them could actually locate a record saying I worked from them. I always have to go to pay stubs.

There are flexibility issues as well.

I have contracted with IBM many times over my career, and when the project goes belly-up, the company reorganizes, the division you were working in disappears, and employment contracts aren't renewed. No muss, no fuss, no layoffs.

At Google, if you don't make it to level 5 sweng within a certain amount of time(2 years from being hired at level 4), you will most likely get fired. After that you can relax more. So the fluff in software engineering(including managers) is minimal. Facebook though... Ask their employees how that build system is working out for them. There's a lot of doing, agreed, but sometimes you need to just take a chill pill, stop the goddamn hacking for chrissakes, design THEN code.

> At Google, if you don't make it to level 5 sweng within a certain amount of time(2 years from being hired at level 4), you will most likely get fired.

Don't know where you get that idea from. I have a lot of anecdotal evidence to the contrary.

I saw it first hand. My manager was pushing everyone to get promoted to level 5, of the few who didn't, one was put on a pip then fired, the others worked like madmen after seeing that, still didn't get promoted and somehow moved into another group. They still haven't gotten promoted but they work insanely hard/ aren't the same people. I guess it depends on the group, but I was told that timelines for promotion to level 5 are an unwritten rule.

there's no 'unwritten rule' for promotion timelines

It was written for a while - the promotion guidelines used to state that the expectation was that all engineers were expected to get to level 5 within "a reasonable amount of time" (generally 2-3 years between promotions), and that managers should identify roadblocks to continued growth for people stuck at level 3/4. I don't think that the "or out" part was ever specified, but that's the implication. Level 5 was continued a terminal level for folks who didn't want to move into management, roughly equivalent to tenure; at that point, you've paid off your value to the company, and can be trusted to pick your own projects.

I heard less and less about this policy over my time in the company - it was stated explicitly my first few perf cycles in 2009/2010, but I don't think I could recall it ever being mentioned after around 2012. And it was always very sporadically enforced; the longest-serving level 4 I knew had been with the company for 11 years.

It's actually a written rule. Look up the job descriptions for SWE2's and 3's and compare them to those for higher levels. It specifically says that you are expected to advance.

Can you elaborate on the build system?

I think GP meant the system at Facebook where the only goal is to build build build, referencing an earlier comment that said that if you can't prove your value at facebook after a year, they fire you.

> Microsoft has been picking up useless deadbeats for nigh on 25 years now. They really need to shed these people.

Isn't the key reason for these layoffs the recent acquisition of Nokia operations ? If that is, the layoff is just about eliminating redundancy post acquisition, and not addressing the deadweight problem.

Or, the redundancy argument could be used to justify laying off dead weight.

Certainly lay offs due to post acquisition redundancy would seem less likely to generate lawsuits compared to lay offs due to dead weight regardless of the true reason.

AFAIK they pulled this during the '08 crisis also. I heard from a few people on the inside that, while the public reason for the layoff was to trim headcount in response to economic downturn, it was more of a house cleaning.

I was at a company once that had layoffs where the company said that they were going to lay off managers. An old timer succinctly summed up how it would go as follows........

You have a tree full of monkeys. So you shake the hell out of the tree to get the monkeys out of hit. How many monkeys do you have in the tree after your done? The same number, they are just in different branches.

Facebook does indeed have a very efficient corporate structure (whether that will last is another question).

Perhaps 7,000 to 7,300 employees (assuming some pickup the last quarter or two), and $8.9 billion in sales.

Microsoft by comparison had roughly 15,000 employees when they had that level of revenue (adjusting for inflation), circa 1993 / 1994.

But Facebook is in a much more tenuous position. Nobody really needs Facebook; it's mostly entertainment and could be toppled in a few years time by the next internet/tech fad. Microsoft is a lot more entrenched in a lot more places that actually depend on their software.

Facebook is entrenched because that is where all your friends are. G+ is much better (circles are really nice), but what does that matter when your friends aren't there?

Let's be honest, Facebook has a much more efficient product, and by that I mean the product does most of the work for them.

Microsoft had a product that was bundled with ever computer sold. If that's not a product selling itself I don't know what is.

Is it really that black and white to compare the size of engineering teams 20 years apart?

20 years ago you could have bought a house for $50k in the town I live in currently. Today same house is $500k. So, 10 times more.

Let's extrapolate that to Facebook vs Microsoft and basically Facebook would have to have $89 billion in sales or alternatively have 70,000 employees.

This is of course rubbish but illustrates the point that you can't compare revenues from 20 years in absolute terms like there is no such thing as inflation and time value of money.

Facebook is in a much more efficient business. Presumably, many Microsofties are doing on-site hand-holding, stuff that Facebook just doesn't do.

I still think about FB spending 3 billion on that app?

Microsoft has made billion dollar blunders too. Look at aQuantive, it torched almost 7 billion dollars.

This is why I believe in a Semco type culture where every team member receives a cut of the profits and there is a peer review process of who gets to join / stay in the team.

It's the most realistic way a profitably franchise can avoid taking in the political rent seekers. What you spend more on profit share you probably make back on cost savings, and at the very least it buys you a more happy, productive workforce.

Would you recommend a reference on Semco's approach?

The books by Ricardo Semler? "Maveric" and the "The Seven Day Weekend". Both worth reading, imho. Even if you won't be able to apply the same principles they can serve as a good eye-opener for the many things we take for granted.

Remarkable story, thanks for the pointer.


Microsoft is a publically traded company, they have a fiduciary responsibility to return their profits to shareholders and not employees, for better or worse.

This is a myth often used to justify unpopular decisions. The board has a responsibility to guide the company in accordance with the articles, but the shareholders are free to define the articles. If they think redistributing profits to employees is a good idea, they are free to do so.

  they have a fiduciary responsibility to return their 
  profits to shareholders and not employees
As evidenced by the fact that executive pay is extremely low, and constantly falling :)

Banks share profits all the time. Microsoft does not have to maximize the share of profits returned, but the absolute value.

That's what Enron did and General Electric under Jack Welch. Up or out strategy. If you put the incentive under delivering something of value too much, you end up with people lying. Sometimes it's not just about star individuals, but star teams.

The DevDiv org at Microsoft definitely had an up or out strategy, I believe there either was not an upper-limit or it was level 65. My understanding is that Google has an up or out as well, which tops out at level 5 (lower than Microsofts).

The issue with the up-or-out at Microsoft was that it was incredibly hard to hire better devs than the ones being forced out (counter-argument- you are supposed to be hiring on future potential). This was seriously compounded by the amount of legacy code- new hires either wanted nothing to do with it, or wanted to replace it en-masse without understanding the requirements.

The up or out became yet another puzzle piece during the review cycle- if you did not give someone a promo then you would be pressured to cut them loose.

> Up or out strategy

I think that 'up or out' is slightly different than 'deliver something of tangible value.' It's possible to keep delivering value without wanting to be promoted into (e.g.) management.

I think the companies with the "up or out" policy also offer career paths that doesn't lead to management. It's viewed as acceptable to keep getting promoted within an individual-contributor role. They've tried to remove any reason to not be promoted.

To provide a counterpoint:

Enron collapsed under massive fraud and GE would have been bankrupt if they had to mark-to-market their debt during the financial crisis.

What does it mean that they would have been bankrupt? I thought (wrongly perhaps) that you were only bankrupt if you couldn't pay your bills. So no matter their true debt back then, doesn't GE's current existence prove they weren't bankrupt?

If your assets are worth less than your liabilities, you're technically insolvent. If you can still pay your bills from cashflows, you don't need to claim bankruptcy, but on a long enough timeline without a significant change, you will go bankrupt. There was a long stretch of time post-GFC where this was the case with GE.

Banks have to 'mark-to-market' their debt periodically, but since GE wasn't a bank, they didn't abide by the same regulations. Say there were two identical houses on a street with $500k mortgages, but the market crashed, the houses are worth less, and only 1/2 of the mortgages will be repaid. If Bank of America owned the first house, the mortgage asset (it's an asset to the bank) would be marked-to-market and what was formerly a $500k asset would now be worth $250k, and the bank would show a loss of $250k.

GE doesn't have to do this and could carry assets at their previous value (and didn't have to mark down bad debt either). To simplify the story, during the GFC, GE had that one mortgage "worth" $500k and debts worth $400k. If they were a bank, they would have had to admit that the mortgage was really worth $250k, and they would've gone into receivership.

If not for that accounting quirk and the massive amount of bailout cash they took from the Federal Reserve / FDIC, they would have actually had to declare bankruptcy. Ironically enough, the FDIC program and Federal Reserve are to shore up banks and GE was explicitly not a bank per the above few paragraphs. So make sense of how they got access to bailout money solely authorized to bail out banks..

[1] - $45B in losses buried on GE's balance sheet: http://www.businessinsider.com/henry-blodget-living-on-plane...

[2] - GE borrowed $16B from Federal Reserve: http://www.propublica.org/article/general-electric-tapped-fe...

[3] - FDIC to back $139B in GE Capital Debt: http://dealbook.nytimes.com/2008/11/12/fdic-to-back-139-bill...

[4] - GE Borrows $59.3B from FDIC program: http://www.bloomberg.com/news/2010-12-01/ge-borrowed-16-bill...

> If your assets are worth less than your liabilities, you're technically insolvent.

No you aren't. It just means you have a negative net worth.

insolvent - unable to pay debts owed.

Think of people instead of companies. I suspect a majority of people in America have a negative net worth. That doesn't make them insolvent. It just means they have lots of college loans and they are working to pay them back.


A situation where the value of a company's liabilities exceeds its assets. Accounting insolvency looks only at the firm's balance sheet, deeming a company "insolvent on the books" when its net worth appears negative.

you added an adjective to the term :-) Yes you can call them insolvent if you want to, but it doesn't mean anything if they can still pay their debts. If the number keeps going down it is certainly a problem, but if it heading back up it is fine.

The housing crisis is a perfect example of this. It only became a problem for people who bought more house than they could afford with hopes of selling at a moments notice. My house was as much as 50k underwater (going based on foreclosure sales of identical townhouses) during the crisis. Fortunately, I didn't buy more house than I could afford. I could easily continue to pay my mortgage each month and was never anywhere near bankrupt. I sold the house recently for closer to 10k loss. Was I insolvent when the house was down 50k? No.

[1]Cash flow insolvency involves a lack of liquidity to pay debts as they fall due. Balance sheet insolvency involves having negative net assets—where liabilities exceed assets. Insolvency is not a synonym for bankruptcy, which is a determination of insolvency made by a court of law with resulting legal orders intended to resolve the insolvency.

[1] http://en.wikipedia.org/wiki/Insolvency

The quote of mine that you first replied to was:

> If your assets are worth less than your liabilities, you're technically insolvent.

Which you disagreed with. It's a literal, factual statement though.

> Yes you can call them insolvent if you want to, but it doesn't mean anything if they can still pay their debts.

It actually does mean something in corporate finance.

> Was I insolvent when the house was down 50k? No.

Stop comparing corporate finance to personal finance, they aren't remotely similar. The only reason that the phrase "Technical Insolvency" exists is because there are consequences if companies breach that threshold.

From the Title 11 of the US Bankruptcy Code[1]:

    The term “insolvent” means—
    (A) with reference to an entity other than a partnership and a
    municipality, financial condition such that the sum of such
    entity’s debts is greater than all of such entity’s property,
    at a fair valuation
This isn't just parsing terms for fun, contract law relies on US Federal code, which has definite consequences for insolvency. All commercial loans come with a plethora of 'loan covenants' that mandate certain actions based on a company's health. Common covenants include coverage ratios, debt/equity ratios, times-interest earned ratios, etc. Technical insolvency would've breached many, many covenants.

If a loan covenant is breached, the debt-holder can demand additional collateral and in some cases, they can demand full repayment of their outstanding debt. At the time, GE had ~$90B in cash/investments and over $300B in current debt. A fire-sale on those assets to pay off that debt load would have killed GE.

[1] - http://www.law.cornell.edu/uscode/text/11/101

Barrkel's got it right. At least in accounting terms, companies aren't people and your college loans don't have covenants that allow the debtholder to liquidate assets.

The world of corporate accounting is very literal. 'Technical insolvency' is a synonym for balance sheet insolvency, while the inability to pay a debt is 'cash-flow insolvency'. You can have either type, both, or neither depending on circumstances.

GE at the time would've still been cash-flow solvent, however they had tens of billions of dollars in debt coming due that likely would've forced them into cash-flow insolvency without taxpayer money.

Also to note: our megabanks would also likely be forced into restructuring (it's probably a bad idea for a bank to go bankrupt) if they also had to "mark to market" with the same strictness as before.

Instead of restructuring, the Fed and FASB decided to blow more bubbles in 2009:


I think what mickeyouse means is that if the taxpayers didn't step in to subsidize all the financial institutions, then GE would have bee bankrupt (just like most other financial institutions).

It seems to be working out well for McKinsey.

and who selects the ones to fire? The ass-kissers of course.

Exactly. Every layoff i have seen personally or through close friends at other company's the ass kissers survive and those t hat threaten them get laid off. Those that threaten them (the competent ones) are always painted as non team players because generally the competent ones are unhappy when the teams gain so many slick talking ass kissers but few doers.

This is good stuff. Thanks for the links. I have seen many times and even done the withdraw and become cynical that they talk about. In my case I hated the entire company so it was a bigger issue.

I will say though that when companies lay off competent people one thing I see time and time again the competent people that survived (who probably are better at the politics) always abandon ship. Its very demoralizing watching the kick ass coder you work with go and Joe Slick that never checked in one line of code stay.

So its really not in a companies interest to get rid of those that can do their jobs well even if they are not wise on the politics.

One remedy for this in the software world would be to publish detailed credits for key employee names/roles associated with software releases, something that is done in the film industry where people care very much about where their name appears in the opening and closing credits.

Great teams are quite rare and short-lived. One specific version of a software product can be fantastic due to the short-lived team that designed and built that one release. You see this clearly with hardware, e.g. the 2012 version of device model XYZ earns a higher price in the resale market. It would be great if the teams building these positive exceptions could be recognized, as they scatter and reform across companies.

I'm sorry but the first one was 20 minutes of "the world is not fair. Some people think its fair but it's not" repeated for 20 minutes.

The second one was "You need to observe people's actions and understand why they do things" repeated for 20 minutes.

Those are both very helpful points but I think people would be better served with some advice or techniques.

For example: Discuss your plans with each person one by one before a big meeting so that everyone will be more likely to see it your way in the big meeting.

One challenge is that naturally talented people often don't understand why the facts are insufficient. Others who are less talented often needed to develop political skills for survival. Agree that those links are basic (hence Politics 101), but convincing technical folk to care about this topic isn't easy.

One needs to understand offense in order to play defense. Here are some books with timeless techniques for those with ethical intentions:

1) Herb Cohen, You can Negotiate Anything (general)

2) Chris Malburg, How to Fire Your Boss (labor)

3) Robert P. Smith, Riches Among the Ruins (sales)

Re: big meetings. Herb Cohen points out that most meetings are the end of the negotiation, not the beginning :)

So the solution is to join the ranks of the lying ass-kissers? Who would be left to do any work?

That's not suggested in my post, the references or most interpretations of the word "defensive".

You don't need to read that much. Most people on HN aren't going to be next Petur Balish, but if you want to protect your job what you should do is find the guy who is the best political player and make sure his job is secure only so long he has your technical expertise. Now it is in his interest to keep you around and your job is almost as safe as his is.

If you want to read more, get a copy of How to Make Friends and Influence People, or 40 Laws of Power (depending on how evil you are okay with being).

In the comments sections someone also recommended this great talk: http://www.ted.com/talks/joshua_prager_in_search_for_the_man...

Companies should lay off products, not people. How can a company expect to support the same products with fewer people? The hoped for efficiencies from laying off the "dead weight" are probably small. The company should kill its loser products and focus on its winners. "More wood behind fewer arrows."

Michael O. Church wrote more about this in a recent blog post:


The last layoffs at Microsoft were largely cutting products, it was an utter failure IMHO- there were very few teams who were allowed to hire and so many rockstar devs taking a chance on products which got cut were left in the cold. I saw way too many awesome devs get pushed out of the company because their product was canceled and the project they wanted to work on had no headcount.

> Companies should lay off products, not people.

Layoffs are generally associated with reduced mission scope, even if its not a reduction in named product lines (it can, e.g., involve reduction in supported platforms for the same area of named products.)

Still, reducing products or otherwise reducing mission scope on its own doesn't save money, you still have to cut people. Especially if the scope reduction doesn't neatly follow organizational lines, there may be no direct mapping between scope reduction and which positions get eliminated, and even if there is, if you have good and meaningful performance metrics and similar skill requirements across the parts of the mission that are preserved and those that are retained, not simply mapping from the functions reduced to the people let go makes sound business sense.

There needs to be an org change first- flattening the hierarchy and a focus on getting shit done. Put in a process where teams can jettison inefficient members.

In most areas at Microsoft above level 65 it is all politics.

That's all well and good, but I bet there's tons of dead weight in the engineering ranks as well. Not just areas/jobs indirectly linked to core business functions as you cited.

I have begun wondering lately if the reasons for the inevitable death of living things has any anologue to the dynamics you describe here for corporations.

Yes, organization development literature is sorely lacking in research and discussion of immortality.

I'm not sure if your joking, but there is research into managing a business' longevity, such as the book "Built to Last":


Not joking :) Built to Last is inspirational, a bit like Warren Bennis "Organizing Genius". I was thinking of research topics like these:

1) Financing, staffing and governance of short-lived organizations which perform complex tasks then disband, e.g. film production, election campaigns, bi-annual sporting events, disaster relief, NGO coalitions. How are institutional identity and knowledge preserved when developed within temporary contexts?

2) What organizational scenarios have been proposed by scifi/fantasy writers for worlds where humans or other lifeforms are immortal? Do those scenarios suggest areas for research or innovation in modern corporations? Olaf Stapledon, A. E. Van Vogt, C.S. Lewis, R.A. Lafferty, Walter Jon Williams, etc.

3) Historically, business has been run by families with inter-generational inheritance. Some public and many private companies still operate this way. There is a body of literature on family-based governance and succession planning. Is that literature relevant to public companies like Facebook and Google where the founders have inheritable shares with unique voting rights on corporate futures? Will other corps seek superpowered share classes? Will consumers seek out B corps, public corps, private corps or hybrids?

Garther Morgan's "Images of Organization" provides food for thought: http://www.ribbonfarm.com/2010/07/13/the-eight-metaphors-of-...

I believe it's "Gareth Morgan" (Here's a link: http://www.amazon.com/Images-Organization-Gareth-Morgan/dp/1...). Interesting read, but dry.

The BCG matrix. Microsoft is in the Cash Cow stage and moving to the Dog stage.

Rules of layoff - Keep the indispensable people in terms of work. - Keep the people who are on good terms with you (the top management). - Keep people who sleep with you (the top management). - Keep people with whom you play poker. - Fire everyone who will create no fuss for firing.

In relation to Google, though, I've read that the current workforce is somewhere between 20-35% composed of contractors. That seems like a reasonable way to A) avoid certain types of costs, and B) reduce or delay the onset of the type of corporate malaise you describe. What do you think?

I have seen way too much dead weight in engineering functions too. Basically, people who keep their jobs because they write horrendous code that nobody else can manage.

So you can BS your way in many functions.

"The deepest cuts are expected to come from the businesses the company bought from Nokia several months ago"

This describes working at NASA down to a T.

Hey! Don't tell people that! We build rockets! And all of our software has to be manned-flight-rated! (Including people.nasa.gov!)

Writing a one-off throw away analysis script? Better consult NPR 7150.2A and matrix it correctly!

It's not about safety. It's about checking of a box and satisfying a process.

LOL The irony of someone getting fired at Facebook for not delivering something of tangible value is so juicy.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact