Hacker News new | comments | show | ask | jobs | submit login
"The carpets are so clean, we don't need janitors" (machinesplusminds.blogspot.com)
311 points by zdw on Aug 4, 2012 | hide | past | web | favorite | 113 comments

In this case, the downside to saving money by laying off expensive experience is clear. Unfortunately, the downsides are rarely clear before the upsides, and by the time the mistake is obvious, returning to the previous state is difficult, and an incentive has been created for short term savings at the expense of long term ones.

By way of example, several years ago I sold tools at Sears. Sears had spent decades building consumer confidence, particularly in their Craftsman brand. If you bought a Craftsman tool, and it broke for any reason, they replaced it, sans receipt. As a result, this warranty was transferable - there were no questions asked. While this policy certainly did not extend to every product sold at Sears, it did exemplify a commitment to service and quality: in the words of one older person I talked to once "if you bought it at Sears, you didn't have to worry."

As time went on, Sears was able to increase profit margins by slowly restricting the tool warranty, and using crappier parts. The obvious problem was that once Sears lost its reputation as a "you don't have to worry about it" store, it had to compete with stores like Wal-Mart on price[0].

The interesting problem here is not the Sears strategy, but the fact that it takes years for the effect of reduced quality to become obvious: in the short term, consumer confidence in the brand is still high, so the worse products are bought for the same prices, under the assumption that the quality is still high. By the time consumers on the average figure this out (when your drill wears down in 3 years instead of 10), someone that made the change has been able to demonstrate clear savings to the company and move on.

Several people have suggested solutions. The first one I hear thrown around is "get rid of executives." I think this is shortsighted in the same way that getting rid of a good IT department is. Certainly some executives are useless, but "get rid of the bad executives" is vacuous. A more compelling solution is to create incentive structures that encourage "bad executives" to be good ones, such as incentives which outweigh the short-term gains gotten by reducing, say, IT spending. For instance, I've heard it suggested that companies use long-term equity, say 15 years out. I'm not sure how that squares with moving from company to company, but it's interesting.

[0]This is not to say that the Sears quality-first business model was sustainable, merely that there was certainly a tradeoff, the effects of which take a long time to see in total.

In this case, the downside to saving money by laying off expensive experience is clear.

Things like time developers spend doing traditional IT tasks are clear to developers. But it's amazing how little of that is visible to executives. They certainly don't get an hour-by-hour breakdown of where developers spend their time. And the time to develop software is notoriously difficult to estimate even using the best techniques out there -- and most people just eyeball it. So executives really can't quantify the loss of productivity, they just hear programmers bitching.

The problem there is that the CIO or equivalent isn't making these things clear to the rest of the executives.

"but the fact that it takes years for the effect of reduced quality to become obvious: in the short term, consumer confidence in the brand is still high," The only thing I'd point out is that the two are related. Sears had this policy for a very long time and so it slowly becomes an established fact. That's why you can make the cost savings change and show new profits for a quarter, 2 quarters, maybe even a year or two. Most likely the "smart person" who made that decision gets promoted onto another group that needs better margins. Then the wheels come off but the exec is long gone onto something else.

I'm asuming there would be an entire board of directors or some such who were aware of this decision. They can't all be so naive that they dont realize the danger of undermining their brand reputation.

They must have made an informed decision.

sure, but what were their individual motives? If the CFO had a plan to migrate to SomeOtherCo(R) in two years if he did well enough at Sears, then maybe that was his horizon...

The board should have been informed, but I see no compelling reason to assume they were.

>In this case, the downside to saving money by laying off expensive experience is clear. Unfortunately, the downsides are rarely clear before the upsides

The general heuristic of pointing out information asymmetries even when they seem obvious is a good one. In fact I'd say my experience has been that most people could benefit by explicitly saying more obvious things. The reason people don't is for fear of looking bad.

Overcommunicate, overcommunicate, overcommunicate. Also, overcommunicate. It's remarkably helpful to overcommunicate-by overcommunicating you'll clear up a ton of miscommunications due to different assumptions you and the other person are making, even when things seem obvious. So again, overcommunicate.

I feel like reputation is an asset that pays dividends. You can sell it off in part or whole for a short term cash gain, but then your dividend payments will shrink accordingly.

> I'm not sure how that squares with moving from company to company...

The stipulation to accommodate moving from company to company could be an artificial constraint. The desired outcome goals of the company outweighs the desired outcome expressed by any individual executive. Incentive structures can reflect that.

There are many ways to accomplish this, but the reality is shareholders who wield sufficient power to enforce this goal-seeking simply are not interested in doing this. That's the real nut to crack. The incentive structure of owners and not just employees subject to the agency problem is part of the spectrum to consider.

Incentive structures all up and down the line of authority are a real tough design problem.

This reminds of a news story I distinctly remember about Circuit City from a few years back. They fired their senior sales associates and hired cheaper presumably less knowledgeable workers to cut costs. They were already on the way out at the time but I have no doubt this sped up the process as it signaled to their customers that experience and knowledge aren't a priority so much as the bottom line.


Ah, I was buying my tools at Home Depot and Lowe's (how the heck it's supposed to be pronounced?) and was thinking that may be I should now shop at Sears since I can afford it. But looks like may be it's not a good idea anymore.

>For instance, I've heard it suggested that companies use long-term equity, say 15 years out.

Not going to work, they will just trade whatever they got on some secondary market.

"(how the heck it's supposed to be pronounced?)"

Ignore the e and pronounce it "lows" as in "highs and lows".

Source: Lowe's radio commercials I've heard.

Got screwed this way be Sears. Won't shop there again.

As fun as getting out our tribal aversion to managers who don't respect geeks is, let's focus on the actionable career point: do not work for cost centers.

Or better. Turn the perception of being a cost center into being a producer.

Most managers wouldn't dare tuch you so long as you say for every dollar put into the department, we save the business x dollars. You could even argue the margins too.

A similar tack is to get yourself perceived as an incremental value adder, as opposed someone filling a role.

For example, IT is generally seen as filling a set of necessary roles where each person is more or less interchangeable. Sales is generally seen as incremental value adds, and sales people get paid accordingly.

So how can you get seen as an incremental value adder if you're working in IT? Simple: start projects and clearly explain the impact to the bottom line. If you show the executives that you've saved the company $2MM/year forever, and that project wouldn't have happened if you hadn't dreamed it up, that's going to get recognized. Even at large, highly bureaucratic companies.

Also: good management is like oxygen -- you only really notice it, much less recognize that you need it, when it goes away.

Bad management is like mud. It get's on everything and is almost impossible to get rid of.

A real life example:

In the last ten years Connecticut Light & Power made major cut backs on trimming the trees near its power lines. The state got slammed by a huge snowstorm last Holloween and a huge portion of the state lost power - for as long as 10 days. Not only did CL&P suffer a financial and public relations disaster in the aftermath of the storm, they decided to double their tree-trimming budget for the next year. But this just led to further problems - people who weren't used to the power company coming in with chainsaws created a fuss at town meetings about CL&P removing or damaging their trees, leading to more ongoing PR problems.

I've read in multiple places and wholeheartedly agree that it's important to focus on visibility in your work.

If your managers can barely describe what you do, or how you do it, then your job is just asking to be cut.

When I show my clients an update, I always focus on what "looks" or "feels" different with the user experience. I wouldn't (maybe couldn't even) convey the awesomeness of all my fandangled SQL subqueries.

I am brought to mind a quote by a god like galaxy: "When you do things right, people won't be sure you've done anything at all."

For a long time, that was my email sig.

Then I realized the problem with it.

Now I still try to do things right, and then try to make sure that a few key people knew what happened.

Step 1 is to start bragging.

Step 2 is to have other people brag about you.

In a somewhat related adventure, I am scheduled to give a talk soon (hopefully the same talk twice) about managing a security assessment project. I am hearing about this same sort of beheading happening in the infosec departments from various clients. The impact on me, personally, is that my client contacts are stretched so thin that they don't have cycles to prepare and make the best use of my time. I'm planning to have handouts with checklists. So this topic has been very much on my mind, lately.

One of the things I've learned from Tom is to keep track of how I spend my time. It is tedious and frustrating, but not unlike dieting: without data, you just have no proof. If I had found myself in the situation that the OP describes, my first reaction would have been to keep a running log of every task I had done, including start/stop time and a description of the work done. I've written scripts to do this in the past that generate PDF timesheets -- you could probably use Trac somehow. I'd keep this stuff on a device I owned, and bring the logs to my reviews. Yes, this adds to the stress. Yes, you'll get criticized for it. But when someone who has no clue why the carpets are always clean starts swinging that axe, you'll be wearing armor.

I have put this in my very small file of "things to share when I leave."

Share it when you're not sure you want to leave. The reaction will tell you everything you need to know.

You'd probably get better results if you share it before you leave. The value and influence of an exit interview is questionable, at best.

Just my 2c. :)

A few years ago, my manager at the time explained to me how our company had acted as an incubator for 2-3 of our competitors. It worked like this: Our company would design a product over the course of a few years and once it was in the pipeline (being bought) it would have a life of a few years. More than once, the then current CEO of the company would decide to cut the engineering staff to save money and move on before the next product to fill the pipeline would be needed i.e. the lack of new products wouldn't show up on the bottom line for a few years. In the meantime, the talented engineering team (that they layed off) would go out and start another company to compete with us....

Everytime I see a "for folks from HN" message at the top of a post I feel like I'm part of a school trip. Can we stop littering the web with these?

Efficiency rarely correlates with headcount and budget. 9 years ago, our IT budget was $11 million. Today, it's a hair over $60 million. Adjusting for inflation, we should be around 13 million.

Most of the added budget has gone to bloated off the shelf software applications, retaining duplicate/redundant staffing from a pseudo-merger, and mismanaged consultants.

The OP has added a lot of facts in replies in this post, indicating that his staff was relatively lean (1 staffer for every 50 users).

But be careful in drawing too many conclusions from one instance of anectdata.

I just started a new job two weeks ago, meaning that I've put in approx 80 hours in. I would guess that I was unable to perform my required functions for about 30 of those, due to setup inefficiencies. This has also delayed the onboard learning curve, because I have to stop and figure out who to talk to to get permissions, or wait for a machine, or an image, or whatnot BEFORE I could start figuring out how to do whatever it was I was trying to do.

It sounds to me (and I know this won't be a popular opinion here) that the original IT department was overstaffed. A printer never out of paper? No projector ever found with a dead bulb?

Much like there being a "healthy" amount of unemployment (~3%, I recall?), there's a healthy amount of smaller items not met. It sounds like the company was overpaying for IT.

What would you say about a building maintenance team that left 3% of the burnt out/bad light bulbs around? Most of the stuff he's talking about can be monitored and caught before they become a real problem or dealt with as soon as they are discovered. (Projector bulbs have expected lifetime, network printers can tell you when they're out of paper, etc.) To me, it just sounds like they have solid monitoring and procedures in place for preventative maintenance.

So, if developers ship on time are they overstaffed? Projects should never ship on time.

Given the anecdotal basis of the account, there may be a good bit of selection bias on both sides of the cut here. He may be forgiving or forgetting those times when something did slip (either because it was so rare or because it was so easy to get fixed), and might be extremely harsh on the failings in the future, due to it being the 18th time it's happened that week or just "I told you so!"

Techstop. That's all I'm going to say.

Executives are day traders. I've known day traders to argue that they make more money trading in the first and last 30 minutes of the day than they do if they're trading all day, because they end up getting involved in "boredom trading" that is break-even at best (and blows out their variance) and quite possibly losing. As for executives, they're disciplined, they can enjoy their easy jobs and high compensation and accept the boredom associated with rarely being needed, because if things are running well, they aren't operationally necessary. Unfortunately, they're rarely content to work 2 hours per day at their cushy jobs. They become forces of nature due to a combination of (a) a need to feel active and important, and (b) a theater in which it's hard for them to have real positive impact (because performance assessment of executives is impossible and most just don't have the talent).

"Trading boredom" sets in. The executive's job is to Make Decisions and to gamble with the company culture and operations. In order to feel useful, executives pull this shit. Like "saving money" by gutting the IT organization.

What's worst about this is that the executives who do this sort of thing rarely suffer any consequences. They shift blame to subordinates whom they can fire, and if the heat is turned up on them, they move on to other jobs. The employees and the owners get fucked, but executives get to flit about from one cushy job to another on account of building a web of connections that makes them effectively invincible.

I hear nightmarish stories of micro-managers all the time. They feel that in order to do their jobs they have to tweak EVERYTHING that crosses their desk.

One memorable anecdote I came across was about an animation artist who was in charge of animating the queen for a 3d chess game. Because he knew his manager's weakness he threw in a duck... Yes, a duck. Fully animated with sound effects and everything.

Sure enough, when he showed his work to his manager he said "It's perfect, but get rid of the duck."

So maybe the moral of this lesson is that when things are running 100% smoothly, it's wise to toss your micro-manager a bone by doing something intentionally a little wrong to give them something to meddle with. ;-)

The duck story is battle chess: http://en.wikipedia.org/wiki?title=Talk:Battle_Chess#The_duc...

I've often had the opposite problem: companies where the CEO is so disconnected from the company (usually by working remotely 80% of the time) that the company doesn't have direction and languishes.

Speaking of managers ignoring things, I'm reminded of the last episode of the 90s sci-fi tv show 'Sliders', apparently the producers thought the network exec werent reading the scripts, so they deliberately broke a rule, and wrote a scene where someone was shot in the head and it wasn't picked up on.

cf. http://en.m.wikipedia.org/wiki/Sliders#section_2

Oh thanks! I was wondering where exactly I had heard that.

Aww, so there never was a duck.

I've heard the same advice applied to PhD theses: Examiners always want to change something, so leave something in which you won't mind having them tell you to change.

I have witnessed this advice applied to contract drafts. Lawyers from the other company were routinely given some bullshit clauses to prove their usefulness in vetting the contract.

The really entertaining part is when they don't catch them. Every clause is negotiable - don't leave it to your lawyers alone to negotiate.

A friend of mine who buys and restores old houses deals with a similar issue. Buyers hire inspectors when they want to buy a house, who are supposed to alert them to major liabilities that might be lurking. But those inspectors aren't well trained and also have a need to appear to be "useful" to the people who pay them, so if you're selling a house that doesn't have any obvious flaws, such as a broken furnace or crumbling foundation, then they will entertain themselves by pointing out trivial items that dont' really affect the house's suitability, but can often scare buyers away or cause headaches for the seller. (For example, cracks in the foundation of an 80 year old house that likely appeared within months of the house being built and haven't budged since.)

To counter this, my friend purposely leaves 3-4 fixes that he intended to complete all along un-done for the inspector to notice and fixate on. Then the inspector gets to look useful, the buyer gets a good house, and everyone's happy.

I had a micromanager who hit everyone with 3 to 4 status pings per day. Even on work-from-home days, which defeats the point. His status pinging was a compulsion like email checking, but it eventually caused the team to fail.

Consequently, I hate status meetings and standups and the various so-called Scr(ot)um tactics used to make people feel watched. I realize they're probably necessary as an occasional annoyance, but I think they're an unnecessary pain in the ass most of the time. Their real purpose is to intimidate people into being productive, and that Theory-X bullshit doesn't work in the modern economy.

I'm not really sure how you're implementing scrum. Our scrum standups are always just the programmers, once a day in the mornings, and are used to keep all the programmers on the same page and talk to each other to find out if features might have a codependency, or if other blockers can be solved by something else another dev is working on. There's no "watching" of people or "intimidation".

You might want to speak to your manager next time something like that is happening, because there's no way it can be good for your productivity or the other people on your team. It's certainly not a normal practice. At the very worst, if someone isn't pulling their weight they should be dealt with in private.

A daily 30 minute standup with 10 programmers (three minutes each) is 5 hours of development time lost. That's 25 hours a week, over half a full-time developer. It's probably more. Let's say stand-up is at 0900. Nobody getting in early is going to really get in the zone because they know standup is coming. Then after standup there's another 15 - 30 minutes getting into the flow. Then it's nearly lunchtime. In other words, standup pretty much kill half the day.

Can you really say they are worth it?

Our standups are at 11:45, so we have impetus to keep it brief and get to lunch. In any case, even if it does cut into a stretch of programming, it serves as an invaluable reminder to go get lunch.

Yeah, I agree that I'd be annoyed with standup at any time before 11:30, but 11:45 is perfect and basically doesn't interfere with anything. In a certain sense it doesn't cost any dev time at all if you take it for granted that your developers should stop coding midway through the day and eat lunch. (We absolutely do not eat lunch during standup - we have standup, then we physically leave the office to go eat.) Sure, some days we end up in endless meetings which mean we're stuck in the building eating takeout, but that has nothing to do with standup.

We experimented with 10 AM scrum but I found that if I came to the office at 9ish, I used to get distracted with the thought of Scrum. Eventually, we decided to move it to the afternoons at 4:30 which works out well because you are almost done with your work day and you can get distracted all you want.

The goal is 45 seconds a developer not 5 minutes. Round up to 10 minutes a day * 5 days a week = one 50 minute developer meeting a week which is fairly common.

PS: You do lose a little time gathering, and people generally spend some time organizing there thoughts etc. But, keeping track of a short synopsis is useful, as is knowing what other people are working on, and most importantly when someone get's stuck working on the same things for a few days. It also adds a lot of pressure to get at least one thing done every day, which many people slack on.

A 30 minute standup counters the whole idea of a 'standup' meeting.

10 people seems like a very large team for a single project, do you have multiple teams in the same standup? If that's the case then you're certainly wasting time as cross project communication is only rarely going to be useful.

You can do standup meetings in lots of different ways. Our standups are maybe 5min max for a team of 6 and they're very worth it.

The problem is trying to communicate this to management. I was told that standups are non-optional.

Yeah, but next you'll tell us at your standups you actually stood up. No one gave me an answer when I asked why they called our meeting standup meetings.

In my current project we do actually stand up, although only due to the pleasure of the team leader. The main reason it's called a "stand up" is to serve as a reminder that it should be very short, 5 minutes max, or otherwise the time most people would feel comfortable standing for a meeting.

Yeah, thanks for the answer. I was being facetious, because our micromanager would divert the meetings to be daily 25 minute meetings! Daily.

There is a large set of humans who get more work done when they're held accountable for it. I'm one of them. Call it "intimidating people into being productive" if you must, but it works. Even in the "modern economy."

The manager in the example you gave was indeed taking things a bit far.

We have a daily meeting that's not supposed to exceed 15 minutes. We're still trying to get into the swing of making that happen. But I find them invaluable for learning what other people are up to that day. And they frequently result in people being able to solve their problems after the meeting because someone else was able to help.

This sort of daily communication can be really helpful. But it's also really easy for it to devolve into a long meeting that's much less useful. And if management is using it to apply pressure to get more work done, they're making a huge mistake.

> they move on to other jobs.

... with their CVs exposing the fuckups as accomplishments: "Successfully managed a cost cutting campaign, sparing the company 2 bazillions in IT costs". The previous company won't admit it was actually a disaster.

Its even funner at companies that cycle managers through business units to get them experience with the whole company. Leads to a "must make changes that are short-term budget wins but long term negatives" attitude.

It's called "a bungee boss": http://dilbert.com/strips/comic/1994-09-07/

This kind of generalization is toxic. The engineering community as a whole should not accept this mindset of Engineering vs Executives.

The executive community is quite on board with the mindset of "engineers are fungible cost-centers to be reduced as much as possible" so why should engineers not be wary themselves?

Engineers being culturally complacent and not willing to make a fuss is precisely why companies like Google, who have revenues per employee twice as high as say Goldman Sachs, get away with paying their engineers a fraction of what financial professionals at Goldman get paid.

It's not that. Good executives are worth far more than the compensation that they get; but as a class, their net impact is neutral at best and probably negative. As with stock traders (where a lot of them just lose money) most of them aren't very competent. That doesn't invalidate every single one of them or what they're trying to do.

It's the CEO-pay debate all over again. No one who actually understands large companies would argue against the notion that it's worth $10 million per year (several times that, actually) for a company to have a good CEO over a bad one. On the other hand, shitty CEOs are paid highly just for belonging to the class of people considered eligible to be CEOs, and increasing executive compensation has paradoxically reduced quality of service: http://michaelochurch.wordpress.com/2010/11/22/pay-more-get-...

Also, I think lightweight management is generally better and I don't see why people find my insinuation that most executives should be working 2 hours per day to be an insult to what they do. If I were at a point in my career where I could legitimately justify the kind of compensation that I make now with a 2-hour workday, that would be awesome. I could go back to school, or volunteer, or take another job. I don't know why people in our culture equate working long hours with being productive. I equate it with being so unproductive that people have to make monstrous sacrifices just to reach par (although the reality is that most people are plenty capable of being productive, and the long hours come from social expectations).

I see 2 flaws in an otherwise fine argument.

1/ I don't think anyone is worth 10 million/year no matter how "good" they are. I sure believe a good CEO should be paid adequately but those are just ridiculous amounts, and the CEO is only one part of the puzzle, even if important. As a CEO its hard to screw up (even thus many do), its hard to be very good, but its easy to be in the middle. Like most jobs in fact.

2/Working 2H a day certainly is an issue. Simply because it doesnt mean you're ultra efficient, obviously. You need more than 2H a day just to be able to talk to people in the company. One shouldn't put extremes like that as they can be misunderstood. It's ok to work 5h-7h a day, or even less some days if you're efficient. But 2H, is just not realistic, you can't even speak that fast.

Imagine a choice among 5 or so options to be made. Choose the "wrong" one and you get nothing. Any of the other 4 and you get at least a few million. But, choose the "right" one and you get $1Bn. There's a guy who already knows which is the "right" one. How much are you willing to pay him for the 10 seconds of work it would take him to tell you which it is?

That's the idea behind high executive compensation. I agree, it very rarely works out quite that way IRL, but lots of people are quite attached to this fantasy.

It's not hard to find an athlete that's 'worth' 10 million / year. It's not hard to find an actor that 'adds' 10 million in value to a film. It's not hard to find people that have destroyed 100 million in value in a single year. Suggesting, that you can't find people that add 10 million in value when then can leverage 10,000+ people under them is suspect. At that scale just not fucking up is worth a lot.

PS: Plenty of stocks have dipped when well known CEO's died or left.

I don't think anyone is worth 10 million/year no matter how "good" they are.

Viscerally, I agree. On the other hand, if we go back to 1995 or any time before it, no 25-year-old programmer was "worth" $100,000 per year. Now that's a typical salary for a strong engineer. I'm cautious about saying people "make too much" money because, ultimately, no one really knows what work is worth. That's why we have markets, and they aren't perfect but they do better for a lot of things (not all, and I'm actually pretty far left by US standards) than any kind of central command and analysis humans have ever devised for that sort of thing. Societies need infrastructure (environmental controls, welfare state) to correct imbalance but markets work. Usually. Executive pay in large companies is set by a parasitic social elite rather than a real market, so that's somewhat of a different story, but I'm not sure where I draw the line.

The difference between how a company operates when there's a good CEO vs. a bad one is easily $10 million, because CEOs have multiplier (or divider, when r < 1) effects and a billion-dollar company is in play. I don't like the message that is sent by having CEOs make 500 times what average workers do, but I wouldn't doubt for a second that the difference between a good CEO vs. a bad one, in a large company, is more than 500 times as impactful as hiring or firing an average worker. My issue with the high pay is that CEOs get it even if they turn out to be total wankbaskets.

It's ok to work 5h-7h a day, or even less some days if you're efficient. But 2H, is just not realistic, you can't even speak that fast.

Why not? Perhaps I'm being starry-eyed and idealistic, but I think there's easily a factor of 10 in variation in a person's productivity in an hour of work, and that properly motivated people can get a lot done in a short time period. Looking back over my career, I'd guess that 90% of the value that I've added occurred in 20% of the time. I think a lot can be accomplished with little time, and why not? If I were offered a $2-million job, the first thing I'd ask is if I could work 20 hours per week for half of that salary. It'd be a win-win bargain, because I'd be getting the time off, and they'd be getting 70-80% of my full-time productivity.

The purpose of the 8-hour day is to enhance the likelihood of capturing the 2 to 4 hours in which the person is actually highly productive. Companies fear (and justifiably so) that if the work expectation were a 3-hour day, people would give their best 3 hours to their personal side projects or interests and let the company have the leftovers. So they mandate a day that stretched from after breakfast to before dinner to maximize the chances of getting peoples' productive hours.

I don't think the solution of keeping people corralled in an office for what is, in winter, a full day is going to hold up much longer. I think the post-managerial vision of Valve is the future, and that once this is established, companies will be more comfortable trusting people to work 4 hours in an average day but get shit done. (Of course, there are times when it's appropriate to work 14 hours per day, but those conditions don't, or at least shouldn't, last for long.)

To make it clear, I'm not advocating laziness here. Working an effective 4-hour day is a lot harder and less lazy than working the sloppy 10-hour day that is typical in the professions.

You're making a simple mistake: valuing the worth of something by its impact. But that's not how worth is determined; you need both supply and demand. If providing something has high impact, but that something is very common, then it has low worth, irrespective of how high impact it is. People would die without oxygen and water, but neither of these things are expensive.

The market for CEO compensation is clearly not very efficient. It's a big risk to put someone into the job to see how they do, so there isn't a lot of experimentation; few people get to build up experience. So the supply is very restricted, and this will inflate the price of labour quite a bit. There's not a lot of hard science between how applicable experience in one company or type of company is transferable to another, so even CVs can't be relied upon. And performance and actual value created may take some time to show up; there's a lot of things that a CEO could do that seem to create short-term value at long-term expense.

The biggest reason CEO compensation is so high is because they are in a pyramid peak position where they can skim off a small percentage of a lot of people's productivity and claim it as their own. In effect, it doesn't "cost" a lot for a large corporation to pay its executives a lot of money, compared to all the other costs.

I wonder if we'll ever have virtual economies / simulators developed enough to be able to learn / train / evaluate prospective company leaders in the future.

> If providing something has high impact, but that something is very common, then it has low worth, irrespective of how high impact it is. People would die without oxygen and water, but neither of these things are expensive.

With employees the heuristic is (very roughly) market price + VORP (value over replacement player). If an average programmer typically adds $300k of value to your company, and market price for an average programmer is $70k, and you find one who is twice as effective-you can pay them a lot more than $140k.

> The biggest reason CEO compensation is so high is because they are in a pyramid peak position where they can skim off a small percentage of a lot of people's productivity and claim it as their own.

This isn't a bad thing. If the CEO can boost the average employee's productivity by 1%, that can be enormously impactful. The difference between a CEO who makes a 1% impact and a CEO who makes a .5% impact translates to huge sums in large corporations.

So assuming that we had the ability to distinguish between good and bad CEOs, and that good CEOs were rare, it would certainly make sense to pay good CEOs huge sums and average CEOs much less. Unfortunately as you said, no one has gotten good at evaluating CEOs yet.

theres an order of magnitudes between 100k and 10M. Plus; 17 years ago 100K was a lot yes. 17 YEARS ago :p Maybe in 50 years 10M will be okay but the point is moot.

As for time, I know in many companies people "work" many hours which consist of non-work chatter, coffee, etc. I'm talking of actual real work time. You can work 2H and sit 8H, then I'd say 2H work is not enough as the actual work discussions take more than that. Of course some barely even work 1H per week. But they don't get anything done, so its not that they're efficient. It's that they don't work.

  Unfortunately, they're rarely content to work 2 hours per 
  day at their cushy jobs. 
These are not the characteristics of executives at successful companies. Have you ever started a successful business? If so, did you work two hours per day? I doubt many YC founders would see themselves in this assessment.

Yes, yes, I see. A person struggling mightily to build a brand-new company from scratch is obviously a perfect analogy for a person who's placed in a high-level position at a strong, established company. I bet that their tasks and workloads are practically identical, even!

YC founders and startups have nothing to do with the story IMO. They're generally engineers that make their own companies.

Executives as depicted are however the general standard for most companies. They generally have never created any company from scratch.

I doubt many YC founders would cut IT because things are working well.

Startup founders are in a different class because there's actually 8+ (sometimes 14+) hours of work to be done in a day in a company that's fighting its way into existence.

  > As for executives, they're disciplined, they can enjoy
  their easy jobs and high compensation and accept the 
  boredom associated with rarely being needed, because if
  things are running well, they aren't operationally
I'm sorry but you have zero clue as to what running a business is like. Zero.

There really isn't much more I can say other than to quote Twain: "He who holds a cat by the tail learns something he can learn in no other way".

Go hold a cat by the tail. Then come back and re-read your post.

In a firm the size depicted in the post business is run by the marketing division. In these firms most of the divisions have executives to sign purchases and attend meetings.

I don't understand the reaction here in the comments. Do the people talking about the uselessness of management think the same of sysadmins whose systems are running smoothly?

There are bad executives like their are bad programmers. You should narrow it down vs being so broad.

No need to generalize based on your work experience. There are good AND bad managers.

The fact that IT gets cut does not mean the manager is bad. You work in IT, so I'm sorry I have to break this to you, but sometimes the universe does not revolve around IT. Sometimes there are more important divisions, and sometimes IT expenses have to be cut.

There's not enough detail in the article to know if the manager was good or bad.

Things were running well. The executive thought that IT was bloated and made cuts. Things were no longer running well. Far more money was lost in the company's dysfunction than saved by the cuts.

Sounds like a bad executive to me.

I never implied that there aren't good managers and executives, but only that most of them aren't, because most of them get caught up in "boredom trading" that doesn't solve problems, but is just managerial action for action's sake.

First, he said the savings almost certainly less than the cost of the waste. This sounds like a guess. He gives no numbers, and doesn't appear to have quantified anything.

Second, maybe IT was running so smoothly previously because it was bloated, and was a huge drain on company resources. Maybe it was running smoothly because a bad manager refused to put a limit on anything, leading to a division whos costs were completely out of control.

We have no idea, because there are no details in the article about the company, or about the IT spending before OR after the cuts.

The exact quote is:

"Of course, it all went to shit. New employees would go a week before they had machines, phones, passwords, and ACLs. Printers ran out of paper, projectors ran out of lightbulbs, servers ran out of storage, networks got misconfigured, and so forth. The total time lost and wasted across the whole company was most certainly greater than the savings of laying off the expensive and skilled IT staff."

The guy might be a bit biased but it's completely clear from his article that even though he might not have exact numbers the company was much better off(not only in terms of total profit) with the old IT department.

Are you a manager?

No, it's clear from the article that he thought the company was worse off.. but really had no details to back it up.

All we know from the article is that IT was worse off. But we have no idea how the company as a whole is doing.

Maybe you aren't aware, but most companies consist of more than the IT dept.

The OP does have details to back it up, just not hard numbers. He states how new employees had problems getting setup, old employees dealing with misc problems, things not running smoothly in general. Yes most companies consist of more than IT but these days the problem is many organizations seem to look at IT as an extra expense they can skimp on or even completely do without("The bosses son is really good with computer's he can fix any problems that arise!"). Theres two sides to every story and maybe this department was bloated and could've used some restructuring we don't know that part for sure but we do know that things went from everyone being able to work efficiently to all sorts of random issues, suffered by general employees not IT that should've been avoided.

You're taking this awfully personally, I'll ask again do you do in some sort of managerial position? Do you have some personal experience with this sort of thing?

The OP does have details to back it up ... employees had problems getting setup

Ok.. tell me about those employees. How much do they get paid? Minimum wage? What do they do? Do they even require a desk?

A week to setup a desk that the employee doesn't even need because they are required to travel constantly is not a problem.

You're taking this awfully personally

I'm not taking any of this personally. You're just sensitive because someone is suggesting that an IT dept might have required cuts.

Geez.. for people who are so scientifically-minded, you sure do seem eager to take everything in the article on faith.

I'm the OP.

The employees were expensive software developers, electrical engineers, RF signal electrical engineers, and microelectronics design engineers. There probably was one person making minwage. She was a very sweet developmentally disabled girl who delivered the mail to our desks.

Similarly, we don't know the names of any of those executives. That means we don't know if we currently work for them. If we work for other executives, we shouldn't assume that we can generalize this story to our experience!

In fact, since the odds of this exact management team starting a new company in exactly the same industry with exactly the same business model are close to zero, I think we can agree that it would be impossible to learn anything from this story even if we DID know all the details.

A few years later, the company itself was effectively dead.

Which leaves the question, was the cut in IT a cause or a symptom? It could have been the last in a long line of bad decisions; I can't see this bringing down a company all by itself.

I'm the OP. After few years of bad IT hell, the rest of the company was almost completely dead in the water, in part because of the bad IT drag, and in part because of other management decisions that had bad outcomes. I then left when it was gutted with layoffs. A few years after that, it was sold for pennies on the dollar prices to a low-quality technology aggregration company.

I think you're trying to argue (in multiple responses) that the evidence in the article doesn't rule out all other hypotheses, so we shouldn't take it as proof.


It's not a formal proof. It's a narrative of a general pattern that we might want to learn from. If we can learn from narratives that are completely false - raganwald's excellent posts come to mind - then surely we can learn from narratives that are not scientifically rigorous.

There is a difference in meaning between most certainly (what was actually written) and almost certainly (your misquotation, and apparently your reading).

He never says what the company is or what it does.

After IT was cut, he says it would take a week before a computer was on their desk.

That sounds bad. But we have no idea what the company does, so we don't actually know if this is a problem or not.

Look, every division is not critical, and sometimes that division is IT.

We have no idea how much the inefficiency costs or how much was saved by the cuts. So we have no idea if this was a bad OR good thing for the company.

> That sounds bad. But we have no idea what the company does, so we don't actually know if this is a problem or not.

That's true. The description of the company does lend itself to the belief that it was sufficiently large enough to make these things a problem. If you have people working in cubicles there is, more than likely, a need for them to have a computer, phone, etc. While it might be possible for them to not those tools, I would expect a different style of working environment if office work were not a significant component of their job.

> We have no idea how much the inefficiency costs or how much was saved by the cuts. So we have no idea if this was a bad OR good thing for the company.

From the sounds of things, they didn't have a problem with inefficiency prior to the cuts. It sounds very much like they had just the right number of just the right people on hand to keep things running smoothly. At that point, it's like you have an engine and decide, I really don't need all this oil, let's get rid of some of it. You might be spend less on the oil but then you have to deal with the increased wear on the rest of your engine or risk having it seize up completely.

Yeah, we don't know the real size of the company or what the company did but, the kinds of things he's describing will lead to a degradation of overall morale and you'll find that "broken window syndrome" starts taking over. You'll find that the IT people hate everyone and that everyone start's hating IT which makes the whole situation worse.

So the question is, do you save a little money now to find you have a massive expense in lingering in the future?

From the sounds of things, they didn't have a problem with inefficiency prior to the cuts.

He includes no details about the IT staffing levels in the company, what the budget was, the size of the company, what the company did, what the employees did, the reason for the cuts (other than speculation), or anything else.

For all we know, it was a company with 5 non-IT employees, and a staff of 50 IT people who flipped a coin once a day to decide which of 1 of them was going to work that day.

Prove me wrong. How many people were in the IT dept? What was their budget? What did the company do?

You have no idea.. because there are no details in the article. So you do not have enough information to say if there was an efficiency problem.

A company of about 400 total on-site staff, about 8ish internal IT staff before, and 3 internal IT staff after.

FOAD, Hope that Helps, Have a Nice Day.

8ish is significantly less than I was expecting. My hat's off to them for keeping things running so smoothly!

8 to 400 just keeping a system going (assumed MS from the vocabulary in the original article) sounds like a reasonable metric to me assuming no big business application development. 3 to 400 sounds very small, especially if there needs to be a Windows roll out (e.g. xp -> 7).

Disclaimer: I'm an observant end user who has worked in organisations ranging from 70 to 1200 staff total, and who has seen huge differences in basic it function. As others have pointed out, the cost of the less efficient IT support is 'hidden' in other budgets and in people 'just getting on with it'. I've seen newly appointed people share logins with established people just to be able to do anything which is an obvious security problem. My current employers have noticeably good IT support, but need to make savings, so I am worried.

Even if they'd kept 3 of the best and brightest they'd be hard pressed to keep up.

thank you.. that tells me far more about the situation

> For all we know, it was a company with 5 non-IT employees, and a staff of 50 IT people who flipped a coin once a day to decide which of 1 of them was going to work that day.

You're right on that, we don't have any idea about the total size of the department or the total size of the company as a whole. So, I'll grant you that there might have been some room for cuts in the department but, that really doesn't matter for his point.

What we see is that prior to the reorganization everything worked and that after the reorganization it didn't. This looks like someone pruned way too much out of an area that they didn't fully understand. The only thing we have to work from the assertion that:

"The total time lost and wasted across the whole company was most certainly greater than the savings of laying off the expensive and skilled IT staff."

I am making what I would consider educated guess on the scale of the organization and the impact of the changes. That impact would indicate that they now have insufficient staffing in IT and inefficiencies elsewhere as a result.

I am making what I would consider educated guess on the scale of the organization and the impact of the changes.

WHAT?! You're making a guess based on what? THERE WERE NO DETAILS. It's impossible for that to be an educated guess.

> You're making a guess based on what?

I didn't write that, but it's clear to me that they're basing it on the results.

They got the "scale of the organization" from the results. No kidding.

The OP posted and they appear to have been correct, while your "for all we know" turned out to be almost exactly backwards.

People match stories against their own experience. Any half-decent sysadmin probably has a story or two like this one. So yes, we do have something with which to assess the credibility of the story.

It would appear that you do not share that experience, though.

There is a strain of thinking evident in the article, and one I think is pervasive in corporate America today, which is that measurable savings are highlighted while difficult to quantify lost benefits are deemphasized.

When you can't precisely measure how much is lost by doing something half-assed, it's probably worthwhile to do it right.

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