[optimiz3]

Steve Ballmer revered Jack Welch's rank-and-yank management style.

During the 2000s and early 2010s Microsoft's stack-rank reviews were infamous for the internal politics created. Didn't matter if you had a full team of top-tier engineers, the bottom 15% WAS going to get managed out. Savvy managers would hire low-tier performers to protect the productive ones.

That's the only lens through which I have to view Mr. Welch's contributions unfortunately.

https://en.wikipedia.org/wiki/Vitality_curve

https://money.cnn.com/2013/11/13/technology/enterprise/micro...

[stefco_]

Contrast with W. Edwards Deming's approach that stresses the systematics of the organization and considers rank-and-yank a symptom of bad management:

https://en.wikipedia.org/wiki/W._Edwards_Deming#Seven_Deadly...

[rhizome]

As far as Welch is concerned, I'm pretty certain that his system was never going to find him in the bottom 10% of any possible ranking. Call it Gödel's Management Strategy.

[rstuart4133]

I identify as software engineer that works with software, processes and a little bit of machines and maths, but have also run a few business in partnership with others.

Years ago in an effort at self improvement I did a postgrad course on software quality. Studying what Deming has to say was a big part of it. Turns out it is not so easy to apply it to engineering (as opposed to manufacturing what the engineers designed), and the ham fisted approaches the postgrad course was promoting at the time (like CMM) didn't look promising, and with the benefit of 20 years hindsight now look farcical.

Nonetheless, his dogma of extracting statistically valid measures of processes, propagating those through the organisation so harvesting everyone's ideas on how to improve them struck a deep cord, and it's something that I've tried to apply ever since. I've been exposed to other MBA related ideas, but that is the only one that ever made sense to me.

It's a pity it's so damned hard to apply to software engineering. It has one one notable effect - I'm now seen by my co-workers as near obsessive about collecting and monitoring failure rates - whether they be bugs or things like people reporting software being difficult to use.

[billman]

+1 for Deming

[amyjess]

IIRC Welch came up with rank-and-yank because he needed to lay people off and decided he wanted to be "fair" about it and not lay off his best performers.

And somehow that got cargo-culted into "we should continuously get rid of the bottom X% of people all the time".

[jonny_eh]

> And somehow that got cargo-culted into…

his book maybe?

He was still defending the practice in 2005: https://usatoday30.usatoday.com/educate/college/careers/Advi...

[hnick]

He seems to suggest not just getting rid of them, rather letting them know where they stand in the hope they move on or take it upon themselves to improve.

> You've got a bad take on it. It's letting the bottom 10% know where they are and then giving them a chance to move on. About 70% (of the bottom 10%) leave on their own. Who wants to be on the bottom once they know it? You don't fire them. That's being mean.

Though of course firing is always an option especially in the USA.

From that article the key point I found was that he would rather expend company resources nurturing top performers instead of salvaging the bottom performers. He believes it's a much better investment.

[asveikau]

Well, at some point in American business culture it does seem like a number of people did lose sight and become convinced that the only way to deliver shareholder value is to lay off continuously.

[neves]

It looks like this became the basis of American economy. And you can't loose your job, because you'll miss your mortgage or education payments and loose your health care. So you must do anything your employer tell you to do.

[WalterBright]

> because you'll miss your mortgage or education payments and loose your health care.

This is why the sensible thing to do is to save up a cushion of 6 months. You can continue with your employer health care with COBRA for 6 months or a year.

[michaelgrafl]

You‘re saying this as if it were realistic for the majority of people to save up a cushion of 6 months in a reasonable time frame (or ever).

[WalterBright]

A majority of people are significantly above the poverty line. If they're living hand-to-mouth, they are mismanaging their finances.

For example, I know a fellow who told me he could not handle the slightest disruption in his paycheck. He was driving a new car, living in new house, wore expensive clothes, and was starting a family.

A friend of mine living in an apartment told me he was unable to pay his bills. I asked him what the payment was on his new car. I suggested he sell the car, and buy a car he could pay cash for. He surprised me by following my advice and was able to reduce his stress about his finances.

My father, an officer in the AF, was once tasked with counseling the privates on their finances, as many were unable to pay their bills on the local economy. (The AF really tries to be a good citizen in the community where the base is located.)

The privates were paid every two weeks. Half of them drank their paycheck away and ate steak for the first week, and begged/borrowed/stiffed others for money the second. They received the same paychecks, but half could not manage their finances. My father would counsel them, work up budgets for them, all for naught.

[kelnos]

It's not easy, but I think a lot of the people who don't save, actually could. Part of the problem is people don't think they can save "enough to matter", so they never start. Even starting with just tossing $5 or $10 a month into savings account can help. It kickstarts the habit, at least, and often people can later justify raising that amount or putting in a little more here and there.

For some other people, if they'd drop the "keeping up with the Joneses" mentality even just a little bit, they could develop significant savings.

Yes, there are many people in the US who are just barely getting by, and can't meaningfully save without making cuts to essential spending, but that certainly doesn't describe everyone who currently lives paycheck-to-paycheck.

[lr4444lr]

If they can't, then they shouldn't be homeowners. It's not a choice for everyone. Home ownership is a liability.

Signed, a homeowner who sees a lot of his less well off neighbors making much more extrvagant purchases than he does.

Braces for downvoting

[AtlasBarfed]

If you're a republican, people can't do this because of their failings and choices.

If you're a democrat, it's because the economic deck is stacked against people.

I thought that a very large majority of credit card debt was healthcare costs and other structural life costs, so that is score one for the democrat view, although we are a very unhealthy population (so back to the republican view a bit)

Then again, the real estate bubble showed that people were getting way too much house (republican view)... of course enabled by almost no oversight of the lending and financials underling it (democrat view).

[hugi]

I feel desperately sorry for Americans when it comes to health care.

[rchaud]

Which Americans though?

There are numerous Americans and American institutions whose work specifically involves pushing the interests of insurers, pharma and hospitals on Capitol Hill, against those of average people who work minimum wage jobs and can't afford basic coverage.

[lotsofpulp]

Why would you need COBRA? You can go to healthcare.gov and buy health insurance that is probably as good or better than what most employers offer. And you wouldn’t have to pay for months you didn’t need healthcare.

[amyjess]

Exchange health insurance is still pretty unaffordable if you're unemployed or underemployed, and it's much less likely than employer health insurance to cover medical conditions that aren't required by law to be covered (e.g. gender dysphoria).

Our system is messed up.

[dnautics]

I would say it's become the basis of worldwide modern central banking capitalism, gas and water socialist countries included.

[random_savv]

Can't tell if this is meant as sarcasm or truthfully :-)

[markus_zhang]

Actually cutting cost usually comes first to the mind of the investment bankers because:

1. Cutting cost will always yield a quick gain

2. It's a lot more difficult to turn it around and win in the long term

3. Even if the banker wants a long term picture, cutting cost is not necessarily a bad idea.

[AtlasBarfed]

Yeah, it should really only be a periodic thing every 10 years, but really that's what "reorgs" mean.

[cletus]

Add to that that stack ranking has a tendency to devolve into a popularity contest and that it most rewards those who do the best job of convincing everyone that their job is hard then one has to ask if this is really the best system for a group of people that trends towards being introverted.

[RandallBrown]

I worked at Microsoft for a short time and got burned pretty badly by the stack ranking system.

In my first year there I had 3 different managers, so by the time my review came around, nobody really knew me very well. At least not well enough to fight for me in the ranking meetings.

So ultimately, I received the lowest ranking. What made it most annoying is that it was prefaced with a speech about how I was doing a good job, but someone had to get that ranking. This was just me.

About the only good advice that manager gave me was that if this happened again, I should leave the company because it was going to be an endless spiral that I couldn't get out of.

It was actually the last time they did the stack ranking and I left before finding out if things got any better. The whole project I was on got canceled a few weeks after I left anyway.

[teej]

Hiring folks to act as telomeres for your team seems so cruel.

[akitzmiller]

Ooh. This is good. When I write my best selling Management Genetics(R) book filled with cellular jargon to describe The Business Double Helix(R), I am totally using this. With attribution, of course.

[jjoonathan]

Ugh, I hate it already. I am sure it would be wildly successful.

[gowld]

Why? Those are people who couldn't get a better job anyway.

[watwut]

They were people who thought they are getting real job, but have been set up to fail from day one. Just because you pay people some money does not mean it is ok to use them like toys.

[Ididntdothis]

Ballmer actually left a better company than Welch.

[throwaway55554]

Because of policies like this or in spite of them?

[hylaride]

Both. I've posted this before about Ballmer, but I'll repost:

He consistently increased earnings via Windows and Office during his tenure as CEO. He was a very good businessman.

He was not, however, a product guy. Witness their failures with the Zune, mobile, search, and early cloud computing. His problem was that he always tried to do what would be best for Microsoft first and not the customer. You can get away with this if you already have an entrenched install base (windows/office), but with net new products people will look at what's best for them first. Even with developers developers developers he tried too hard to get people to use pure MS tech and made developing for the web on MS ugly if you are using anything other than .NET. He didn't see the geeks learning to program in their basement or dorm rooms who ended up using free software instead of going through any "free" channels that may have been available to them (were they?). The result is a generation of developers on Macs and to a lesser extent linux. MS is only now catering to them.

Funny enough, thinking of the customer is what MS did with the xbox and it worked out well enough.

[Ididntdothis]

Ballmer also didn’t create a personality cult around himself. He actually seemed ok with being viewed as a clown as long as it helped Microsoft.

[olivermarks]

Great point.

[vl]

Welch also invented corporate raiding by management. When he started, employees owned 0, when finished 31 percent of GE, with majority of 31 owned by management. This effectively transferred 1/3 of value from shareholders to management. Since then same thing happens with most other corporations.

[technotony]

But at the same time enterprize value increased 40x. Paying 1/3 of the gains to management seems fair given the performance.

[vl]

But it's more complex than this. Would it increase as much without paying 1/3? Were short-term gains optimized at the expense of sustainable business?

[cma]

Didn't they start doing leveraged finacial product investments and stuff since it was heads their massive stock option incentives pay off, tails no big loss for them.

[gowld]

Was that 31percent purchased or seized?

[vl]

Granted as stock options or restricted stock units.

[TuringNYC]

Granted by dilution

[hintymad]

I find Netflix's model works better: allocating resources by visibility of a team. Roughly speaking, a manager gets a chunk of budget from her manager, and she can decide how big a package she gives to each of her team members. The more important a team becomes, the more budget the team gets. The more visible a person becomes to her manager, the more likely this person will get larger package. This comes with the following nuances: * A team member is responsible for her visibility only to her team and her manager. There is really no need for any popularity contest if the team is small enough. I deem it as fair as it can get. * A manager can definitely abuse the system, but then the team would suffer churn and productivity loss, and the manager would be punished. If the team performed anyway with little churn despite the manager's misbehavior, well, did the manager really misbehaved? * To make the system work, the management chain needs to be perceptive, and teams need to be small. Netflix delivered both, which showed that Reed Hastings is a truly great CEO.

[lazyant]

I may not agree completely with Netflix' philosophy but I love that it has "tests", like "would you fight for a person to stay? if not, let her go" https://jobs.netflix.com/culture

[mywittyname]

With the caveat that you need to be like Netflix to pull this off. Most companies don't have the endless talent pool to feed from that makes this strategy work. So many have to settle for, "Is this person better than no one?

[kamaal]

>>she can decide how big a package she gives to each of her team members

In my experience this turns out be a huge scam. Worse a legally sanctioned scam.

Say she had 11 members in her team. She will give $90 to her pet, and $1 each to the remaining 10. That is how this works on the longer run. The definition of a 'high performer' is often hazy here, and a case can be built up to justify giving the money to whom she wants.

Eventually it turns out there are enormous cartel like structures, you have to be in them to get paid up well. Or you slog like a donkey and quit eventually.

[hintymad]

Nope. Netflix would not allow that to happen. There is always a balance and check, freedom and responsibility, right?

[kamaal]

Who is watching? And who will keep things in check?

People who are supposed to approve, audit and verify are a part of the same system playing with the same incentives.

[shekharshan]

They called him “neutron Jack” for that very reason.

[projektfu]

That was more because of his restructuring in the early 90s. "As with a neutron bomb, the buildings are still standing but the people are gone." He didn't seem to understand GE's broad products businesses but did understand finance, so he continuously closed those operations.

[triceratops]

> He didn't seem to understand GE's broad products businesses but did understand finance, so he continuously closed those operations.

Uh...no. He was a trained chemical engineer, and understood products, manufacturing, and operations perfectly well.

In his own words, here's why he emphasized finance:

"My gut told me that compared to the industrial operations I did know, this business [GE Capital] seemed an easy way to make money. You didn't have to invest heavily in R&D, build factories, and bend metal day after day. You didn't have to build scale to be competitive. The business was all about intellectual capital - finding smart and creative people and then using GE's strong balance sheet. This thing looked like a 'gold mine' to me." (emphasis mine)

[keenmaster]

"[GE Capital] seemed an easy way to make money...This thing looked like a 'gold mine' to me."

A good MBA finance curriculum might kick off with the Modigliani-Miller theorem (M&M), which basically says:

1. Theoretically, you can't create value through your mix of financing. It doesn't matter whether you use 100% equity, 100% debt, or 50/50, your mix of financing won't create value (unlike good R&D for example).

2. #1 above is predicated on the absence of tax incentives, bankruptcy costs, agency costs, and asymmetric information, and the market being efficient.

The creators of M&M believed that the assumptions in #2 are mostly false most of the time. The beauty of the theory is in the assumptions themselves. Prescriptively, it basically says that your mix of financing matters only insofar as the assumptions in #2 are false for your project at your company in your country. If you start a division like GE Finance, and you start using complicated financial derivatives, you have several things to prove. You have to show which M&M assumptions are false for your company, to what degree they are false, that your suggested method of financing will take advantage of the M&M violations, and that the magnitude of financing matches the magnitude of M&M assumption violation (this last bit is important, GE really went hog wild with derivatives in a way that was completely not justified). GE Capital never passed the M&M sniff test.

[fsckboy]

huh? you don't have this right.

Modigliani-Miller refers to the financial structure of your corporation, GE in this case: it refers to the mix of capital that GE raises from investors (debt + equity). To give a simple example to illustrate, the value of your lemonade stand as a business is based on your revenue minus your costs, it's based on your "business". You need to raise money to buy your raw materials and equipment to get started or expand? Whether you borrow that money or sell shares in your lemonade stand does not change the value of your business. The bit about absence of tax incentives is because you can write off the interest on debt from the income taxes, but the dividends you pay actually get taxed, so there is an incentive for the equity holders to raise some money via debt rather than equity; but the point that M&M makes remains true.

GE Capital did not fund GE, so M&M doesn't apply.

M&M would apply to GE's ownership of GE Capital, but if all the debt and equity is owned by one entity there really isn't a difference between them (see Humpty Dumpty, back together again).

It's quite common for large industrial companies to have large credit arms: oil companies extend credit to gas stations to buy their gas, just like automobile manufacturers finance car purchases. GE was no different. Leveraging their expertise in finance was a natural extension of the business, and financial firms sometimes fail like any other business, whether they are standalone or wholly owned subsidiaries.

[keenmaster]

I'm aware that GE Capital extended credit to suppliers (good, sometimes). However, my understanding is that it also enabled greater use of financing through off-balance-sheet derivatives for the core entity (bad). GE tried to create value through GE Capital, and it nearly destroyed the company. It's actually a good illustration of the agency costs of debt mentioned by M&M.

[projektfu]

I stand corrected.

[rmason]

I think a better way to remember him is his insistence that all GE's company's be either first or second in their market. So under performers got a chance to go for it and if they failed they were sold.

Welch more than quintupled GE during his time. None of his successors could manage the company so what he built is slowly being dismantled. What I never understood is why they didn't just split the company in half which would have returned more value to the shareholders.

[lotsofpulp]

> None of his successors could manage the company so what he built is slowly being dismantled.

Or GE Financial was hiding a ton of bad debts while he was in charge.

[martin1975]

The problem with performance reviews is that they are mostly subjective instead of mostly objective. I've yet to see a company specify a clear objective criteria for advancement.

[PopeDotNinja]

I have the same problem with behavioral interviews.

"Tell me about a time when..."

"I want to see how you think."

"What your greatest..."

95% of my philosophy is don't be an asshole, work hard, know my stuff, and don't let the turkeys get me down. I spend no time dwelling on mistakes, as I internalize the lessons and throw out the reason I learned those lessons.

[tlb]

The reasons might come in handy later when you're trying to mentor younger colleagues and you're trying to explain why things are the way they are.

[amznthrowaway5]

> Steve Ballmer revered Jack Welch's rank-and-yank management style.

So does Jeff Bezos. 10% forced "least effective" is still implemented here at Amazon and is truly destructive. The quota often gets passed down to small teams. The worst part is that performance evaluations have almost nothing to do with performance.

[AtlasBarfed]

"Savvy managers would hire low-tier performers to protect the productive ones."

Juking the stats! I'm shocked there isn't a middle management bible dedicated to this.

Where is the hackers guide to middle management?