Fog Creek Software
Discussion Board




Joel is wrong about performance appraisals

So says this dude:

http://wmf.editthispage.com/discuss/msgReader$3514?mode=day

-- QUOTE --

BTW, Joel's paper is mostly wrong regarding the review process at Microsoft. He first confuses incentive/motivation, quoting Herzberg, and then fails to mention that Kaplan's, "The Balanced Scorecard" is another HBR classic. While Herzberg's article talks about motivating people, it does *not* talk about evaluating people. Joel's admonition to always give everyone a glowing review is stupid...

-- END QUOTE --

Ralph
Sunday, June 23, 2002

A blind slam without a reference isn't fair.  To which of Joel's articles is "this dude" referring?

Hardware Guy
Sunday, June 23, 2002

frankly, I find the idea that Joel could be wrong about anything, ever, offensive.
How dare 'this chap' disagree with Joel?  and did you notice his lame efforts to justify his opinions using lazy misquotes from unknown authors?
who does he think he is?
...nasty little swine....

outraged, offended, and slightly threatened...
Sunday, June 23, 2002

The guy writing this article appears to have failed to understand Joel's writing. Or maybe I misunderstood Joel's writing.

Quote: "And all of Joe's talk about different people contributing different things (like the guy who made everyone happy, or whatever) sounds like "But we ALL are special!" whining to me."

Obviously this guy is failing to understand how some people in a team can act as catalysts for other members of the team. (And I don't think that Joel was suggesting that this as a sole contribution would be enough).

Robert Moir
Monday, June 24, 2002

I wish he had posted some sort of Joel article reference.

Performance appraisals are for the non-productive anyway.  If productive people cared for the empty opinion of others they wouldn't be productive in the first place.

Setting of a performance object is essentially useless in many environments - the "requirements" and "direction" can change so quickly - even over a three month time frame, that previous promises to achieve something can't even apply in the new situation.

Joe AA.
Monday, June 24, 2002

the "requirements" and "direction" can change so quickly - even over a three month time frame, that previous promises to achieve something can't even apply in the new situation.


  -- Would you agree that a company run in this manner has a very real lack of vision?  (IE No vision at all or a vision that is never realized?)

  Of course a very small company might have to play the shifting goals game, due to cashflow/financing issues, but I think even they would wish they had the luxury of not doing so.

  My conclusion:  Sure, Sept. 11th changes things, but goals are goals are goals.  Unless they aren't - in which case you're in trouble.  (See "The Practice of Management" by Drucker for my $0.02 on Managing by Objectives) ...

Matt H.
Monday, June 24, 2002

A company that understands that it has to constantly adjust to realize it's vision is the only one that will.  Vision doesn't help you avoid an Enron, nor "deregulation" as a couple of examples from the electric utility industry.

To eliminate necessary change is usually a whiny front-line desire of those that don't realize that such an objective is really not in their best interests.

MBO is an interesting study in failure, and the general inability of people to learn from it.

Joe AA.
Monday, June 24, 2002


"MBO is an interesting study in failure"

okay, buddy, there's a big in-your-face kinda statement.

You can to start a thread and defend your statement, or just post it here?  I'm willing to listen, but I'd like to here why.  (Plz don't just come back and say that MBO is locked and can't address change.  I heard that the first time.  I'd like to hear _how_ that is measurably bad, and _how bad_ it is objectively.)


looking forward to it,

Matt H.
Monday, June 24, 2002

>> Performance appraisals are for the non-productive anyway. If productive people cared for the empty opinion of others they wouldn't be productive in the first place.

Joe AA nails it.  IT is very difficult to asses the  productivity and worth of say a Program Manager, a device driver writer, and UI designer and/or implementer, etc.

And yet, it beomes necessary to have some system in place whereby you can coach underperformers.

Interesting thread...

Nat Ersoz
Monday, June 24, 2002

Sure, the setting of goals has its place.  However, if done wrong it can be soooo demoralising.

A smart manager knows his people.  He knows their strengths and their weaknesses.  He knows when they are going the extra mile and when they are slacking.  The smart manager knows his team's dynamics - its what his paid for.

The stupid manager tries to measure his team members like they're inanimate objects.  Subordinate X is superior to Subordinate Y because he wrote more lines of code in the specified period...

If team dynamics and human interaction were this simple, then managers would have no right earning such high salaries.  Smart managers are worth every penny.  The number crunching manager could be replaced by a Commodore PET with significant improvements.

Heres a scenario:  Bill and Ben work on a call centre.  When the queue gets busy Bill just cuts of the calls off on after another until they are all gone.

Bills call response time is excellent.  His satisfaction rate is high (he never answers calls under pressure).

Ben answers all calls properly, even when the going is rough.  His response times are inferior.  On top of this, this majority of a random sample where dissatified.  The reason given was that they had to redial several times before being served.

Bill is promoted, Ben become disillusioned.

(Based on a true story)

Ged Byrne
Monday, June 24, 2002

MattH... why would I want to argue with you about MBO?  The only ones that don't want you to know it's been a historical failure (and do your own research) are the people that are still selling it to you under the snakeoil of the month club. 

Not trying to read your mind, but I would expect your response to any failure point would be much the same as the snakeoil salesmen:  but you didn't implement it correctly.  That's always the excuse for snakeoil failure and the primary means of detecting snakeoil.

Joe AA.
Monday, June 24, 2002

>"snakeoil of the month club" ....


This is great.  No reasoning, and now I'm stupid for not agreeing with you.  What is this, proof by repeated assertion?

I typed in the phrase "Problems with Management by Objectives" into Yahoo.com, got 1 hit.

(If you're interested:

http://www.infed.org/rank/managing_projects/managing_workers.htm
)

Typed in "problems with the waterfall model", got 41 hits.

Now, MBO has been around since 1954, and is still taught in most MBA cirricula.  The waterfall model has been around twenty years less, and is taught and esposed less.  I know that _SOME_ of the extra (x5 or maybe x10) hits are because the web is a CS-centric media and The Waterfall Model is a CS-Centric topic.  (Similar results on searchs for (Waterfall Model I hate it) and ("Management by Objectives" I hate it))

But 41x hits?  That tells me the waterfall model, while used, is much more strongly questions. AND YET EVEN IT is still taught and used in > 50% of companies.

Got a better link?

thanks,

Matt H.
Monday, June 24, 2002

No, no links. 

Tom DeMarco (in places besides Peopleware, like "Why does software cost so much") has probably been most vocal pointing out the falacies of MBO.  But since you primarily seem a Drucker fan, I can understand why you still buy into it.

Don't get me wrong.  Management is a difficult business to do correctly.  But MBO is snakeoil and only helps the salesmen sell it to their victims.

For example, and I don't remember the source of this story either, and the way I heard/read/remember it, the sitting was Russia... so don't any of you Russians out there take offense.

There was this factory in Russia that produced nails.  Things were going along fine.  Then the western concept of MBO was introduced to improve things.

The factory was given the object of producing some amount of tonnage of nails.  The factory responded by producing only railroad spikes.  The Powers That Be soon saw what was happening, and revised the object to state that the factory had to produce some number/count of nails.  The factory responded by producing tacks for picture frames.

The scene changes.  You now see a Russian carpenter holding in one hand a small tack, and in the other an equally useless railroad spike.

The grand fallacy of MBO is easy to see.  It places you, the manager, in a position to focus on the wrong thing.  If you miss your objective then what are you really focusing on?  A failure. 

The worst manager is the one that focuses on failure.  Like driving a car, you will go exactly where you focus.

But the worst, worst manager is the one that believes avoiding the appearance of failure - like by meeting MBOs - is the exact same thing as success.  These are the clueless ones.

Joe AA.
Monday, June 24, 2002

"But the worst, worst manager is the one that believes avoiding the appearance of failure - like by meeting MBOs - is the exact same thing as success. These are the clueless ones. "

A fairly good implementation of clueless would be, for example, claiming that MBO is always flawed and in fact, hey, its "snake oil" because its not suitable all the time and may have been poorly implemented in some cases.

Everything is flawed if its implemented by someone who doesn't understand it properly.

Robert Moir
Monday, June 24, 2002

I've worked at a few companies, and I've never seen MBO implemented correctly (I gave one example earlier.)

If a software system is incorrectly implemented, it falls over.  It obvious that it hasn't been implemented right, and steps have to be taken to fix it.

Unfortunately, the same cannot be said for HR methodoligies.  Almost every large companies have one installed, but for the majority it crashed back in the 70s.

If a methodology can consistantly damage a company over years and years without it being noticed, I see this as a major flaw.

Ged Byrne
Tuesday, June 25, 2002

"Everything is flawed if its implemented by someone who doesn't understand it properly."


Good point.  One of the key factors is the MBO is that the goals have to actually be aligned with the goals (& success) of the business.  Another is that they have to be written by the executing manager and approved by his superior.

That being the case, I think the nail/spike story is, well ... a story.  I heard a similar one in business school about glass. (The company in russia had to produce glass, and was given a tonnage of output, so they produced huge raw blocks that were good for nothing, instead of mirrors.)

I have to agree that MBO would work much better in a more static industry (Office Furniture, Utilities, etc.) than in a newer or more dynamic one.  Still, at it's heart, all MBO is is setting correct goals and measuring yourself against those goals.

I'm not a Drucker Fan more ... I just happened to have read his book most recently.  Someday I'm going to order Slack and "Why does software cost so much", but I've got a pile of Baby books to read first .... :-)

regards,

Matt H.
Tuesday, June 25, 2002

Hey, Ralph -

We're still waiting for the reference to the (alleged) article that kicked off this thread.

Hardware Guy
Tuesday, June 25, 2002

I think Ralph was referring to the article "Incentive pay considered harmful".

It's at http://www.joelonsoftware.com/articles/fog0000000070.html

Having been forced to conduct several painful performance appraisals as a manager (painful both for me and my subordinates), I pray that they be abolished in every bloody company.

BTW, I found this interesting site:
http://www.abolishappraisals.com/

Incidentally, among W. Edwards Deming's 14 admonitions is this one (#12b):

"12b. Remove barriers that rob people in management and in engineering of their right to pride of workmanship. This means, inter alia, abolishment of the annual or merit rating and of management by objective. "
(Read the rest at: http://www.khosla.com/softwaremgmt/deming.htm )

Here are some books by Deming:
http://shorl.com/famohaprajoki

MadMan
Tuesday, June 25, 2002

Deming is not as "popular" as the snakeoil salesmen.  The quest for the silver-bullet must be a strong force indeed. 

And I still stand by my previous assertion - snakeoil is easily detected by it's failure excuse: you didn't implement it correctly.

I've always liked Deming for demonstrating the popular way is not the only way.  Instead of "manipulation of people through their vices (e.g. greed)" as practiced by MBO - he encouraged internal motivation by the individual with basis in their virtues (e.g. love of their work, pride in their workmanship).

And I really like the "obstacle" of meeting specifications - one of the biggest excuses used in our industry to pretend success.

Joe AA.
Tuesday, June 25, 2002

Maybe one of these.

http://www.joelonsoftware.com/articles/fog0000000070.html
http://www.joelonsoftware.com/articles/fog0000000038.html
http://www.joelonsoftware.com/articles/fog0000000039.html
http://www.joelonsoftware.com/articles/fog0000000040.html

Nat Ersoz
Tuesday, June 25, 2002

" Having been forced to conduct several painful performance appraisals as a manager (painful both for me and my subordinates), I pray that they be abolished in every bloody company. "

Madman, could you explain why they were so painful?

I don't see anything wrong with them _if the employee and the manager both buy in to the concept_. At a company I used to work for, we had "appraisals" every 6 months. It was a serious discussion of how I was doing, where they thought I could improve, where I thought I could improve, what I thought of my managers, where I saw my career within the company going, etc. The key point is that my manager took it seriously. He didn't just fill out the forms and then file them away, it was clear that he'd thought about the appraisal, had solicited comments from people who had worked with me on projects, was serious about discussing my objectives and worries, etc.

I suspect this is quite rare. I've also had appraisals where I got the feeling that the person giving the appraisal didn't really take it seriously. There was no thought given to my appraisal before we sat down to do it, I didn't have time to prepare anything, etc.

So why do some companies do it well and others badly? Clearly just setting up a process for appraisals isn't enough.

Incidentally, have we figured out whether the original poster was discussing appraisals, or incentive pay? My "good appraisals" company never said that appraisals were linked to bonuses. Obviously, you knew that you'd be more likely to get a bonus if all your project leaders said you were a great worker... But the appraisal wasn't "about" the pay/bonus.

Adrian

Adrian Gilby
Tuesday, June 25, 2002

. . . and that is the key. A good company has one form of punishment: firing. There is really nothing else you can do to discipline an underperforming employee. A chat, "If you don't get better, you're gone." And then the ax. We're adults, after all.

A.
Tuesday, July 02, 2002

*  Recent Topics

*  Fog Creek Home