Stack Ranking

  • Comments posted to this topic are about the item Stack Ranking

  • I think maybe the manager who thought that system up was a '2'.

  • Thats a crazy system!

    If you did have a group of people to begin with that had people who were 1 and 2, after you had weeded them out you would then be forced to rank people who previously 3,4, or 5 downwards even if their performance improved

  • Sounds utterly ridiculous. Yes, you need to identify the severely underperforming people and either fire them or move them to a different job more suited to their talents, but in this stacking system, if you do that somebody more competent then becomes the "bottom tier" and is suddenly in trouble. It seems to me that the whole system is designed to make less work for managers--instead of fairly assessing their staff all they have to do is stack them against each other using some KPIs and make all their decisions based on that!

  • This type of system seems to be a way of finding excuses to keep wages down to a certain extent. The rankings can be used to limit raises. As there will be a quota of staff in each quota it has nothing to do with ability. It could also be argued that it is open to abuse - ie the top ranked staff just happen to get on best with their bosses. I am glad I do not work in this sort of environment - as I have never been a 5 or a 1 I think I would get lost in the middle ground. It must be horrible to be be labeled this way - the job titles and grades can be bad enough at most organisations in setting the hierarchy let alone having this sort of grading thrown into the mix.

  • I think it's a terrible system. This means you could have a room full of the best and brightest, all of whom would be a '5' in a random sample of workers, but someone has to still be a 1 or a 2 because that's how the system works.

    This reminds me of a couple of other barometers for standards that work in the same way

    - Employment demographics and targets. I'm pretty sure from the depth of my mind that some companies and public sector work places aim to employ a certain percentage of the workforce from minorities or based on gender, so they're letting a persons race, age or gender play a part in their decision making process. Given two applicants of approximately equal ability, they let their targets decide who should get the job.

    - Exam grades. Some exams use a system where only the top n% can get the top grade, which is great in some ways as it prevents everyone passing an easy exam with a high mark but there are so many ways this can work against a person. One exam group may be composed of more talented people on average than another group sitting the same exam. Because of that, someone with a higher actual mark might end up with a lower grade than someone who scored a lower mark but was still in the top n%

    There is nothing wrong with assessing ability relative to others as long as it's used alongside consideration of expectations for the job etc. Scoring someone with a low mark simply because relative to their colleagues, they're not as good as them without considering their raw ability to do the job is the mark of someone who frankly isn't very good at being a manager in my opinion

  • This loathsome ranking system has been in use at Logica, where I work, for several years.

  • This is the "bell-curve" performance evaluation system that is unfortunately, in practice in most IT consulting service based organizations that I know of. The sub-processes have very attractive names, but in the end, it's exactly what is mentioned in the editorial.

    In the end, the managers have a hard time convincing their team why someone was a "3" or a "4" (or in some cases even a "2"). In many cases, the unfortunate team member would have been left out from a band just because of a point or two in the system.

    The good part is that many organizations are now realizing the drawbacks with respect to employee morale and have started experimenting with hybrid systems - which will mature over time, but is a move that I support.

    Thanks & Regards,
    Nakul Vachhrajani.
    http://nakulvachhrajani.com

    Follow me on
    Twitter: @sqltwins

  • As a techniucal hiring manager, I strive to hire only the best, the 1s. If I make a hiring mistake, and I have made a couple in 50 years in the field, I correct the problem immediately. That being said, how do I rate people as a 1,2, 3, 4, or 5 at the end of a rating period. The whole forced distribution system of rating staff is detrimental to maintaining a high quality staff. As I look at young companies that I worked for, there was no forced distribution until HR staff was added. I can't help but wonder if distribution is to help HR justify their existence. The concept sounds like something devised by a 5, who should have disappeared long before his/her rating was due.

    Gary Jones

  • paul.knibbs and skanker, your comments are spot-on!

  • Stack ranking sucks. I Loved doing reviews for my employees for all the teams I managed at Microsoft, but detested the necessity of using the stack-ranking system.

    Microsoft's implementation of it sucks even more for three reasons:

    1) the manager delivering the poor rating has to 'own the message' - meaning they have to convince the employee that the manager truly believes the employee only deserves the poor rating compared to their peers. This is the case even if they strongly disagree with the rating being given.

    2) the rankings are determined for a group by having all the managers get in a room and decide how various team members compare to members of other teams, for team members at the same seniority levels. Many times this is done with very little to allow direct comparisons between people - meaning the people with the stronger manager, or manager more able to vociferously argue for their team members got a proportionately better number of higher ratings (my teams benefited from this under me 🙂

    3) the rankings set by a certain level of managers get tweaked at every higher level of management up to the Senior VP of the Division. This means a manager may suddenly be told 'you have to give out another 3.0' - which is grossly unfair.

    The only way to fairly review employees is to do what I do at SQLskills - everyone is judged on their own merits, without comparison to others. This only works with competent, caring managers though - which is the unfortunate small minority of managers at Microsoft IMHO.

    Cheers

    Paul Randal
    CEO, SQLskills.com: Check out SQLskills online training!
    Blog:www.SQLskills.com/blogs/paul Twitter: @PaulRandal
    SQL MVP, Microsoft RD, Contributing Editor of TechNet Magazine
    Author of DBCC CHECKDB/repair (and other Storage Engine) code of SQL Server 2005

  • Gary Jones (7/20/2012)


    As a techniucal hiring manager, I strive to hire only the best, the 1s. If I make a hiring mistake, and I have made a couple in 50 years in the field, I correct the problem immediately. That being said, how do I rate people as a 1,2, 3, 4, or 5 at the end of a rating period. The whole forced distribution system of rating staff is detrimental to maintaining a high quality staff. As I look at young companies that I worked for, there was no forced distribution until HR staff was added. I can't help but wonder if distribution is to help HR justify their existence. The concept sounds like something devised by a 5, who should have disappeared long before his/her rating was due.

    Gary Jones

    Gary, yes, of course you only aim to hire the best. And that exposes the biggest flaw in the system - the implicit assumption that the distribution of abilities

    will conform to a normal distribution. Whilst I accept the entirety of the population will conform to a normal distribution, tech companies hire in a vertical band well to the right of the median. I did try asking my employer for the justification of the normal distribution assumption but did not get a reply.

  • Paul Randal (7/20/2012)


    Stack ranking sucks. I Loved doing reviews for my employees for all the teams I managed at Microsoft, but detested the necessity of using the stack-ranking system.

    Microsoft's implementation of it sucks even more for three reasons:

    1) the manager delivering the poor rating has to 'own the message' - meaning they have to convince the employee that the manager truly believes the employee only deserves the poor rating compared to their peers. This is the case even if they strongly disagree with the rating being given.

    2) the rankings are determined for a group by having all the managers get in a room and decide how various team members compare to members of other teams, for team members at the same seniority levels. Many times this is done with very little to allow direct comparisons between people - meaning the people with the stronger manager, or manager more able to vociferously argue for their team members got a proportionately better number of higher ratings (my teams benefited from this under me 🙂

    3) the rankings set by a certain level of managers get tweaked at every higher level of management up to the Senior VP of the Division. This means a manager may suddenly be told 'you have to give out another 3.0' - which is grossly unfair.

    The only way to fairly review employees is to do what I do at SQLskills - everyone is judged on their own merits, without comparison to others. This only works with competent, caring managers though - which is the unfortunate small minority of managers at Microsoft IMHO.

    Cheers

    Bravo! Reviews should be a three-step process. 1) is the employee performing their required job duties? 2) How does their current period work compare to their past period work? 3) from the employee's perspective, how is their manager doing in coaching them?

    ---------------------------------------------------------
    How best to post your question[/url]
    How to post performance problems[/url]
    Tally Table:What it is and how it replaces a loop[/url]

    "stewsterl 80804 (10/16/2009)I guess when you stop and try to understand the solution provided you not only learn, but save yourself some headaches when you need to make any slight changes."

  • "There are three kinds of lies: Lies, damned lies and statistics." If we replace the word "statistics" by "the abuse of statistics" then this use of the stacking system is a good example of the third kind.

    If your sample (department, team, etc) is big enough, the stacking system is the ideal one to use. Using the examination example quoted above, the New Zealand School Certificate system in the old days (approx the same as the old English O-Levels) passed 50% of the candidates and failed the other 50%. This avoided arguments about the relative difficulty of the exam between years and made it much easier for higher educational establishments, employers etc to select applicants based on their exam scores. This is much better IMO than a system where (say) 90% of candidates get a pass grade, rendering the necessary choices much more difficult. But - and this is the point - it only works fairly if you have a sample with a good spread of abilities. This is OK when you have a country's worth of exam candidates. But it is also something which a manager trying to get a top-flight team together will be consciously trying to avoid!

    These days, we don't need more information - we're already awash with it. We need more understanding of certain key implications of it.

    Must say that the above comment about things going well until a certain type of HR gets seriously involved rings certain bells with me.

    Mark Dalley

  • Thank you , Ben. Brevity can be so elegant.

Viewing 15 posts - 1 through 15 (of 77 total)

You must be logged in to reply to this topic. Login to reply