Employee Evaluation Formulas & Curves 1

[vonsteimel]

Greetings,
I've been trying to resolve an employee evaluation problem that I'm sure is common in any medium/large company.

At my company, employee evaluations are directly tied to raises. It starts once a year when, to leave out the complicated details, the accountant and the CEO decide on a lump sum of raises that can be distributed (i.e. 10K year.etc)

The supervisors then evaluate each of the employees that they supervise using a standard form, which results in a number score. This number score is then input into a formula which decides what percentage of the "lump sum" each employee will get.

The problem occurs when you have multiple supervisors conducting evaluations. Because supervisor "A" may be giving out a 4.5 average when supervisor "B" is giving out a 2.9 average. This problem is amplified by the fact that each score is going to have a direct effect on the amount of raise each employee will get.

Some companies go to a flat "across-the-board" type of raise but we don't want to go to such a system. We want to reward hard-work.

I assume this problem is fairly common. How do bigger companies solve this?

I have written a formula for "equalizing" or "curving" the scores. (See the attached file)

I wanted to see what other companies were doing & seek suggestions before I implement this.
Thanks,

VS

 

[GregLocock]

Where I work we have forced ranking. So at a fairly high level all the engineers are listed by perceived performance (not just assessed by their supervisor) and pay. Then performance is divided by pay, and the argy bargy starts.

If you want to know why forced ranking is a shockingly bad idea as a permanent method then there are plenty of google hits. You might also like to read up on Tournament theory.

I personally can't complain about my pay rises relative to the rest of the organisation, but I also can't believe that such a stupid idea still has legs. We are supposed to work as teams. So why are individuals being encouraged to outperform their peers, and grab all the glory, rather than helping/educating the also-rans and young-uns? As an example, once upon a time, I prepared a graph in excel and sent it out as a png. A lowlife then hand transcribed the data off my graph onto an all but identical one, and issued it as his own work (yes MP, if you are reading this, I have kept that powerpoint). He won an award for that one.


Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[KENAT]

What if supervisor A's staff really are better than supervisor B's staff?

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[GregLocock]

That gets sorted out because supervisors A and C and D will point out that B's guys are a bunch of unhelpful nuff-nuffs. So even if B sticks up for his guys everybody else will enforce the majority view. That part of it is the least worst aspect of the process. Yes it is a bit of a beauty contest.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[tickle]

Or Group A's engineers do not get recognised proportional to the other groups, the good ones (and ones with potential) move on, the engineers left are either below par or less experienced, they do not get "decent" projects or decent roles on projects, they do not develop, the focus is on the other groups so Group A do not get recognised further. The supervisor gets bashed because the group are not performing in comparison to the other groups, he has to step in to get projects over the line due to lack of experience, then is not giving opportunities to the engineers (the ones left with no experience). No work comes in, or is unprofitable.

It then becomes a downwards spiral. Different levels of management focus on (and blame) different parts of the problem and put forward valid arguments as to why there is a problem. But if you do not retain the experienced staff with potential, the rot sets in. The companies need to look at who adds value to the company and provide incentives for them (remuneration, responsibility, training) and also focus on succession planning.

 

[MacGyverS2000]

Things get funny when supervisor's A, C, and D have an axe to grind with B, too... and punish B's minions in the process. It's a sad little game people play.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[vonsteimel]

Yes, this system has to assume all of the supervisor's groups are equal. Though they will probably never be in reality.

It kind of goes back the Winston Chuchill saying: this is the worst system except for all the others that have been tried....

Any details on how your current employers conduct evaluations? And raises?

Thanks,

VS

 

[zdas04]

I've been involved in a number of these "forced ranking" schemes and they all suck. They kind of work OK for the middle 90% of the population. Stars and slugs get screwed/unreasonably compensated. When you figure that in an Engineering sense, a hugely disproportionate amount of innovation comes from the stars, screwing them is a bad idea. Also, a disproportionate amount of your problems come from the slugs that you really don't want to disproportionately reward.

Any plan that has an equal sign is trying to replace decisions with policies. I have found that these schemes tend to drive the stars out and encourage the slugs to never leave.

My first boss had it right. His policy was that he compensated his stars first, and spread the rest among the troops in an absolutely arbitrary manner. He was the one that said, "I plan to pay you so much that you don't think about your pay, because if you are thinking about your pay you are not thinking about the projects that make the company money." He never had a star quit him. He had more than a few slugs go looking for greener pastures. He passed away before the "personel department" became HR and took over the process. Everyone he had nurtured moved to greener pastures within a year of his passing.

If you take the decisions away from the supervisor and give it to an algorithm, the people who make you the most money will be the first ones you lose.

David Simpson, PE
MuleShoe Engineering

Law is the common force organized to act as an obstacle of injustice Frédéric Bastiat

 

[truckandbus]

Each supervisor classifies staff as 1, 2, or 3. Those ranked 2 get nominal % raise. Those ranked 1 get a little more, those ranked 3 get a little less.
Numbers get run to see if the total falls within budget, and then the list get reviewed for anomallys - high salary '3' having larger bonus or increase than low salary '1'.
We repeatedly made the mistake of rounding the increase to get the numbers easier to read and would deal with fallout along the lines of "I got 3.32% inrease and so-and-so got 3.34%"

Then you get in to arguements over the increases between the different departments....

 

[KENAT]

We also have the fun that people are meant to get paid at not much more than around the 50th percentile of their Radford Survey job description.

I was grossly underpaid for my nominal title a few years ago and so got to have some nice big pay raises, but now they're tapering off.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[SNORGY]

Some things just need to be left to human judgement, and to the discretion of those promoted into supervisory positions - if they are indeed correct for the job, their judgement will invariably be sound. As for team play versus beating out the competition, the folks who are always stepping in with a "can I help with that?" attitude have always been higher on my radar than the folks who crank out voluminous quantities of high quality work from their silos. What I have observed is that the individuals who go out there and get engaged at a working level with as many on the team or in the department as possible soon are able to integrate everyone's different perspectives into a grasp of the big (overall) picture, and those are the people who ultimately become more valuable, versatile and promotable.

The only times I tried a hard core ranking scheme to make a staffing decision, I scrapped it and trusted my gut anyway - once to hire an intern and once to help staff a department. Both times, it was the right thing to do.

zdas04 has the right idea.

 

[GregLocock]

"I tried a hard core ranking scheme to make a staffing decision, I scrapped it and trusted my gut anyway"

This happens very often with any formal evaluation process, whether it is for people or design alternatives. Once you've got the stakeholders together, established the factors and responses, and the weightings, and then scored each proposal, the result is often odd. Then you have to reverse engineer the decision matrix and see what you missed out or did wrong. This can be useful. Or, as you did scrap it.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[ajack1]

If you do not have a documented policy heaven help you when someone throws the discrimination card at you. To be fair that might only apply to here in the UK?

 

[CELinOttawa]

Hahaha... No. Not just a UK problem.