Cost per Unit of Productivity (Employees)
blog.joelx.com
blog.joelx.com
> Find the most productive employee you have
which the author admitted you can't accurately do.
This is a recipe for disaster, not because reducing people to cold, heartless quantitative metrics is necessarily bad, but because this is totally subjective, and will ultimately devolve into simply rating the people you happen to like as "more productive" or giving them free passes with the "special skills/circumstances" column (for example, imagine if that column came to simply represent diversity criteria).
Someone considered for termination due to being low in the CPUP ordering is probably more likely just someone that people don't like, or someone who does a great job but within a business unit that some other political rival wants to defeat, disband, or re-org in order to consolidate internal political control.
Also, it's very reasonable for people to discount profit sharing pay. Such pay is generally less specifically defined in employment agreements, easier for an employer to omit for arbitrary reasons, affected not just by one's own performance but also performance of others.
These are wage risks that a given employee largely cannot diversify away. Whereas with salary, there are many legal protections preventing an employer from omitting salary without documented cause.
The more of your pay that is arbitrarily discretionary and at the whim of your employer's choice, the more variance there is in your overall wage outcomes, and if you're risk averse (as many people are when it comes to wages) you'll very rationally discount the value of that kind of wage stream compared to a wage stream with far lower risk.
I also wonder if this employer does basic things, like giving all knowledge workers private offices. If the employer doesn't do that, but then does judge the workers on their productivity, it's exceedingly hypocritical, and a red flag of a bad employer.
I liked the way this was stated in the classic software productivity book Peopleware (paraphrasing ...)
> If the employer did not care enough about protecting everyone's productivity [e.g. by using obviously cost effective technology like private offices in the first place], then I'm not going to kill myself to make up that productivity lag.
An employer who actually cares about productivity will provide workers with all the things they need to actually be productive: state of the art tools, adequate access to privacy and quiet space, adequate vacation and rest time, adequate work/life balance, positive and collegial atmosphere, transparency about performance expectations and appraisals, etc.
They will make the office (and the non-physical workflow structures of the organization) into a place that facilitates getting work done instead of being something that you get work done in spite of.
Obviously, I can't say how it plays out for this particular employer, but most often all you ever see is that the employer points fingers at employees and makes arbitrary and erratic and politically motivated decisions about who to promote, fire, or give a raise. Rarely will they ever point the finger at themselves and say, "How is it that I am failing as a manager to provide the right sort of workplace that would help this worker reach a higher CPUP [or whatever]."
Given all of this, it's a very reasonable heuristic to simply call bullshit on employers who say this crap. Yes, maybe this is the one-in-a-million employer who actually takes a humanity-affirming perspective to it. But you're probably better off just assuming this employer is not such a rare case, and just avoiding them because of all the red flags this emits.
You raise some reasonable concerns about the fairness of the system and especially if workplace politics could unfairly influence which individuals are laid off. For my business specifically, I am the owner so playing any games with politics would only damage the business and my equity in it.
In general, isn't this how managers determine which employees to promote / keep / fire anyways? They try to identify the worst performers and let them go, and identify the best performers and give them raises or promotions?
My thought is my system just takes that subjective methodology and simply makes the manager think about it consciously rather than subconsciously.
Is there a better method?
If owners enter 'dictator mode' it can be very hard to separate out what is cognitive dissonance vs. what is a genuine attempt to manage a business without politics.
As far as your question goes, I think it begins with a fundamentally different outlook. You write about it in terms of who to promote / keep / fire as if that is the right menu of options to orient your thoughts around.
Instead, I'd say you should think more about how best to develop the workers you have. Firing people ought to be a distant thought that is only brought up under serious conditions, like overt misconduct, extreme underperformance over time or incompetence (that can't be corrected with training and investment).
You strike me as someone who is not thinking in terms of employees as investments that payoff over time, but rather you are using some kind of hyperbolic discount function that only looks at some short window of recent productivity or some short forward-looking window of possible future productivity -- classic short-term thinking that leads to a lot of bad outcomes.
If you just want some kind of cookbook style management process, you can find books endorsed by millionaires, billionaires, political leaders, athletes, coaches, etc. etc. and just pick whichever one tickles your personal confirmation bias fancy. (I'm not speaking just to you specifically, we all do this, certainly me included). So you can always justify pretty much anything you choose by saying, "isn't this how other people do it?" or "Wealthy Person X said to do it like this..." -- it doesn't actually mean you're making the right choice for your business.
The other thing is that you conceive of your employees as falling on a linear ranking, which I feel is misguided. I think it is not possible to put a useful ordering on employees like that, because they are individuals and you can't discount their potential, drives, passions, personal circumstances, etc., all of which are changing often.
There are all kinds of dimensions in which a person can be good or bad in their job for you, and there is almost surely no coherent "norm" you can define on that high-dimensional space that adequately orders employees ... and certainly not something entirely subjective like "who do I happen to think is most productive modulo special circumstances". Even if Person A is "more productive" than Person B, you don't know that will always be true. Maybe Person B is a faster learner, a more creative thinker, or has other latent talents that you have not properly tapped into yet.
What you propose almost sounds like a proto form of stack ranking [0], which is a notoriously horrible way of doing things. But at least with stack ranking, you tended to receive scores from multiple people who function in mostly uncorrelated ways in the firm, so as to (mostly) remove dictator effects or sabotage effects. I don't think you could easily get that in a very small company where everyone knows everyone else and where the founder/owner is highly visible in day-to-day actions.
Overall, I'd just generally recommend a totally different mindset. Stop trying to apportion wage to micro measurements of productivity. Just admit you (and everyone else) can't measure productivity well enough for that to be accurate and robust to social biases.
Don't conceive of your lowest performers as candidates for termination. Rather, spend lots of time first seeing them as examples where your management style, workplace offerings, equip/tools/etc., have not been good enough to coax good performance out of them, and invest in them. View their payoff as something that happens over time, yes still with some discount rate more like bond pricing, but not so extreme that their short-term productivity counts for so much that it could lead to them being fired. If their productivity is low for a long time, after you have implemented many strategies for helping them improve, then you can be more confident that you're making a fair and correct decision to let them go.
I am not sure that what you recommend would work for the type of business I have. Perhaps it makes sense for a product based business where you can invest in someone for a year because they may create a revolutionary change to your product after that. My business is a service business - we operate on very tight margins.
I like everyone I hire and generally wish I could keep all of them. I don't have that luxury. I am forced by the economics of my industry to let people go... so I have to pick using some method. Stack ranking based on productivity seems to be the most fair way to do it. Is there a better method than stank ranking for selecting who to let go?
The brass tacks economics always matter, of course. Your situation suggests to me that you have over-hired, or hired incorrectly. Rather than creating a company culture whereby people are pruned, it would create a better workplace to simply not hire too many people to begin with. Operate in a leaner fashion, even scaling back revenue opportunities and saying "no" to new business when you have to, so as to disallow the company to grow in unsustainable ways that then lead to a hard requirement for layoffs. This won't be possible 100% of the time, but going for a stack ranking approach instead is like you are just giving up on intelligent, sustainable growth and instead saying you will grow randomly and unsustainably, and then simply fire people when you have to pare back.
There is also the question of how much of a financial cushion you leave yourself. Instead of being "forced by the economics of [your] industry to let people go" you should use your knowledge of those industry economics to build a safety cushion that allows you to continue supporting everyone's payroll through the lean times. Yes, sometimes the lean times go on so long that you can't do this indefinitely, but the more you budget for it, the better.
Basically, I do not see anything about your situation that makes it special in these regards. I've worked in statistics consulting firms before, where the work is highly variable and where the business really needs a high level of productivity from everyone. Even in that situation, the company was able to avoid layoffs by planning ahead, hiring conservatively and hiring for versatile workers with broad skill sets, and budgeting for safety payroll cushion to express the company's value that avoiding layoffs until the absolute last minute (and communicating very transparently about it all the time) was deeply important.
If you read more from the Wiki article that I linked, you'll also see some of Deming's well-known criticisms of stack ranking -- and Deming was a well-known productivity expert [0] who helped many Japanese firms gain international reputations for efficiency and quality (firms that faced tough brass tack market economics just like you do, but without resorting to stack ranking).
Many of the negative sides of stack ranking could have deleterious effects on your business even during times when you do not need to fire anyone. It is known to affect morale and productivity very badly. It is known to be highly susceptible to office politics and cronyism, especially when times are tough and employees scramble to secure their position. This in turn often leads to a combative and overtly zero-sum / competitive workplace, which is well-known to be bad for productivity (I imagine especially in a service business).
At any rate, it seems like the specifics of your situation which you are using to justify stack ranking don't actually indicate that stack ranking would be useful or productive for you, and in fact suggest that it might even be counter-productive.
The best recommendation I can give is for you to deeply question and re-evaluate some of the assumptions you seem to be working on (for example, that there's no way to address both the market economic constraints on your business and also the need for company values which preclude frequent layoffs and promote longer-term investment in employees). Without re-thinking some of these assumptions, I think it's doubtful that any method of determining who to layoff is going to be constructive for you.
[0] < https://en.wikipedia.org/wiki/W._Edwards_Deming#Seven_Deadly... >