Once your organization start talking about performance reviews you usually hear about some formal system with the same structure for everyone involved. That doesn’t really sound like a good idea, right? Why companies are using this approach then?
If you have like a couple hundred people on board C-level exec can’t really say anything reasonable about vast majority of people in the organization. However leaders have to make some decisions basing on employees value, like firing rotten apples or promoting best candidates for managers.
This is the point where management is tempted to build an appraisal system which makes it possible to compare people easily, so all these decision can be made basing on hard data. The system ends up as stiff and structured checklist which produces grades in the same categories for each employer.
And this is utterly wrong.
Actually I believe you can hardly do worse. This approach not only makes an illusion of producing comparable results but also harms relations between managers and their subordinates since performance reviews following this pattern just suck.
Do a simple exercise: take a description of a few requirements and send them out to a bunch of managers working in software development teams. Now ask them to judge each feature in a few categories in a scale from 1 to 5. Let them judge difficulty, work consumption, innovativeness and business value. Grab these numbers from managers and compare them.
You will see that someone hit average of 3,5 and another barely 2,5. You will see how differently people look as specific categories. You will see how vague one-word category definitions are. Basically, you will learn what subjectivity means.
Now, I have a message for you: people are hell lot more complex than software requirements.
If you used the same system for people what you would get is a set of top marks for a handful of organization’s gurus, handful of worst grades for a bunch of incompetent slackers and like 90% of random results for the rest of people.
In uniform appraisal system this is taken as reasonable data which decides on a number of things, starting with promotions and money and ending with general respect. These numbers make or break careers. And yes, I’ve just called this data random.
But that’s not the worst thing which is introduced by uniform appraisal system. Yes, it can be worse.
Formalized, homogenous appraisal system degrades performance review to simple mark trade instead of making it an occasion to exchange feedback.
You get what you measure. If you measure few criteria, and these criteria are uniformed among the organization, you create incentive to fight about better marks, so people would get more money, have better chances for promotion and would be able to boast in front of their colleagues how cool their marks were on the last review.
There is a side-effect too. This approach creates an incentive for managers to run crappy reviews. Instead of focusing on two-way communication, learning what motivates their people, they just go through a simple script: programming three, communicativeness two, quality four, team work two, creativity five, next please. Hey, this is what the system expects from us, doesn’t it?
Running performance reviews is pretty damn hard job. I always feel stressed when I’m going to talk about one’s performance, no matter how official or unofficial it is. Yes, it is easier to just go through a number of marks and call it a day, but that’s not the option which works for reviewed people. Unfortunately not everyone understands that, so we should build systems which create incentives for positive behaviors, not the negative ones.
So while I don’t agree that performance reviews are evil in general, we can hardly think about worse approach in this area than a formalized, homogenous appraisal system which unifies measures among all employees. That’s just not going to work.
Subscribe RSS feed
Follow on Twitter
Subscribe by email
11 comments… add one
All true. But what is the alternative?
A simple-to-implement alternative is to give managers freedom to choice how they would do their appraisals. Let them decide which criteria are relevant for every single member of their team. I guarantee you they will differ.
But that’s not the clue. The clue is you’d force them to think about every employee individually. Oh, maybe not force, but create an incentive. If they fail you have at least some information about managers.
If you have managers you trust I’d encourage you to go further – give them complete freedom how they run their performance reviews.
Let’s assume (and I do assume this) that no matter what system you use it is impossible to get comparable results from appraisal made by a couple of different managers. The only measurable you can get is within a team. Why don’t we give managers freedom in this area then?
What more, if I was a CEO I wouldn’t expect to get a list of marks once per quarter. I wouldn’t care about that. I would expect that every manager would be able to give me their well-grounded opinion about every single one of their team members instantly. And then I’d ask if I can freely share the opinion with discussed employee. If the answer was positive, and wouldn’t turn into a lie after discussing it with the employee, what else would I need?
I am completely on board with the premise of value over structured checklists in employee evaluations. One huge challenge, speaking from the USA, is the employee legal protections that exist. These protections favor the employee when company treats the employee unjustly. These protections force companies to have more quantitative data to support employee hiring/firing/promotion/raise decisions over quality. In a loose corporate system, a employee who is not given a raise or bonus or promotion/etc. can bring a legal case against the company using the theme of being denied based an individual bias against them in some form or another (age, race, religion, weight, hair color, shampoo choice, etc.). If the company has as loose system, the employee as the advantage in the legal matter. If the company can point to a very quantitative system backed by paperwork supporting a consistent application of a reasonable industry wide HR vetted methodology … blah, blah, blah, the company is in a stronger position in the legal matter.
My point is, at least in the USA, that the legal protections afforded an employee work against the qualitative over quantitative approach to performance reviews.
And this is basically sick.
You bring another argument in favor of structured system which I didn’t think about. But it’s like making people’s life more miserable just to avoid a couple of cases. Fortunately in this area US culture hasn’t spread all over the world.
If you do your reviews fairly and your managers are good the risk of having lawsuit on that is minimal. On the other hand I understand that if you have a few thousand employees you don’t really know what kind of manager you have. Then feel free to implement structured and homogenous appraisal system but at least be aware how much harm you do for those reasonable folks working for you.
And be assured best brains are likely to leave soon. They rarely suit very formalized environments.
I think the key thing here is that the review is only one small part of the process, a chance to reflect on goals over a 6 month period rather than much more immediate feedback delivered informally on a day to day basis or through regular 1-1s.
As part of this continuous feedback there is value in talking about specific areas that you as manager consider to be important to their job, this will be slightly different for all employees, even those in the same team with the same job title. In this instance having an objective way to provide feedback and track progress can be a hugely positive thing.
In this case the classification of employees is focussed on the individual and talks only of performance. If done well it’s transparent and the employee is under no confusion over what it is that their manager expects of them or considers important.
I think that the problems arise when the classification scheme is used solely for the purposes of salary review in an isolated remote manner. It’s not clear to the employee how these figures were calculated and indeed how they might go about improving on them over and above ‘get better at programming’.
Managers themselves then feel disconnected from the process and rather than thinking about the underlying benefits of coaching and personal development, they follow the process so that they can tick off the next item of their list.
Neil,
At the end of the day every performance review will be subjective. It doesn’t really matter what criteria we choose since it ends up with one person sharing their opinion on another.
Of course we may bring some more objectivity by adding more and more people to the process (like on 360-degree methods), but it would still be just an average of subjective opinions. And it takes hell lot of time to do it well.
You make here a very important point – as long as manager talks with a team member regularly and delivers feedback whenever there’s a need (and not quarterly or yearly), the review shouldn’t be an issue. However you make one assumption here: whole review process is aimed to deliver reasonable and fair feedback to employers so they possibly can improve.
Now, the problem with appraisal systems is that the goal is usually quite different. Appraisal systems are implemented, because everyone has to have one, or to make promotions and raises easier, or, as John points in his comment, to avoid lawsuits, or to distribute merit pay. Typical appraisal system is treated as an evaluation tool, not as something which should help employees to improve.
And this is where disconnection appears. If we want to have all results easily comparable in the first place, we’re likely creating appraisal system which won’t help people improve. Of course good manager should find other means to achieve goal of delivering feedback (1-on-1? anyone?) but, let’s face it, average manager won’t even think about any other tool besides those enforced by top management.
And that’s why uniformed appraisal system are evil.
Yeah, that’s the key distinction, if the need for review comes from, and is driven from within then it can be an extremely powerful.
I guess this applies to pretty much any management evil. When you hear of teams railing against agile frameworks because they perceive it as micro management from above then you know something is wrong.
Over the past six month I’ve being trying to write a feedback focussed performance review system that’s flexible enough that folk don’t feel like they’re in a box ticking exercise. There really isn’t much in the way of resources available to draw upon so I’m hoping that I can share this via my blog.
Then I’m looking forward to see this.
Actually if I had to think about performance review system the thing I’d focus on would be coaching managers. Any system which helps managers to start thinking, to get out of their boxes and to perceive their teams as groups of individuals would be much better than what we typically see.
There checklists and points always make me laugh. To me, they look just like stats in a role-playing game. I am a 5th level developer (5 years of professional career) with 3D6 in WCF, 1D6+2 in WebForms and +25% to SQL query performance. Real life is far more complicated and smart people who invented RPGs knew that and they created a Gamemaster role.
It is not a coincidence that the more advanced RPG players become, the greater is the power of their Gamemaster and lesser the value of statistics and dices.
Actually RPG-like stats would be crazy awesome. In all RPGs I played you get experience points and stats for achieving something, killing monsters, completing quests etc. Why it isn’t so in many companies?
What more it is you who decide what you spend your level-up points on. And you never get money directly for leveling up.
“Over the past six month I’ve being trying to write a feedback focussed performance review system that’s flexible enough that folk don’t feel like they’re in a box ticking exercise.”
So here it is, thought it might be of interest
http://fragile.org.uk/2010/09/how-to-appraise-a-developer-part-1/