Recently, in our company there was a decision that everyone needs to fill out skillset spreadsheet, so we could better utilize our resources on internal and external projects. While the whole idea seems to be quite awesome from the manager point of view, I think standard skillset spreadsheets are mean to be faulty by design, because of at least few reasons. Here is my point of view:
There are no standard skills
In almost every business now, simply there are no standard skills. You can’t just say you know “Linux” because there’s a huge difference between knowing how to install required packages to set up a server and writing kernel patches to improve security, yet both require knowledge of “Linux”. Also, there’s huge difference between using Photoshop as a photo-editing tool and using it to do pre-printing fixes on the provided files. Tools we use have become so complex, it’s really hard to keep up with every possible feature they offer. And here we come naturally to my second reason…
There is no good rating system
I’m using tools like Photoshop, InDesign or CSS daily, yet I’d never put 10 using 0-10 rating scale in them on a skillset spreadsheet – why? Because for me, 10/10 means you know absolutely everything about every single feature and module that the software or language or library offers and it’s often simply impossible. I know of CSS problems that I wouldn’t be able to solve and I know that I suck at preparing @media print stylesheets, for example, because I hardly used them in every project I’ve done so far. I would never put a 10 in my Photoshop spreadsheet because my work in it is very specific and I’m not sure I’d be able to work on any Photoshop project I’d get. Not to mention the fact that most of the time I’m using CS3 or CS4 and I’m sure I’d be pretty lame at using older versions of the software that didn’t include features I use on daily basis now.
While we were discussing how to rate our skills, Adam proposed one very good system – every skill would be defined by “Level of knowledge” and “Last used”. This guarantees that even if someone thinks he knows – let’s say – C++ 8/10, but used it 3 years ago, it’s fairly possible that his knowledge is really outdated and he would have to learn a lot. With this reasoning system, I’d also add third indicator that would be “Colleague rating”, put on by someone random that worked with you for a while but doesn’t necessarily have to do the same thing. That means that someone else thinks that your skills are – for example – 8 out of 10 in random project you’ve worked together on and that makes pretty unbiased skillset rating. This is connected to my third reason…
Everybody lies
This famous House MD quote works surprisingly often when it comes to things like skillset spreadsheets. Why? Because people are biased – they are often too fond of themselves and put higher values than they should or too shy and put lower values than they should. This means that “Colleague review” would be a must when it comes to spreadsheet, so random person that worked with you could say how the given skill looks like from their perspective. But yet, people interact. They often like or dislike each other, not to mention usual undermining because of negative feelings between two people. It’s all cool if everyone likes everyone and everyone knows everyone, but in big companies where we often see “rat race” it wouldn’t be as obvious.
Our last and final reason is…
Seeking information is the ultimate skill
I finished technical high school and got a MSc degree in technical degree that is Computer Science. My teachers in high school and then our tutors at university always repeated that – in modern world, where you can’t know everything about everything, the easiest way to get information is to search. With just base knowledge, you can often find very quick how to solve your problem using given tools, so you don’t have to really know a lot about tools. We even had experiments by our university teachers that allowed us using our own notes during exams and tests – because you don’t have to know every detail of what you learn, but if you know where to find quickly what you need, you’re a winner. Taking my “Linux” example from the first reason, you don’t have to know how to write kernel patches but with basic foundations in Linux, C and security you could probably write a pretty decent patch just by digging through the documentation and in end it would probably take you less time than learning all three foundation skills so deeply that you could write this without any external help.
I hope you understood my reasoning and this post will help you improve your skillset spreadsheets, if you’re using them in your company so they become more unbiased. If you’ve come up with new ideas on how to measure people skills in an effective and objective way, feel free to drop me your thoughts in comments or Twitter. Thanks!
You’re so right, but for this world… let’s say, you’re too honest with this opinion. Today’s a lot about pretending skills to others, and looking behind the “masquerades”, to the real competences of typical 10/10 colleagues, is mostly deflating.
I think you’re a bit harsh using the work “fail.” It seems to depend on why the data is being collected and on how it will be used.
If the data sheet is meant to be an accurate reflection of reality, then indeed failure is a likely outcome.
If it’s only meant to stimulate conversation about potentially underutilized skills, then it can hardly fail unless no one really cares.
I’m using word ‘fail’ because I personally think that for measuring skill, spreadsheets fail and that’s basically what this post is about.
Nice post, Mariusz.
Your proposed system has some clever additions to the old-school skillset spreadsheets, but perhaps it still might get biased.
If having a skillset spreadsheet is that important to a manager, he should really run exams about the subjects which matter most for the company. And because you’re right saying that “seeking information is the ultimate skill”, those exams should be made with Googling and maybe even asking guidance in forums allowed.
On another level, a company who has to build ad-hoc systems just to know its own employee’s capabilities is reaching the sad-stage of becoming too corporate. Trust your mid-level managers – or get smaller.