Tuning up employee performance
A better way to manage employee skills
Most companies, by their nature, are driven by profit. Since profit is just the excess money you have above what you spent, there are two ways to make your profit bigger: make more money when you sell it or cut your cost to make it, aka revenue generation and cost-savings.
Spending money on things like licenses for software products, that shiny new laptop, or that multi-million dollar backend infrastructure project, are all exercises that routinely go through a cost-justification exercise: If we spend this money now, is it going to reduce future expenses enough to offset the cost? Likewise, keeping things that have been paid for, like software, hardware, machinery, and office equipment in good working order makes sense, too. You paid good money for it, you want it to stay in top working condition for a long time. That’s why companies spend around $600 billion a year in software licenses to keep their software up-to-date.
Why, then, does it seem that companies seem to pay such little attention to the resource they pay the most for: labor. Labor costs can be up to 70% of a companies’ total business expenditures.
It would make sense then, to keep your workforce working in top condition as well. It stands to reason that a highly skilled employees would lead to good business outcomes, or at least, would help eliminate one way of getting bad outcomes. One way to do that would be to approve that conference that your employee asked to go to. It’s a knowledge upskill, and it’s part of the cost of keeping your workforce tuned up. But an often overlooked method, is the employee skills review.
Unfortunately, typical employee reviews don’t look at what you know, they look at what you do. Your pay, title, and position are all driven by how much you can accomplish. Implicit in your ability to accomplish anything though, is that you have the skills to do it. And rating skills can be a tricky thing to do. So the question is:
How can I help a person grow their skills over time, making sure their skills are aligned with the company goals while still giving them space to grow in areas that are personally exciting for them?
What we need is a way to evaluate what a person knows, not what they do. This is typically done by some form of a skills matrix.
Anyone who has tried to use, or been forced to use, a typical skills matrix will inevitably find one of these to be true:
Wrong level of detail
The matrix has either too much or too little detail to be useful.
Not aligned with position needs, company goals, or is outdated
Maybe the skills matrix was accurate at one time, but position requirements may have changed because the company has needed to adapt to a changing business landscape, so the role requirements changed but the skills needed were not updated with it.
Mixing skills (what I know) with role responsibilities (what I do)
Having skills mixed with role responsibilities makes it difficult to assess both. Is the manager who is rating the person qualified to assess their skills? What about the other people in the 360° review?
A better matrix
Facing all these same problems, I decided to look for a better way to measure our employee’s skills. I picked a population of our workforce to start with, our designers, and focused on creating matrix that can:
1. Accurately reflect their current skillset
2. Highlight areas for growth
3. Track growth progress across time
4. Be used to map skill sets to project needs to so we could match a person’s skills with the skillset needed for a particular project
The core of this idea was inspired by the Figma team’s process for building their design team. We started by mapping out what skills we felt were actually needed to be a designer at Lextech. Every company has different needs, and we wanted to make sure our skills were aligned with our companies’ needs. So I ran a mini-workshop with 4 of our designers with varying experiences and backgrounds.
Step 1
First, I asked them to list out all the skills they thought were necessary to be a designer at Lextech, and to group them in whatever way made sense to them, no constraints. Then I showed them the company goals, and asked them to add any additional qualities that they felt they had missed initially, but that should be added to support the company goals. The result was 4 boards of concepts that were organized however the designer wanted to.
Step 2
The next step was to aggregate all the various answers into buckets of similar things. For this step, it helped to try to group them starting with the most core set of skills, like sketching, typography, and color, and then moving to the more esoteric or advanced skills like business acumen, curiosity, and mentoring. We did 4 rounds of this, each round we’d carryover the unmatched items from the previous round.
Our final aggregation looked something like this. Keep in mind, this is what we felt like was needed for us, at our company. If you’re looking at it thinking, “Shouldn’t such and such a skill be core to being a designer?”, then you’re applying your particular lens, which is highlighting the biases and nuances that come with your knowledge history and work experiences, which may not apply to other companies.
Step 3
With the initial aggregation completed, we needed to group them into buckets to which we could apply meaningful labels. The rows are the skill level which correspond roughly to a job title. The columns are just groupings that felt meaningful for us. There wasn’t an easy way to do this, it took a few attempts to feel good about it. Here you can see the “matrix” starting to take shape.
While this captured the essence of what we wanted to track and would be easy enough to update, it still had two major problems:
- It didn’t have a good way to track progress
2. It wasn’t easy to compare across individuals
After many explorations and revisions, I settled on a representation that was more or less like a sunburst diagram. The skills are built from the inside out, so the core skills are in the center ring, and then more detailed or specific skills that build on that core skill are on the outer rings. The chart is roughly split in half, with the top half containing all the hard (craft) skills and the bottom half containing the soft skills.
The degree of saturation represents the competency in that skill. A gray cell means no competency and a fully saturated cell means mastery. It’s a bit subjective, of course, and its not even 100% accurate as far as sunbursts go, because there are skills in the outer rings that don’t necessarily have corresponding core skills. But again, its tailored to our companies’ needs so its accurate for us.
One of the key attributes here is that skills are not only relevant to what is needed for the company now, but also what will be needed to support current and future company goals.
For example, technology is constantly evolving. Not so long ago, you’d be happy to hire a designer who knew how to design for mobile devices. That’s table stakes today. Now we are up-skilling to make sure our designers know how to design for ML and AI interfaces. That type of fluidity in the skill matrix is what makes this so powerful. But rather than just constantly adding to it, the design is constrained such that it cannot be endlessly expanded. So choices must be made what stays and what goes, and this can be done yearly, or on whatever schedule is appropriate for your company.
Putting into practice
Reviewing a person’s skills was relatively simple. We either had them self-evaluate, or had a direct manager do the initial evaluation. Then the manager and employee would have an open discussion about how they each felt each skill should be rated. Once that was complete, the manager asked the employee to pick 3 skills they wanted to work on for the next 6 months. We indicated those simply by putting stars on them. This made it easy to keep track of what they should be working on.
One thing we thought would be useful for us was a way to see a blended skill set. For example, if we have an upcoming project that has budget for 2 designers, and needs someone with strong prototyping, visual design, understands Material Design, and has some AI / ML knowledge, how do we find that?
By designing the matrix to use colors of varying levels of transparency, we were able to create a Figma prototype that stacked all the designer skill matrices into a single view where designers could be toggled on and off and the resulting skill matrix indicates the strength of their combined skills.
What about extra skills?
One thing we found out relatively quickly was that people have other skills that, while not part of our defined set, are nonetheless valuable skills to be aware of. Either they are something a person wants to learn, or they are a skill they already have that we might be able to utilize if we know about it.
Enter Adjunct Skills.
We added an outer ring to the diagram to capture what we call adjunct skills. These can be anything at all, but should still be somehow related to the main defined buckets. For example, one of our designers is exceptionally good at drawing, and even used to draw comics for her college newspaper. It was a skill that ended up being super useful on a project where a client wanted a video to get people excited about their new product, and they wanted a warm, friendly vibe….perfect for a hand-drawn look. That skill wasn’t captured in our original matrix, so we augmented it to allow things like that to be added.
Go forth and prosper
With that I present the final version of our skills matrix. We’re giving it to the community for free. Just fire up Figma and look in the community section for the Skills Matrix template from Lextech. You’ll have to do all the initial work to figure out what skills you need for each roles you’re supporting, but the template will provide an easy way to create the matrix once you have that defined. If you find it useful, leave comments in Figma, we’re always happy to talk about how to keep our most important assets, our people, tuned up and working at maximum effectiveness.
Does your company need help making your employee experience better?
