Small Orange BurstLarge Blue BurstLarge Yellow BurstLarge Green Burst
The Inventivity Pod
Better Employee Evaluations

How do you measure the performance of people whose achievements are hard to measure? Building on the work of Harold Fethe, Jeff Lyons founded MindSolve, a company that developed a technology which made employee evaluations more accurate and more reliable. The company did well and was sold, and Jeff made the challenging transition from founder to employee. A self-described “nerd,” Jeff as a kid used to secretly reprogram Tandy computers at the Radio Shack in the Jacksonville mall. He said “not a lot of planning was involved” in his career, “it was more “just being open to stuff and people who say, ‘come solve this problem for me.” *This episode was originally released on August 14, 2019.*




Intro (00:01):

Inventors and their inventions. Welcome to Radio Cade the podcast from the Cade Museum for Creativity and Invention in Gainesville, Florida, the museum is named after James Robert Cade, who invented Gatorade in 1965. My name is Richard Miles. We’ll introduce you to inventors and the things that motivate them, we’ll learn about their personal stories, how their inventions work and how their ideas get from the laboratory to the marketplace.

Richard Miles (00:38):

I’m going to call HR if you work for any size company, that sentence has the appeal of I’m going to tell your mom, but it turns out HR has a fun, sexy side. I’m your host, Richard Miles and today we will be talking about how much fun it is with Jeff Lyons, founder of a company called MindSolve and currently the Senior Vice President of Global Professional Services at Sum Total, welcome to the show, Jeff.

Jeff Lyons (01:00):

Thanks, Richard. Excited to be here.

Richard Miles (01:02):

So Jeff, when I look up fun and sexy in the dictionary, there’s a picture of you, right?

Jeff Lyons (01:05):

Definitely. And HR as well.

Richard Miles (01:08):

I also forgot to mention actually the pinnacle of your professional career has been serving as a board member on the Cade Museum, right? It’s pretty much downhill from here.

Jeff Lyons (01:16):

Yes, thank you.

Richard Miles (01:17):

And normally we’re talking about topics that are unfamiliar to a lot of people, advanced medical technologies, engineering marvels, that sort of stuff. Today, we’ll be talking about something that actually most people understand pretty well, performance evaluations, training skills management, but let’s talk about first of all, how those things in an organization can be a problem or at least what was the problem you saw in organizational process and what was the solution that you came up with?

Jeff Lyons (01:43):

So the first thing I should say is that the core technology that MindSolve took to market was not something that I came up with or, or that the folks at MindSolve really came up with. It was the brainchild of a guy named Harold Fethy who ran the HR function at a pharmaceutical research firm in Palo Alto. And so what was interesting was the problem, there was a unique, because he’s trying to do performance assessment with a company full of PhDs that are going to argue with any kind of measurement model or metric or calculation you can put in front of them.

Richard Miles (02:12):

What year are we talking about Jeff roughly?

Jeff Lyons (02:14):

This was 95, 96. Um, so sort of just prior to the founding of MindSolve, and so it was an idea that Harold had, and then we help develop the technology and then licensed to take that to market.

Richard Miles (02:25):

So what is the problem in principle that you were trying to solve?

Jeff Lyons (02:29):

So the main problem is how do we evaluate employee performance in a way that’s relevant and in a way that has a lot of quality in the data where you can make solid decisions on it, but in a way that’s also easy. Performance appraisal, as you said, everybody’s familiar with it, universally, everybody hates it. It doesn’t matter what you do. You’re never the popular guy walking into the building when it’s performance, appraisal time. But I think we actually did come up with a way to make it pretty sexy and pretty easy and at Aliza because they had a very high bar for data quality and for a robust measurement metric behind everything that was challenging to do in a way that was fun and easy. And Harold had an idea of doing a visual ranking. And I hope talking with my hands on the podcast that comes through, you could describe this to our listeners, but a way of doing just a very simple drag and drop stack ranking on a screen and took Aliza from a process that had very high data quality and was well-respected, but was miserable and onerous and people would do over the weekend with a case of beer and complain about it to something that people were finishing very quickly on time. Not only felt good about the validity, but it was easy to do, and that helped also make it well received. And so that translated very well to a broader audience, that wasn’t a company full of PhDs. And that’s what helped grow the company.

Richard Miles (03:45):

Instead of a long series of questions, for instance, like, is this employee good at X, Y, and Z, it would be more of a graphical interface or?

Jeff Lyons (03:53):

Exactly. So what else it was coming from was an idea called paired comparison, which has a lot of data validity. And you build up a data model by comparing A to B and then B to C and then C to A, and kind of making these one by one paired comparisons. And so the psychometrics behind that are great, but people are pretty skeptical and it’s a hugely painful exercise. Normal performance appraisal looks at every employee one at a time. And you just do like a one to five ranking on a bunch of questions and people try to make it better by asking more questions. And there’s a lot of counterintuitive stuff. You get better data up until about eight or nine factors or questions after that, your data quality drops way, way off. The more questions you ask because people get tired of it and they just start Christmas treeing. So this was a way, instead of doing each employee one at a time, we would take your whole team, put them on screen and say, let’s talk about communication skill, put your best communicator at the top, worst communicater at the bottom, kind of rank people in there. And then we’ll look at decision making and self management. And we ended up with about five criteria for most employees. And I think we added two or three extra for managers. So it was very simple, very fast, but we did a lot of work to look at the data quality and we ended up with very, very good decision quality coming out of the exercise.

Richard Miles (05:05):

That’s fascinating. I wish they had had that when I was in the federal government, which is probably the worst possible example for performance evaluations. But I remember in the army, they had a problem with the officer evaluations and that there were really only two types of officer evaluations. One, your the next Dwight D. Eisenhower and you should be promoted immediately. And the other one is you’re basically a trader to your country and you should be taken out shot. There’s nothing in between. And then the way the army saw that problem is it sounds like something similar. What they started doing is they started putting, I think they called it a diamond and it was after you’d gone through all this verbiage of how wonderful this person was. You were forced to say, well, this person is among the top 5% of officers I’ve ever commanded and so on. And the second and third tier, but then they would add a reality check and that you could then check what your average was. So you kind of knew that this raider was full of it because he gave everybody top diamond or whatever. It’s something like that. Is this something similar in principle to where you’re sort of forcing an accountability so that you can’t just go on and on about somebody’s qualities without comparing them to something? That that kind of it in principle?

Jeff Lyons (06:10):

You’ve hit on a lot of really, really dense stuff. There’s a lot of psychometrics wrapped up in what you just said. So the first is, yeah, if you evaluate people on a 10 point scale, you might have one person who rates his employees, all eights, nines, and tens, and then you have another person, especially in a room full of PhDs. Who’s just much more critical. Nobody’s perfect. And he rates all his employees five, six, seven, right? So the best guys getting a seven versus the best guy getting a nine. And when you want to make a decision across a broad organization, that’s not really fair. So the first thing that we did was we had to normalize that data. So, you essentially come up with a percentile. So what’s my number one compared to your number one. And we were doing multi-raider assessments. So it wasn’t just a manager evaluation.There were peer evaluations, direct reports would evaluate managers. And so you ended up with a lot of different perspectives about a single person’s performance and backing up for a second. When we took the data, we worked with a couple of really great people on the data model. And one of them, they had the advantage of being in Palo Alto is one of Harold’s close friends is a guy named Brad Ephron who was head of Stanford statistics department at the time, and a MacArthur fellow studying small datasets statistics, which is exactly what we had, right? So we’ve got not classic statistics, but five or six or seven ratings about an individual from a lot of different people. So he worked with us to say first normalize the data average that, and what we found was the absolute rating matters, right? A 3 out of 10 is not as good as a 9 out of 10, but the relative does as well. And what, having multiple people on screen at the same time does use your thinking, not just about what is decision making, but you’re thinking, how does Richard make decisions? And I can benchmark that against how does Jeff make decisions? And it helps me as a evaluator ground, something in reality and make better decisions. And there’s also an element of fairness to it. And then you mentioned kind of this idea that we would call like a forced distribution. Like if everybody fits into a bell curve, you can’t have all tens, you can’t have all ones. Where we ended up after lots of trial and error and back and forth and working with people is that it would be invalid to look at a large group of folks and not make decisions about who you’re starting five are, and who’s going to be cut from the team, right? You’ve got to be able to make those hard decisions in any organization. And it’s difficult because people say, well, we only hire A’s, everybody’s an A, but then you can’t get anything done if you’re not able to make those decisions, but we would not force a ranking. You could tie people, you didn’t necessarily have to fit percentages into those sections of the diamond, but you also couldn’t be flat. And what we would do is provide reports back to show where there wasn’t good differentiation in the ratings and go ask the question and you will get situations where we put our starting five all with this manager. So they’re all going to get high ratings or vice versa, but it was pretty rare. And you could look at the data and at least ask the question of, are we making a good, valid decision?

Richard Miles (08:57):

So you started out trying to solve the problem, or at least make more efficient performance evaluations, but then the company MindSolve that you originally founded, started doing other things, right? Like skills training and other types of management process. Can you describe, or the evolution from going to the performance evaluations to the other function.

Jeff Lyons (09:14):

Yeah, absolutely. What got us into more things was that we licensed that technology back from the company we built it for and started selling it to other companies. And what happens, I think this is true of almost any startup situation is if you go in and you help someone solve problems, they turn out to have a lot of problems and are struggled to solve them. And so they end up giving you more work. So if you look in HR, there’s a bunch of different functions. There’s performance, appraisal, there’s compensation, succession planning, learning, and development. And so you do good work here and they say, well, now we want to push that data into our comp process. For example, we use Excel spreadsheets, it’s miserable. We need to automate it. Can you help us automate that and just tie it right in that was the first sort of adjacent space we went into and then kind of worked our way around the wheel of HR. As customers started asking us to do more stuff. So we really grew in a direction dictated by our customers or requested by our customers.

Richard Miles (10:07):

Is there an optimal size of company that’s sort of like your ideal client for whom this is the most useful? Is it relatively small companies that for them you’re taking a huge burden in terms of HR off of their shoulders, or is this ideal for our company of say a thousand employees or more?

Jeff Lyons (10:24):

I think there’s a better ROI larger. And we used to talk about, if you’ve got 10 employees, you can kind of sit around a table and do this.

Richard Miles (10:31):

And rank them one through 10.

Jeff Lyons (10:32):

Yeah, it’s pretty straight forward, and everybody knows everybody. And the value of automation is greater when there, the data gets so big, you just can’t manage it. Compensation is a great example. People would send Excel spreadsheets out to every manager in the company, pull those back together, copy paste. It was a huge just labor problem. If you only have a few dozen employees, anything about maybe a hundred and fewer, is pretty easy to do, above that it gets very difficult.

Richard Miles (10:58):

And so some guy or gal spend their entire day just trying to figure out what everyone should get paid.

Jeff Lyons (11:02):

Yes. Every case that they’re tested around for everybody versus real time, everybody’s kind of in the same data.

Richard Miles (11:10):

Now, you have had as an entrepreneur being in that field, sort of one of those experiences that is both, I guess, a Mark of success, but also a challenge. And that is a company that you helped found, MindSolve became acquired by another company or sold. And then you became an employee for that company. So you’re making the transition from being the top guy to being a guy who probably has to fix a lot more your own coffee and that sort of stuff, right? So tell us, what’s that like mentally or professionally, how do you make that transition from being the person who started something to being the person who is at work.

Jeff Lyons (11:41):

I feel like I should lay down on a couch for this part of this session, that you’re, there’s a lot of scar tissue, your brain,

Richard Miles (11:47):

I just started my clock. I am billing you for this job.

Jeff Lyons (11:51):

Well, first I’ll correct. You going a bigger organization was nice because then you actually had people who would help with administrative stuff. At a small startup we were making our own coffee, we would draw straws on who got to clean the bathroom. You know, the biggest thing though, was the change in the level of control that you have. That was hard. But I think as we got closer to our acquisition, I was really becoming aware of our limitations, which is a really polished way of saying I had no clue what I was doing. And so we had kind of maxed out what we could do with the organization. We needed more funding. We had bootstrapped the organization, meaning just grown out of revenue. We weren’t burning through a ton of VC money, but we also a couple of guys straight out of college who had no idea about enterprise software. And so we really didn’t know how to sell well. And we had kind of maxed out the organic growth model. So I was actually very excited about talking to people who I thought knew how to run an air quotes, real company. There were definitely a lot of frustrations. Things move so much slower. I was not very politically astute at MindSolve our, our decision making model was yell at each other until somebody gave up and that did not serve me well as part of a bigger organization. And then I came to find out that,

Richard Miles (12:58):

So you’re really a consultant is what you’re telling me. You just tell other people to yell at you. And it sounds like a title of a great book or, you know, yell until you win right?

Jeff Lyons (13:06):

It’s probably a best seller, but it’s not a very good model. I’ve gotten definitely better models since then. But no, I think we definitely learned a lot post acquisition about the corporate world, how to sell to that world. Surprisingly, there were a lot of things we lucked into doing better at MindSolve. Then we’re done at the big publicly traded company that we went into. And we found that after a few years, that company was acquired by a private equity firm who was extremely focused on operational efficiency. And we looked at massive changes to how we approached management. So that was a big learning curve.

Richard Miles (13:39):

One thing that a lot of people talk about is AI, artificial intelligence and it’s going to take everyone’s job, right? Is this a sector loosely described as you, you weren’t consultants, but basically you are helping businesses do their business better, and by making the HR process across the board more efficient, is this something that you could write into a code, right? Where basically you’ve now got an automated way to swart and judge employees and give them training and so on. So is this in any way going to be, or is it already being affected by AI?

Jeff Lyons (14:10):

It is, at some total and it’s Skillsoft we have AI built into our code now and I think it’s an amazing tool. I think it can help you, but I don’t see it really replacing management judgment strategy, things like that. A good example is we use AI to look at, what do I know about you? What do I know about folks who are similar? And we can recommend, for example, developmental training, that’s better for you than if you just did a random search and found 200 courses on management communication. We’ll find the one that’s most relevant to you, almost like an Amazon matching, but there’s limits to that. As you know, you go into Amazon and you’ve bought a bathtub. Amazon thinks you want to buy five more bathtubs in the next week. It makes no sense, right? So there are those kinds limitations.

Richard Miles (14:52):

I stop at three bathtubs. I never buy that forth bathtub.So we’re not at a point where you ask Alexa what the weather is and she says, Jeff you’re fired, right? We’re not there yet right?

Jeff Lyons (15:02):

I don’t think so. And I think there’s cultural hurdles to that as well. People want a human safety net on that stuff. I think the technology can get you closer to a small set of decisions with good data to help you make a decision. But I think unless it’s just sort of a repeatable cookie cutter, kind of a problem, I don’t see AI solving a what’s best for the company.

Richard Miles (15:23):

And it seems to be the consensus on AI is that it will take away some jobs, but it really just helps people do their existing job better because it cuts out some of that mundane data gathering, I guess, or sorting. Right?

Jeff Lyons (15:34):

You know, I think people never ask the question of what new jobs is AI going to create. Right. And people think, Oh, well it’s just coding AI. It’s not that at all. What we saw with our technology is HR is spending 90% of their time on tactical logistical, moving data around not really adding value stuff. And when we can automate that, it frees up their time to do interesting things right? Drive the strategy of the business, which then creates more work and more growth and all of that. We never really downsized HR because we automated part of what they did. We freed up their time to add more value, to do more things.

Richard Miles (16:06):

So Jeff, now we sort of shift to the best part of the show and the one most likely to get subpoenaed in a few years. And that is what were you like as a kid, where you smart, curious, are you just someone whose parents drop them off at the mall as fast as they could, you know? And your a Jacksonville boy as well, so tell us a little bit about growing up in Jacksonville. What were you like? What did you do? That sort of thing?

Jeff Lyons (16:25):

I was a nerd that kind of sums up most of it.

Richard Miles (16:28):

It’s amazing how many Radio Cade guests describe themselves as nerds, it’s gotta be over 90%. So we’re doing something wrong here. I don’t know.

Jeff Lyons (16:33):

You’re definitely hiring to a profile. Look that just cuts out about 20 minutes of description right? Um, I was not at all athletic, I was super uncoordinated. I liked to do a lot of different creative stuff, all the normal nerd things in terms of reading and movies and watching Star Trek and I never really got big into the Star Trek versus Star Wars debate. I was more of, we can like everybody,

Richard Miles (16:58):

We can all get along here, we can.

Jeff Lyons (17:00):

Yeah, exactly, always did well in school to spite myself. I never applied myself at all until I got to college.

Richard Miles (17:07):

So you’re a little bit younger than I am. What was the cutting edge technology when you were say in ninth grade, what was the thing that everyone was talking about? Can you remember, or that you just had to have.

Jeff Lyons (17:17):

This is horrible. What we used to do was go to the Radio Shack in the mall and they would have their Tandy computer sitting out there and you could walk up and immediately just interrupt it and write little basic programs to scroll words at random, across the screen and do stupid stuff like that.

Richard Miles (17:30):

So Radio Shack, Tandy computers, maybe you are as old as I am. You just look, younger.

Jeff Lyons (17:36):

Keyboard built right into the monitor. You know, that kind of thing. I mean, that was just when Atari was coming out and Kaliko Vision and, and television and all that stuff. So that was kind of the hot stuff we wanted with just the home video games. We would spend all our time at the mall, arcade,

Richard Miles (17:52):

Re-programming. Okay. Was there a certain point in your childhood or later in high school where the idea of going into business of some sort of, kind of entrepreneur appealed to you? Or did you think about it? Did you have your own business? Did you know lawn business or whatever in high school, or did that come later?

Jeff Lyons (18:07):

I always worked. I was cutting yards when I was young. I worked through high school at a shoe store. That’s a nice embarrassing podcast we can save for later time. But I was never, I need to go start a business or dream of being an entrepreneur. It was more, I needed money.

Richard Miles (18:22):

So it’s a fine motivator. It works for a lot of people.

Jeff Lyons (18:27):

It went from, you know, wanting to be able to play video games at the mall to wanting to buy beer. There are always staples of life that I needed. No, it was more about that. And I think that’s one thing that served me well, it’s always had a decent work ethic. I was never afraid of working late.

Richard Miles (18:41):

Now you come from a family of engineers. Correct? Your father is a civil engineer. Right? And you have a couple of siblings that are engineers?

Jeff Lyons (18:47):

I have an older brother who yeah designs subdivisions.

Richard Miles (18:51):

Alright. But your degree was in, what? Was it software engineering?

Jeff Lyons (18:55):

No, my degree was in mechanical engineering. So back to your question of wanting to start a business, now, I thought I’d go into engineering and I used that approximately zero days after graduation.

Richard Miles (19:07):

So you graduate your mechanical engineering degree and what did you do?

Jeff Lyons (19:09):

Well, I was working part time for some folks in Gainesville doing software development. That’s what got me into software and then when,

Richard Miles (19:15):

Again, what year are we talking about here?

Jeff Lyons (19:17):

I started working with them in 90 and I graduated in 94.

Richard Miles (19:22):

So software was still kind of in its infancy in terms of,

Jeff Lyons (19:25):

Very much so.Yeah. I mean, we were writing really rudimentary code, but also doing really neat stuff. We doing three dimensional models and walkthroughs of, of hotel ballrooms, really, really neat stuff. And when I graduated, we had been developing some software that we decided to take to market. So that was kind of the first startup pre MindSolve, which was a big failure, but fun. And so I had this offer to come be employee number two, working out of a defunct dentist office in Gainesville. And my other offer was a company that was in the fortune one at the time. And so, uh, those were the two ends of the bell curve. And I said, well, I’ll go give this a shot. And if it doesn’t work out after a year, I can go back to being an engineer. And I did that these little one year, i’ll just give it one more year for quite a while and that led to today, basically. Yeah, so that was the last time I got a job was straight out of college.

Richard Miles (20:17):

Okay. Well, I hope you’ve worked on your resume recently.

Jeff Lyons (20:21):

Yeah. There was not a lot of planning involved or this was not a, this is what I want to do with my life. There was a lot of being open to stuff and working really hard and people going, Hey, come solve this problem for me.

Richard Miles (20:31):

Well, so that’s kind of a nice segue into my next question is asking you your words of wisdom and maybe you don’t have any words of wisdom, Jeff, I don’t know, but most people do or they make it up on the spot, but let’s say you magically encounter that the 22 year old version of Jeff Lyons, probably in the arcade at the video games, what would you say? What would you tell him aside from always wipe off the fingerprints, what would you say to that person?

Jeff Lyons (20:53):

You know, it’s really funny, I’m really of two minds of it because I think I’ve had a really fun life. I think it’s been really rewarding and I’ve liked the journey, but there’s a big part of me saying, don’t do what I did. I mean, we made like every mistake you can make. I was very lucky to have great mentors and advisors early on, right? Even though one of my co-founders, Dan and I were sort of straight out of college. Our third co-founder was a guy who had been an entrepreneur for a long time, was able to give us great advice was a very calming influence on, on a couple idiots, straight out of school. So I did have that, but I still think, just get more advice of people who had done it. There was no real startup community. And in Gainesville, um, as you said, software and the technology,

Richard Miles (21:32):

There wasn’t a startup community until like 2006 or 2007? You waited a long while.

Jeff Lyons (21:35):

This was before boom. I mean, there was no model. And so we were just kinda making it up as we went along and our story is great and it sounds fun and everything until you realize that we had a competitor of similar size that we had better technology, but they knew how to sell things and were connected and invested. Right? And that company later sold to SAP for $4 billion. So I probably would have preferred to run that company. Um, all things being equal so,

Richard Miles (22:05):

Well then you wouldn’t be in a booth with me, it’d be on your private jet somewhere. So lets just be honest here right.

Jeff Lyons (22:10):

So we probably tried to do things too much on a shoe string. I think being well-funded especially now is even more important. So that’s a pretty easy lesson to share is don’t be afraid to give up a little bit of control to people. You’d benefit from them having a little bit of control and who can bring a lot of funding and not suffocate the business.

Richard Miles (22:28):

Well, it’s interesting because you hear a lot from other people saying, give up control, any control at your peril and don’t take any money because they’ll take over and so on. But it’s interesting counterpoint that that may limit a lot of what you can actually do. You don’t have the resources.

Jeff Lyons (22:43):

Yeah. Very much. And I’ve seen the downside of that as well. The other thing I’d say is more on a personal level versus a professional for me coming out of engineering school and just being a very technical oriented type of a person we joked around before about kind of the communication style and the debate style that decisions got made. But in reality, it took me about 10 years to realize that other people have feelings and that most people don’t enjoy vigorous debate as much as I do. And that I think held me back from being an effective leader for a long time. So to somebody who can recognize that handicap in themselves, paying more attention to the people side versus the technical side will serve you very well.

Richard Miles (23:25):

Well, Jeff, my invoice for counseling is already hit probably about a thousand dollars here. So I’m going to have to wrap this up, but Jeff Lyons author of the soon to be written book yell until you get what you want. Jeff, thanks very much for coming on to Radio Cade, wish you all the best in your professional career. And I look forward to having you back on the show.

Jeff Lyons (23:43):

Richard, it was a lot of fun. Thank you.

Richard Miles (23:45):

I’m Richard Miles.

Outro (23:46):

Radio Cade would like to thank the following people for their help and support Liz Gist of the Cade Museum for coordinating and vendor interviews. Bob McPeak of Heartwood Soundstage in downtown Gainesville, Florida for recording, editing and production of the podcasts and music theme. Tracy Collins for the composition and performance of the Radio Cade theme song featuring violinist, Jacob Lawson and special thanks to the Cade Museum for Creativity and Invention located in Gainesville, Florida.

share this article: