Interview with HR.com, part two: T.J. Rodgers on Compensation and Goal Setting | Cypress Semiconductor
Interview with HR.com, part two: T.J. Rodgers on Compensation and Goal Setting
DC- No Excuses Management describes your compensation practices. Have those changed at all?
TJR- I called the second chapter in my book Reviews and Raises: How to Sustain Outstanding Performance. Looking back I would change the title to Reviews and Raises: How to be Fair. I no longer believe that giving raises by a certain algorithm creates outstanding performance. In fact, if you go back to our underlying philosophy of what motivates people then you know it can't be true.
On the other hand, if you look at the focal review methodology, ranking and rating performance, giving bigger raises to people who perform better and making equity adjustments to drive toward fairness these are all valid ideas.
I would publish everyone's salary and willingly defend what each person was paid if I thought I could pull it off—and I have pulled it off with the VPs—just not with the whole company. At the VP level they all know everybody else’s salary. They know I am fair. That's not to say they don’t think I make mistakes. However, if one of them gets a low raise that person is right in my office because he or she knows it means I believe they have a real problem. He knows he has a choice; either start looking around for a new job or come in and figure out how to fix the problem.
DC- You had a very demanding goal setting process. How has that evolved over time?
TJR- The process described in the chapter Real Work: How to Set Demanding Goals and Meet Them has been one of the most challenged systems at Cypress Semiconductor Corporation. When the company was started, we were tightly managed. I don’t mean micro-managed. It never meant, I give orders and you follow my orders. What being tightly managed means is you tell me what you are going to do in detail and I will take the time to understand that detail. When we have agreed on the right thing to do, I want you to do it on time.
I believe it is imperative for any manager to be immersed in the details. In the beginning we required managers to meet with their direct reports weekly and discuss their mutual agreed upon written list of goals. A goal is any activity that was important and measurable. It could be financial, technical, or administrative. It could take anywhere from a day to a month to achieve. We actually had software track the goals in aggregate to make sure that the company is on track. We reported delinquency rates in aggregate at the vice president level.
It is all about a fetish for execution because if you haven't done what you said you were going to do it doesn’t matter if you have a good plan or not. You can be a great strategic thinker and nothing comes of it.
In the goals meeting, the manager and staff would talk about what got done and what did not. At this stage the manager might set four more objectives, but this step must be collaborative. The employee has a right to say I can’t do four things next week, you better tell me the two you really want done. This collaborative aspect is important because one of the ways to ensure a lack of performance is simply to cause people to be signed up for more than anyone can achieve.
We tracked goals along two dimensions, by manager and by project. So you could see all the goals up the management hierarchy. A VP could see everything going on in his or her division. At the same time, you could sort by project, so that projects as executed by different parts of the organization could be tracked.
We had the computer tools that supported the goal setting process and the managers were not allowed to blow off the system. We ran this goal system for ten years from the founding of Cypress Semiconductor Corporation in 1982.
DC- What caused you to change an approach that had worked for ten years?
TJR- As we grew, we hired employees more quickly than we could assimilate them into our culture. Assimilating people into the Cypress culture takes about a year. When we were growing very rapidly, we sometimes hired 500 people in a year, and we had 500 people suddenly getting goals jammed at them. They would wrongly say, "Oh no this is micromanagement. I am in the army. I am getting orders now. Look at this list. I have a list of four things to do this week." That is an unfair and unreasonable criticism from a person who doesn’t understand the system.
However, even though those criticisms may have been unfair, an understanding of the system is not instantaneously achieved, particularly if you bring in somebody from some touchy feely culture. I started getting a lot of negative feedback about the goal setting system, which led me to reconsider if it was the right way to go. The question became, was the inherent value worth the anxiety it was causing?
A second issue was that the functionality of the goal system was diminishing as the company got bigger. In the early days of our rigorous goal setting system, Cypress had perhaps one hundred people. Once you get several thousand people, the database requirements and sorting capability required made it increasing difficult to get any meaningful data. For example, if I sort the goals and find that one of my vice president's goals are 23% late and some other vice president goals are 3% late what does that mean and what do I do about it? The ability to give meaning to the data diminishes with scale.
So between the diminishing value and the bitching, I gave up on the system about 10 years ago. I still think the approach is right. I still think it is do-able. However, when I analyzed what I really wanted, it turned out I could put much less rigor into the management and still get most of what I wanted.
We still have some mandatory goals in the company. I expect every employee to have a set of a few key objectives on a quarterly basis. People set goals for themselves in negotiation with their boss. I expect those goals to go into a database and I expect them to be retrievable through the management structure or through the project structure.
For example I can still take an important project, retrieve the goals from that project and look at the big picture of what is being done to see if it makes sense.
Goals are now negotiated with employees and used for employee feedback at the end of each quarter. They provide an objective basis for ranking and rating at the end of the year. One message to employees: tell me what you’re going to do; tell me when you’re going to do it; then do it.
DC- How do you handle the detailed objectives for project management?
TJR- I created a whole secondary system that is a very elaborate detailed project planning system along the lines of Microsoft Project. So we do track the nitty-gritty of our technical projects in a different system.
I have very little push back on that because anybody managing a project understands that unless they have some fairly sophisticated tools they have no chance of achieving their goals.
When I am running projects, and I do always run a couple of projects, I still use the weekly goal system. I negotiate every week what we are going to get done and we have discussions about what happened last week. If you don’t negotiate objectives every single week, you lose track of the project.
The other time I demand goals in detail is when we identify an under performing program. We have an exception system that shows if a group is off track. When that happens I usually get involved. When I'm reviewing projects I want to hear two things: You admit you are off track so I know you know you have problems. Secondly, I want to hear how you are going to get back on track so I can be confident that your plan to recover will work.
When I go into a meeting for under performing projects I usually find that they don’t use the goal system. In these cases, I demand that they go back to basics and use the detailed goal system.
DC- Thanks for sharing your ideas on compensation and goal setting. We'll look at capital allocation in our next conversation.