David Parmenter was recently interviewed by David Creelman, a Chief Correspondent, at The Human Capital Institute. Below is a copy of the interview.
DC: There is a lot of interest in KPIs. How do we get a handle on using them?
DP: The first thing is to get an understanding of what KPIs really are. Kaplan and Norton’s work on the balanced scorecard—which I think was the most important piece of management work that came out in the ’90s—never defined what a KPI was. This has led to a lot of confusion as people measured anything they felt like and then branded it a KPI.
When I go around the world I find people who have monthly and quarterly KPIs, but I define a KPI as being something you measure 24-7. How can something be key to your business if you just happen to glance at it once a quarter? Hasn’t the horse well and truly bolted? When you ask people, “How useful have these quarterly measures been in transforming your business?” they all look pretty blank.
Another aspect of a set of true KPIs is that they must be few in number, probably less than ten. We’ve got famous cases of companies that manage to boil down to one or two KPIs, whereas many companies today have 40 or 50 so-called KPIs—and if you look at the health sector, you can even have them playing around with 300.
DC: What else characterizes a good KPI?
DP: One very bright woman once pointed out that my examples were all non financial It was like a bolt of lightening had struck me: a KPI can never be a dollar measure.
Once you measure an activity by putting a dollar sign on it, you know you aren’t delving deep enough. A classic example is sales. What we really need to find out is what makes a sale. It could be maximizing the time of key sales people in front of the customer. If that were the case, the CEO would want to know—on a daily basis—when the next point of contact is, because it’s that contact that is driving revenue, and you cannot let it slip.
Let’s look at a classic measure: customer satisfaction. What is happening to customer satisfaction? Is it going up or coming down? If it’s going down, then obviously the board should be concerned. However, you can never ring anybody up in the business and say, “Can you please increase customer satisfaction today?” They wouldn’t know where to start.
Customer satisfaction is an interesting measure, but it could never be called a KPI. You have to delve into customer satisfaction to find out what activities are making the customers happy. As you dig down, you will find that some of those activities should become the basis for KPIs.
DC: What are typical KPIs?
DP: Let’s just take one step back. To do performance measurement well you need to understand your critical success factors. This is a step that very seldom takes place.
When an organization first makes a list of critical successful factors (CSFs) they may come up with 20-30 of them. I would call these “success factors” as it is only when you have whittled them down to between five to eight that you can call them CSFs.
A critical success factor may be increased repeat business with key customers. If that’s the case, then we can begin to think in terms of actions specific individuals might take to have an impact on that factor. Think of somebody in dispatch. Now, we’ve never told them that a particular customer is really key to us and that it’s very important they treat an order going to that key customer differently than to any other customer. In other words, we would say to them, “Please double check this order, be sure of the quantity, be sure of the quality—don’t ever get an order to this customer wrong.”
DC: That’s a great example.
DP: As I’ve said, one of the main distinguishing characteristics of a true KPI is that it’s monitored on a daily basis. It should be monitored by the senior management team, and to me, a measure hasn’t really arrived as a KPI unless the CEO is spending at least an hour a day working with it—and making phone calls to staff about it.
The phone calls the CEO makes about the KPIs are essential because KPIs are the only things in the organization that actually link the day-to-day operations of the workplace with the strategic objectives. The CEO is going to be ringing a front-line supervisor asking, “Why did we send a shipment out to ABC Limited, who is a key customer, with four defective goods in it?”
Let me give you another example: you have an organization that is shipping freight. A critical success factor would be optimization of the fleet on a day-by-day basis. So the CEO should get a summary every morning of what trucks went out underweight. (If you have a truck that can carry 40 ton, but you only have 10 ton on it, it would be significantly underweight.) The CEO would ring up the dispatch person to find out why this happened; he or she would ask “Who were you sending it to? Did you phone up the customer to see if you could change the delivery day? Could it have gone the day before or the day after on another trailer unit going that way?” Inevitably, the dispatcher will say, no, he didn’t ask, he was just delivering it to the customer. The CEO would explain that this is costing the company a huge amount of money, and that quite frankly, optimizing the fleet is the most important thing the dispatcher can do. Perhaps the CEO will even say, “I’ll give you the authority to subcontract out that 10 tons rather than send that 40-ton unit of ours on that kind of journey.”
The CEO doesn’t have to make many calls like that before the grapevine works for him. Behavior starts changing. Hopefully the CEO will balance the critical phone calls with congratulatory ones: “By the way, I’ve noticed that in the last 45 days you haven’t sent one unit out underweight. I’d like to come down and take your team out for lunch.”
DC: Take me into the domain of people management. If I believe that people issues are central to our success, what are the KPIs around that?
DP: In most organizations the retention of key staff is a critical indicator, although it is probably not in the top five. There are some manufacturing plants where only a few people are expert in a particular field, and if they aren’t around the machine can’t be run at all. So we’ve got to be monitoring the training that’s been going on to develop these rare, very skilled operators. We want to constantly generate a skilled workforce in this area so we are not caught short if someone leaves.
There are a lot of performance measures when it comes to people. The one I’d like to see in a CEO’s office is the number of recognitions the CEO has given out this week. If you tracked this you would find some CEOs only give out three recognitions in six months. A CEO like Richard Branson would be very high on any measure of recognition. It’s his natural way. It’s the reason people gravitate to him. But if it doesn’t come naturally to you, then measure it. I know a CEO who has something called ‘the success express,’ a weekly newsletter where they talk about the success stories in their business. We are not talking about employee of the month here, but an on-going recognition for performance, without a set limit.
Would a measure of recognition be a KPI? Certainly if the organization has poor recognition—and I’ve worked in some very sick organizations—it could be the measure that changes the culture significantly. Remember it would need to be measured at least weekly to be a KPI.
DC: Give me another possible KPI from the people management domain.
DP: I want to be careful about this because a measure that’s a true KPI for one organization might not be for another. But if we go to the more general concept of a performance indicator, then one of the important ones in HR is the ability to fill supervisor positions internally. There are huge costs in having to go outside to fill these roles. Certainly, it is nice to have new blood in occasionally, but the most successful organizations have managed to have well over two-thirds of supervisor positions filled from internal resources. Unlike a KPI, you may not measure that daily. It is something to discuss on a monthly or bi-monthly basis with management.
DC: What else might be an important measure in the human capital area?
DP: We’ve talked about recognition. Let’s talk a little about retention of staff. Why do staff want to stay? One reason is that senior management delivers on what they promise. There’s nothing worse than being told that something is going to happen and it never does.
What we promise is seldom measured. Didn’t we promise that Mary would be given this responsibility in three months? Did we do it? We should be measuring the delivery of what we promised staff.
It’s very important to deliver on promises to staff in their first three to nine months on the job. A lot of perceptions are set in that period. We’ve got to make sure that we get staff quickly on the right track and that they do not start off resentful.
What I do is look around the critical success factors in each organization and drill down from there. In aviation, planes running on time is a critical success factor and the key performance indicator may be late planes in the air over 2 hours late. If a plane is an hour or two late, an organization needs to chase down the KPI and figure out why it happened, and what needs to be done to get it back on time. That measure has an HR component. What we want the staff to do is take ownership of the problem. That’s a very important thing in human capital; staff need to be trained not to dismiss responsibility with a “this isn’t my problem” or “it’s not my fault” attitude. Most KPIs have a very important behavioral component to get people doing something to fix it.
DC: Should rewards be tied to KPIs?
DP: Reward is something most organizations get terribly wrong. I’ve seen huge payouts based on reaching a pre-determined target. But we don’t have a clue whether these targets are too hard or too soft. Particularly in growing economies we can be paying out huge bonuses to staff who have been, frankly, playing golf.
Jeremy Hope of Beyond Budgeting (http://www.bbrt.org/) fame, calls this approach to pay “the annual performance contract.” His groundbreaking work is trying to rip it out of companies. What I’m talking about is throwing out the annual planning process, because it’s destroying so much wealth. You cannot accurately set a target for a staff member because we can never know what the future is going to be like. We end up in endless debates where the staff member wants to set a soft target, management is trying to set a stretch target but no one knows what’s soft and what is a stretch in reality.
Jeremy says we need relative performance measures. What that means is that we measure you against reasonable comparators; for example, your sales peers, not a pre-determined target. We might measure you against how our competition has performed. If your sales go down, but you’ve done better than the competition and managed to increase the market share, then you have done a good job and your rewards should reflect that.
It is a massive change, and I know people hearing this will be thinking this will never happen in my lifetime—but it has to. When you look at the wealth that we are dishing out so inappropriately, you know it needs to change.
For more information access David Parmenter’s white papers that contain his latest thinking along with supporting electronic media to get you started.