DC- What do you think of the balanced scorecard described by Kaplan and Norton?

AS- One concern is that what is done in practice is far different from what Kaplan and Norton have been writing about most recently.

I have visited between ten and twenty companies that have implemented the balanced scorecard and spoken with an additional ten or twenty companies. If you ask anybody what they are using their balanced scorecard for they will reply with the latest claim about the scorecard: they are using it as a tool to link strategy to action. However, when you look at what they are actually doing, they are not doing that.

DC- That’s a wonderful criticism because so much of what people say they are doing is different from the reality. That makes it very difficult for managers to learn from one another.

AS- You are absolutely right. There are a lot of contributing factors to this. One is the western management model that says if I have a lunchtime conversation with somebody and they talk about the balanced scorecard, I now know enough to go ahead and implement it.

So the CEO has lunch with someone, then the orders go out that we need a balanced scorecard. Organizations respond to that kind of thing in a variety of ways. Probably the least satisfactory way is to take something that they have already done, erase its title and write "balanced scorecard" on top.

Another problem is that a lot of companies rely on consultants to help them implement the balanced scorecard. Consultants operate under the reality that there is a certain budget for an assignment and they need to have something deliverable by the deadline. So their focus is more on having something by the time the money runs out, rather than on having something which is really going to be useful to the organization. As a consequence they rush through the process.

The net result is that I have seen seven different actual uses of the balanced scorecard, only one of which is as a tool for linking strategy to action.

DC- Can you take me through those seven uses?

AS- The type one scorecard is a tool that comes out of the world of project management. In it we are tracking the progress of initiatives against plan. It is born from a situation in which an organization lacks accountability. People sign up to do certain things but as the year progresses nobody is really tracking whether or not they are doing those things.

So some organizations use the scorecard to track, the intermediate steps between the beginning and the deliverable so they have measurable indications that people are making progress towards their goals.

DC- And type two?

AS-The second type of scorecard is born from the recognition that for generations we have managed our organizations only by financial measures. We have begun to recognize that those are lagging indicators and if management is to be proactive, not reactive, it needs to manage by leading indicators.

So a scorecard is used to educate people that we can use non-financial measures to manage the business in a proactive way. It is not so important what those non-financial measures are, just that people look beyond the financial. So the selection of measures becomes less critical and often can be delegated down to the people whose process is going to be measured. You say, “I want you to come up with measures that you are going to use from a predictive perspective."

The third type of scorecard is a communication tool. It arises because stated strategies often becomes less and less meaningful, less and less tangible, as you move down through the organization. So this type of scorecard tries to translate the language of strategy to terms that are more meaningful to the people lower in the ranks.

So if you are communicating to in a call center operators that your strategy is to improve customer satisfaction. They will say, “You mean we should answer the phone more quickly?” and you say “Yes that is a manifestation of improving customer satisfaction”.

DC- Type three sounds familiar. HR people have said for years that performance management systems should work by cascading strategy down to specific objectives at each level of the organization.

AS- There is a distinction that is a little hard to articulate. This type of scorecard starts in the world of measurement and finds a subset of measures that link to the strategy. In other words they ask, “Is this measure linked to strategy? Only slightly? Well then we are not going to put it on our type three scorecard." The measures they choose haven't necessarily cascaded down from higher-level objectives.

Let me jump ahead and contrast level three to a level seven scorecard. In level seven you start with the strategy and translate that strategy into the language of the stakeholders of the company. That is something that is rarely done. You need to sort out all of the stakeholders and the meaning of your strategy to them so that you surface any conflicts and make sure that your strategy is not improving the satisfaction of one stakeholder group at the expense of another important stakeholder group. For example, employees want more pay but stockholders want more profit so there is a potential conflict there.

Let's focus on an example, customers as stakeholders—keeping in mind you do the same analysis with each stakeholder group. You first have to identify what market segments you want to compete in. Once you have identified the segment you identify and rank order the requirements of that segment. The next question is how are we doing both absolutely in terms of the perceived needs of that segment of customers as well as relative to competition. That lets us identify gaps in performance from a customer's perspective. It's all very logical.

DC- That’s one standard approach to strategy—a rational analysis.

AS- If you stop there, everyone in the organization nods their head and everybody feels it is somebody else’s job to close the gaps. You need to make a link between the customer perspective and what people have to do. That link is through the process focus, the view that the organization is made up of interacting processes.

DC- HR tends to be weak on the process perspective. They tend to adopt the perspective that the organization is made up of jobs or individuals.

AS- There is not a right and a wrong way of looking at it. It depends on what you are trying to do. If your objective is to come up with a performance measurement system that links to your strategy then you need to take the process view.

You ask, “What is the impact of each process on the customer improvement priorities that you just came up with?” Which processes are the primary drivers associated with the gaps you have identified? Let’s say that your number one gap is in product quality. The question is what are the internal processes that are the drivers of that? You generate a set of process improvements that give the greatest leverage in closing the gaps. Those processes become the ones you measure for the level seven balanced scorecard.

DC- Is level seven the same thing Kaplan talks about as a strategy map?

AS- I think Kaplan's followers often end up doing a type three scorecard not a type seven. If you start from your strategy, if you identify what important performance gaps exist, and what the processes are that drive closing those gaps then you can generate performance measures. The performance measures in all likelihood will not be a balanced set. It will not fit in with the four perspectives defined by Kaplan because you have not forced it that way. It may not have any measures associated with learning and growth because learning and growth may not be strategically leveraged in the scorecard’s time horizon. From a strategic perspective your organization may have done an adequate job in learning and growing. If you don’t have to improve them, don’t clutter a balanced scorecard with them.

DC- You are describing a way of focusing the organization to close certain strategic gaps and presumably once those gaps are closed you move on to different things. However, the model that Kaplan talks about is a little bit more timeless. The strategy maps he creates are the expression of what matters to carry out an on-going strategy rather than closing strategic gaps.

AS- Let’s return to the type three scorecard which starts with measures, of which there are an infinite number, and says which of these measures link up to strategy? My claim is that anything that you could conceive of measuring has some link to strategy. If you just use subjective judgment to determine how important that link is you may be fooling yourself. From a mathematical point of view that is called backwards chaining. What I describe as a type seven scorecard is forward chaining. Start with strategy and work your way down to a set of measures rather than starting with an infinite number of measures and trying to work your way back to strategies. There are many, many paths that lead from the infinite set of measures back to strategy but not all of those paths have the same impact.

When I look at people who have done Kaplan inspired balanced scorecards they often start off with the four perspectives. Then it goes to objectives for each perspective, then measures and then initiatives. That is the logical path. The problem is that in practice the perspectives and the objectives may be done strategically, top down, but there is a disconnect and the initiatives, measures and goals are bottom up. The bottom up people take whatever their current initiatives are and they come up with a set of measures for them and then they give an argument as to why they link to strategy.

DC- In terms of what actually goes on in organizations that rings true.

AS- If you ask people who have just done a strategy map how their initiatives have changed as a consequence, nine times out of ten they will tell you they have the same initiatives as before.

The reality is that organizations are poor at choosing the right initiatives. There is an immense amount of waste. The only thing saving companies from the competition is that everybody is making the same mistakes. Everybody reads the same literature and they know that now we have to have data warehouses because everybody has data warehouses. Nobody asks the question, "How are we going to use data warehouses to improve our competitive position?"

DC- Let's move on to the other types of scorecard, types four, five and six.

AS- A type four scorecard is used for control purposes. A control measure has a goal that is equal to the historic average value. For example, if historically we averaged 95% on time delivery we may conclude that that is good enough. We want to maintain 95% so we put in on our scorecard and measure it just to make sure it stays there.

Control measures are very important; in fact every process has to have a set of control measures. This is the whole world of statistical quality control and statistical process control.

There are a very large number of control measures that exist in any organization. Control measures don’t belong on a scorecard, yet you often find that 80% of the scorecard measures are control measures. The reason is that people are forcing measures onto the scorecard to match the four perspectives. A type four scorecard that is used for control purposes is not a balanced scorecard --- it is a process control report.

A type five scorecard exists to introduce balance. The best example I've seen is in a hospital in San Diego. They have measures of cost, of quality and of responsiveness. They did this because in the healthcare industry they tend to swing on a pendulum where one year everybody focuses their attention on costs and nobody is paying attention to quality or responsiveness to patients. Then all of a sudden they find patient satisfaction declining and people swing all their attention to responsiveness. The purpose of a type five balanced scorecard is to highlight potential trade offs to stop the pendulum swinging wildly from one focus to the next.

A type six scorecard is as a deployment tool. The best example of a type six scorecard is a methodology called Hoshin Kanri. It is a management approach developed in Japan in the 1970s. It starts with the CEO identifying one or two clear breakthroughs that he views as important, for example they need to go from 5% revenue growth to 15%.

They need to rally their forces around these special initiatives. So the CEO asks each subordinate how the company can grow 15% and what they are going to do to make it happen. They in turn ask each of their subordinates what they are going to do to help me achieve the target. For example, if the Vice President of Sales has an objective of 15% sales growth he or she will go to all the district sales managers and say, “How are you going to help me achieve 15% growth this year?”

It's not just cascading goals down the organization, that is management by objectives, not Hoshin Kanri. As it gets cascaded down through the organization, not only do the goals get deployed but means get identified for the achievement of those goals. It then bounces all the way back up to the CEO and, for example, the Vice President of Sales will say, "I can meet that 15% goal if you allow me to hire 100 more sales people and I'll need support from the Vice President of Technology in marketing to this particular customer." Hoshin Kanri demands all of the requirements in all the gory detail of more than just the goal.

In the end the work plan of each sales engineer gets changed. For example, he or she will cut back on visits to traditional customers and spend that freed up time visiting customers in this new market segment.

The process bounces up and down over several months as this very detailed planning process is carried out in order to make a good match between the goals and the means. They call this “catchball.”

DC- What is the main difference between a type six and seven scorecard?

AS- With a type six scorecard, the CEO may or may have not linked his objective to strategy. In a type seven scorecard you very meticulously link those objectives to strategy. A type six scorecard is really used for deployment and alignment of activities.

DC- The seven types present a useful typology of what kinds of scorecards exist in practice.

AS- It is important to distinguish the various uses of a balanced scorecard because unless you settle on the use, the balanced scorecard for you can’t figure out a process for getting you there. You need to decide what is it that you are trying to accomplish. Are you trying to bring accountability to your organization? Then you want a type one scorecard. Don’t say that you want to link strategy to action.

DC- I suspect that most people, and that includes me, thought there was really only one type of balanced scorecard and hence, if they set out to build one are bound to get to the right destination.

AS- People don’t realize that there is a very complicated process that you need to use to ensure that what you produce is a level seven scorecard. If you decide that is too complicated a process, if you feel you already know what you need to do and just need to lay it out in the framework of the balanced scorecard, you are likely to end up with a type three scorecard. You will have the measures on a scorecard that will link to strategy but they will not be the right measures.

DC- Do you have any closing comments?

AS- The last issue is why balanced scorecards fail, or to put it more positively, what are the requirements for successful implementation of the balanced scorecard.

There are six cultural requirements, which apply to any change initiative, and an additional six requirements specific to the balanced scorecard.

AS- It certainly is true in the West; patience is in short supply.

From 1986 to 1993, Art was Vice President of Quality and Productivity Improvement at Analog Devices, Inc. He was responsible for planning, facilitating, and supporting Analog's worldwide implementation of Total Quality Management as well as its performance measurement and balanced scorecard systems.

Art was a Senior Examiner for the Malcolm Baldrige National Quality Award and served on the Conference Board's US Quality Council II. He is a visiting fellow at the Centre for Business Performance at Cranfield University's School of Management (UK) and a tutor in the University of Limerick's Master of Quality Programme (Ireland).

He has authored many articles concerning quality goal setting (the half-life concept), quality costs, performance measures, and the balanced scorecard. He is a member of the Editorial Advisory Board of the Journal of Cost Management.

Art is a graduate of MIT with a BS and MS in Mechanical Engineering and an MS in Management from MIT's Sloan School of Management.