The 5 Rules of Maximizing Performance

Over the last 30 years I`ve been and garage entrepreneur, a deal-making CEO, and a highly paid performance management consultant to some of the largest companies in the world, pursuing the answer to one fundamental question: “How do you really maximize the performance of a business?”

About seven years ago I began a close collaboration with Mahmut Karayel, a Berkeley Ph.D. in Operations Research. Working together, we’ve developed and refined the answer to this question and demonstrated it in business after business.

We have turned our approach into a management technology platform that helps management teams deliver consistent performance improvement, quarter on quarter, in every aspect of the business where sufficient data exists. But in the process we learned that this journey requires a different appreciation of the business management problem that is able to question and reject even deeply-ingrained ideas, like KPIs and activity-based costing. We don’t presume to have all the answers but we have proven over and over again that these five rules put companies on the path of maximum enterprise performance. Clients who have taken our advice seem to think that we are right.

Rule 1

Treat Enterprise Performance as a Production Problem

Align Resources with Goals
If you survived in business for some time you must be a goal oriented person. You are burning resources (capital, energy, time, or brain cells) to achieve a goal. Pick the right goal; make the resources last.

In an enterprise, you invest inputs (cash, personnel, trucks, vendors, distribution resources, square feet, customer time, etc.) and seek to maximize outputs (revenue, profit, market share, customer count, etc.) If you sense you are faced with a production optimization problem, you’re absolutely correct. You are searching for the best mix of products for maximum profit; you strive to meet market demand with most efficient use of man and machine capacity, you are trying to squeeze a dozen products into the same assembly line; you are optimizing stock to simultaneously minimize working capital and lost customers. Or maybe you are juggling project and service priorities to keep customer satisfaction at the highest level. Regardless, as a manager you are challenged to make the most out of what you’ve got.

In automotive manufacturing, you can confidently stock one steering wheel per car assembly. By contrast, it’s not possible to confidently allocate four minutes to serve a bank customer or fifty dollars per day for sales expenses or ten cents of advertising per ticket sold. So what is different? In today’s world the challenge is to plan and optimize considering that both on the producing and selling and buying end, you are dealing with decision makers with different tastes and preferences and problems. In the modern enterprise, the bill of materials recipes may have to be flexible when applied across the wide range of business units and geographies. But there’s a much more important problem: how do you arrive at the right numbers – the optimal numbers – in the first place? With a fixed recipe production system like a car, you can refer to a blueprint to understand the number of screws, washers, door panels, etc. required to build a car. Efficiency and quality are defined as matching a pre-existing specification as closely as possible. With most service processes, there is no “perfect specification.” Rather, you derive service policies from adaptive pattern recognition and continuous optimization. You are chasing continuous improvement, not a fixed ideal.

How do we recognize improvement possibilities and continuously strive for them? The answer to optimal performance does not live in a manager’s head or in a clever spreadsheet model. It lives in the actual performance data of your organization itself. Revealing current performance levels across all business units, geographies and economic conditions is the key benefit of a good Business Intelligence knowledgebase. With this data, you can start to compare your average performers and laggards to the best performers in the organization, and understand what optimal resource allocation looks like. Then you can work to continually improve, to do more with less, and drive the performance of each unit toward its optimal potential, rather than towards the average performance embedded in most KPI-based plans.

This is the power of enterprise performance optimization.

Rule 2

Eliminate KPIs

KPI is not a Religion
Avoid making formula based KPIs the company religion. Formula KPI’s are prone to manipulation and often used beyond their expiration date. Real KPI’s are very simple or behavioral. “Insanely great!” is not a KPI, yet it fires people into doing the right thing.

In an enterprise, you invest inputs (cash, personnel, trucks, vendors, distribution resources, square feet, customer time, etc.) and seek to maximize outputs (revenue, profit, market share, customer count, etc.) If you sense you are faced with a production optimization problem, you’re absolutely correct. You are searching for the best mix of products for maximum profit; you strive to meet market demand with most efficient use of man and machine capacity, you are trying to squeeze a dozen products into the same assembly line; you are optimizing stock to simultaneously minimize working capital and lost customers. Or maybe you are juggling project and service priorities to keep customer satisfaction at the highest level. Regardless, as a manager you are challenged to make the most out of what you’ve got.In automotive manufacturing, you can confidently stock one steering wheel per car assembly. By contrast, it’s not possible to confidently allocate four minutes to serve a bank customer or fifty dollars per day for sales expenses or ten cents of advertising per ticket sold. So what is different? In today’s world the challenge is to plan and optimize considering that both on the producing and selling and buying end, you are dealing with decision makers with different tastes and preferences and problems. In the modern enterprise, the bill of materials recipes may have to be flexible when applied across the wide range of business units and geographies. But there’s a much more important problem: How do you arrive at the right numbers – the optimal numbers – in the first place? With a fixed recipe production system like a car, you can refer to a blueprint to understand the number of screws, washers, door panels, etc. required to build a car. Efficiency and quality are defined as matching a pre-existing specification as closely as possible. With most service processes, there is no “perfect specification.” Rather, you derive service policies from adaptive pattern recognition and continuous optimization. You are chasing continuous improvement, not a fixed ideal.How do we recognize improvement possibilities and continuously strive for them? The answer to optimal performance does not live in a manager’s head or in a clever spreadsheet model. It lives in the actual performance data of your organization itself. Revealing current performance levels across all business units, geographies and economic conditions is the key benefit of a good Business Intelligence knowledgebase. With this data, you can start to compare your average performers and laggards to the best performers in the organization, and understand what optimal resource allocation looks like. Then you can work to continually improve, to do more with less, and drive the performance of each unit toward its optimal potential, rather than towards the average performance embedded in most KPI-based plans.This is the power of enterprise performance optimization.

 

Rule 3

Create an Internal Market for Resources

No Silos
Eliminating silos means more collaboration and less energy spent on turf wars. Discourage empire building. Encourage sharing of ideas, resources and contacts. Allow managers to sell their vision to the resource in order to get more of his/her time.

Using common resources efficiently requires competition. Otherwise you get all of the problems associated with improper pricing: hoarding, spendthrift behavior, and shortages.Consider that a large automotive company will buy transportation services that exceed the revenues of a medium sized US company. The marketing spend of a US beverage company is larger than revenues of most manufacturing companies in Europe. Yet all of these resources are actually distributed, allocated, spent, and invested throughout the enterprise in a quite haphazard way.It is common practice for allocation to happen either with a few people behind a desk pushing the buttons on a spreadsheet, or through a process of competitive rain dances in front of the executive board. Clearly, neither approach is a substitute for a pricing mechanism.This type of resource allocation is the birthplace of mediocre enterprise performance.Now I can hear voices asking:

“So what do you want to do? Have all our stores bid for fancy packaging supplies before the shopping season?”

“You want our sales reps bidding for transport services every week, to deliver their orders in time?”

“Should individual regions bid for air time or marketing inserts?”

In a word, yes.

But I am not suggesting that we conduct call ins, Sotheby’s style. Rather, we should leverage our massive investments in information technology to develop a system that is more akin to eBay or Google Ads than it is to the Resource Rain Dance. We have been up to our ears in Business Intelligence since 1995. It’s time to use it to plan intelligently.

The Alta Bering EPO platform provides an easy methodology for optimal resource allocation, whatever the changing resources and whatever the changing needs of the organization may be.

 

Rule 4

Cure the Flaw of Averages

Averages are flawed
Do not drive performance toward an average. Investigate the maximum potential instead. Leverage outliers and use them to discover new possibilities. Our modern world is volatile, involving hazards, options, and discontinuities. These phenomena render averages as partial truths, or worse yet, misleading.

Measuring performance with the purpose of improvement and target setting is a time-consuming activity of the modern firm. Estimating performance requires selection of critical factors that are ingredients of important metrics representing resources consumed and values added. Non-parametric approaches such as EPO™ use these factors directly whereas parametric approaches assume a priori functional form. Parametric approaches such as regression may be appropriate when observing the behavior and characteristics of a population without the possibility of intervention. Outcomes are not classified as desirable or undesirable, nor do we contemplate changing the outcomes. In a regression model, the errors (deviations from average) are assumed to be random. However, as our objective is to measure and improve performance, we need to be searching deviations that are due to inefficiency. Typically these deviations are not due to random exogenous factors and good outcomes can be replicated or imitated, improving the performance of the overall population.A parametric regression model requires re-estimation of parameters when new factors are introduced in the environment. This is time consuming re-work. The EPO™ platform handles changing environments non-parametrically and incrementally.In regression, parameters are estimated by the average. If all members of the population tend to the average, estimation is said to be more accurate. Yet, in our competitive economy, “Average resources per average output with as little variance as possible” is the axiom of mediocrity and no degree of parametric sophistication can change the domination of the average. On the other hand, EPO™ tries to seek out the desirable outliers and draws a path from non-performance to these “outliers”. In short, regression estimates what is while EPO™ discovers what could be.It is in fact fortunate that the members of a branch network, in a bank or store chain, are diverse. Each branch is unique in its production function given the internal and external resources as well as management resources. However, regression based allocation of targets stifles overall performance, simply because it is too eager to enforce narrow variance bands around some average, when opportunity to learn and innovate actually lies with the “outliers”. When a branch discovers a new business process and starts to perform better, regression estimate of average performance moves only a little. But the business objective should be to guide other branches to adopt this improved business practice. But how? EPO™ shows which factors are below target and by how much. Regression does not give guidance as to how to move towards the better performers.When targets are set with limited recognition of the unique opportunities, performance defined as compliance with these targets becomes counterproductive. There is no incentive or guidance for each manageable unit to do better than its peers. With EPO™ we are able to discover these opportunities and set guidelines. Alta Bering EPO™ accelerates evolution whereas regression simply observes it.

 

Rule 5

Abandon the Idea of Allocating Fixed Costs

There are three dangers to fixating on Fixed Costs:

1) Since they are “fixed”, managers treat them as a given. It discourages them from challenging the assumptions and innovating.

2) Managers can hide behind the Fixed Costs to justify performance shortfalls.

3) When fixed costs are sunk costs, managers sometimes fall into the trap of feeling obligated to use them.

If you don’t like the cost of something, re-calculate it for the shortest possible term.

In business planning and management accounting, usage of the term fixed costs depends on the intended use. This can be confusing and controversial. Some cost accounting practices such as activity-based costing will allocate fixed costs to business activities, in effect treating them as variable costs.

In accounting terminology, fixed costs will broadly include almost all costs (expenses) which are not included in the cost of goods sold, and variable costs are those captured in costs of goods sold. In practice, depending on whether the management is accounting for past performance or planning future performance, definition of “overheads” tends to shift between “short-term” and “long-term”, (last term and next term). It is unreasonable to expect agreement on the definition of “overhead” outside the accounting department.

So what is the sales manager or the regional manager who feels he is over-burdened with his slice of overhead to do? How can this “hit” to profits be explained to the sales force that just made their revenue targets but missed the overall profitability target (not to mention their bonuses)?

The solution to this controversial management issue appears when economic and accounting treatment of fixed cost meet, in other words, in “the long run”.

The more effective alternative is to focus on ways of understanding operational efficiency first, not just costs, regardless of how much has already been invested in ABC and its various versions.

This is yet another issue directly addressed by the Alta Bering EPO platform.