Recently in one of Justin Roff-Marsh’s SPE Practitioner Tips (which you can subscribe to here), he asserted that salespeople and their managers spend too much time trying to forecast sales revenue.

Most sales groups forecast revenue by adding the value of all opportunities and then adjusting them for risk. To make the adjustment, they assign a probability that each deal in their pipeline will close, multiply the product of the probability and the deal size, and add them all together.

The calculations look something like this:

Credit: Justin Roff-Marsh

This statistical method of forecasting is inherently unreliable because the sales group has combined uncertain risk-adjustments with small sample sizes.

Roff-Marsh identifies the main problem with this method. The risk adjustments are nearly always a guess. They are often manipulated to achieve a forecast that will satisfy sales management.

Even though the sales forecast is unreliable and requires a lot of time to estimate, sales teams are under pressure to produce this number month after month. Why is that?

Executives, financial planners, and production planners all want to see an estimate of future revenue. They naturally ask sales to provide this estimate to them.

There is a further reason that sales teams use this method to forecast revenue. They are chronically starved for good sales opportunities, so they scrutinize every deal for its potential to convert. If they had a bigger pipeline then each single deal would not have to be scrutinized so closely.

The consequence: The whole company needs an accurate revenue  forecast, but sales uses an unreliable method. This approach to sales forecasting leaves the company unable to accurately plan for its production and financial needs.

Roff-Marsh suggests a different approach. His method uses scenario planning instead of statistics.

The Scenario Planning Method for Sales Forecasting

In his approach, Roff-Marsh coaches salespeople to do several things differently:

1. They carefully align each deal with the pipeline stage that reflects objective observations of customer behavior.
2. Instead of building a forecast based on all deals in the pipeline, they ignore early stage opportunities and focus only on late stage opportunities. By restricting themselves to late stage opportunities, they have less need to assign a probability to the deal because these are all high-probability opportunities.
3. The scenario planners review the late stage opportunities and assign them to one of three categories: possible, probable, and highly likely.
4. Next they assign each deal to the month in which it is expected to close and they give each deal a dollar amount.
5. With this knowledge the team can produce three month-by-month scenarios for business that is highly likely, probable, and possible to close.

Visually this can be presented in a chart like Roff-Marsh’s chart below. This chart conveys a lot of information because it’s limited to late-stage deals which have been broken out month-by-month into three scenarios.

Credit: Justin Roff-Marsh

The Conflict Between Sales Forecasters and Scenario Planners

Although some of his clients use the scenario-planning method for forecasting, the Roff-Marsh solution would create conflict in organizations that currently use conventional sales forecasts to plan for future revenue.

While it’s true that those who support conventional sales forecasting and those who advocate for a scenario-planning method would agree on the same goal, the need to plan for future revenue, they would disagree strongly on how to achieve the goal.

The sales forecasters would say that they need to give management a forecast that it can incorporate into their planning systems in order to forecast future revenue. Further, to give management a forecast, they must provide a single number that can be entered into the planning system. The way they arrive at a single number is by calculating the risk-adjusted aggregate  estimate of future revenue from all salespeople.

On the other hand, the scenario planners would say that they must give management a truthful forecast of an unknown future in order to plan for future revenue. Further, to give management a full picture of a truthful forecast, they must produce multiple scenarios of the future.

These two approaches to achieving the same end, the ability to plan for future revenue, are in direct conflict. One approach forecasts revenue based on calculating a single risk-adjusted number and the second forecasts revenue based on a set of scenarios. A company can’t do both simultaneously.

Is there any way to resolve this conflict? To do so, we have to look at the assumptions that each side makes and see if any of them are faulty. If they are, then we might be able to inject an alternative way of working that resolves the conflict.

Questioning Our Assumptions

Let’s look at the assumptions that each group makes and see if any of them can be called into question.

The sales forecasters want to forecast a number that they can incorporate into their planning system. The plans for production, financing and staffing all depend on the revenue forecast. Further, they use the planning system to report to the Board and shareholders.

Sales is the only group that can supply information about customer purchasing plans. However, they could supply this information in many ways. It doesn’t have to be a single forecast number (especially if we suspect the accuracy of the number).

Next we’ll look at the scenario planners. What assumptions do they make?

The scenario planners want to give management a truthful forecast to plan for future revenue because many other plans are dependent on the truth of the forecast. If the forecast is too optimistic or too pessimistic or simply doesn’t convey the range of possible outcomes, financial planners and production planners will make significant errors.

The scenario planners make the assumption that it’s better to illustrate the uncertainty in the forecast rather than hide this uncertainty and possibly mislead planners and management.

They base their forecast only on late stage opportunities and they believe management should see a range of potential scenarios. They also believe that we should increase the size of the pipeline so that sales is no longer chronically starved for opportunities. A bigger pipeline would make it no longer necessary to place so much emphasis on each deal and its probability of closing.

The Resolution

A. Sales needs to be the one that forecasts revenue, but it doesn’t have to be a single number in order for the forecast to be useful.

B. Sales does not have enough information to estimate the probability of deals closing. The only thing they can do with confidence is take the late stage deals and break them out qualitatively into three buckets.

C. Executives and planners need forecast accuracy more than they need a single number. By giving them a range of scenarios, they can combine this knowledge with other information and make their own estimate of the forecast. This method lets sales provide a richer set of useful information to the planners and it does not force sales to make impossible predictions.

The concept of a system in the Theory of Constraints is analogous to a chain with one weakest link. If you pull from both ends of the chain, one link, and one link only, will eventually snap. In the chain, the weakest link that snapped was the constraint.

If you can strengthen the weakest link in a chain, then the chain can accept a greater load. Similarly, if you strengthen the weakest step in a process, then the process as a whole is strengthened and the load it can carry is improved.

What happens when you strengthen one of the links in the chain? Then another link becomes the weakest link. Similarly, if you improve the capacity of the constraint in a process, then the constraint moves to another stage in the process.

Look at the example below of a simple process in a dental laboratory from William Dettmer’s Breaking the Constraints to World-Class Performance. Each stage in the process can produce a certain number of units each day.

Can you see the step in the process that produces the lowest number of units? At 15 units per day, the Porcelain step produces the fewest units. Therefore Step E is the constraint in this simple process. Even though the market demand is 35 units per day, this company can only manufacture 15 units per day. If it accepts more than 15 orders per day, it won’t be able to deliver its orders on time.

What if the dental lab focuses on the constraint and doubles the capacity of Step E to 30 units per day? In that case the constraint would move to Step B which produces 25 units per day. Step B is now the new constraint. The overall production capacity of the process has moved from 15 units per day to 25 units per day.

The market demand remains 35 units per day. At 25 units per day the lab is still producing less than the number of potential orders from customers.

Now what if the lab continues to improve the production capacity at each of the internal constraints to the point where it can produce more than the market demand of 35 units per day? Now where does the constraint lie?

The constraint is no longer in your system; it lies in an external source, the market. About 70% of companies find themselves in the position where they have excess capacity and the market demand is their constraint.

When we think of constraints in a system, the first image that usually pops into our mind is a physical constraint, like a machine that holds up the workflow. But this isn’t the only kind of constraint. And physical constraints are not usually what cause problems in a sales organization. So let’s learn more about the three kinds of constraints.

The three types of constraints are physical, policy and paradigm. All three of these inter-related constraints exist in any system.

Physical constraints

Physical constraints are resources that physically hold the system back from increasing its throughput.

To locate the physical constraint, Lisa Scheinkopf says we need to ask this question: What is the resource, that if we had more of it, we could increase the throughput of the system?

 

 

 

 

This physical constraint can be internal to the system or it can be external.

Where inputs enter your system, external physical constraints could be due to a shortage of raw materials. If you cannot purchase enough of what you need from your vendors, then you have a constraint.

At the other end of your system, where you release your transformed products to customers, you could be constrained by lack of sales. In this case your constraint lies in the market. You have plenty of raw material and plenty of capacity, but not enough customers to consume what you can produce. Most organizations today are constrained by the market.

Internally, you may be limited by a lack of capabilities or a shortage of capacity inside your organization. We likely first think of being constrained by a machine but today it’s more typical that our constraining resource is people or skills—not enough of the right kind of software engineer or the inability to hire enough quality customer service agents.

Organizations are composed of a system of interdependent resources. The system is designed to perform the process that enable the organization to accomplish its goal. Every organization has one primary physical constraint. If you can improve the throughput of this constraint, you improve the throughput of the entire system.

Many organizations have successfully used the five focusing steps to improve the throughput of their physical constraints.

Policy Constraints

Policy constraints are harder to visualize than physical constraints. They are the rules and the metrics that an organization uses to govern how it goes about its business. You could call them managerial constraints.

Policies define the source of your inputs—who you buy from, how much you are willing to pay, and the terms of your agreements.

They also define the direction of the output of your system—which markets you will serve, how you will compete, and your selling model.

Internally, policy constraints include the myriad of work rules inside your organization.

If you ask anyone in a company, they can tell you about many stupid policies in their organization. No one intentionally develops a stupid policy. People work in groups and through their belief systems they develop policies that seem correct at the time. Once policies are in place, they take on a life of their own and people feel obligated to follow them. We believe that these policies will enable us to make decisions and take actions that will produce the results we want for our organization.

Paradigm Constraints

Paradigm constraints are the beliefs, attitudes, and assumptions that inform the policy constraints we develop and follow. These paradigm constraints cause us to see the world in certain ways and prevent us from considering alternatives. In a famous example, companies that ran railroads viewed themselves as being in the railroad business. They couldn’t see an alternative paradigm until Theodore Levitt pointed out that they could view themselves as being in the transportation business.

Process Improvement

The three types of constraints affect each other. Lisa Scheinkopf says that “paradigm constraints cause policy constraints, and policy constraints result in mismanagement or misplaced physical constraints.”

The five forces have been used widely to manage the physical constraint and improve the throughput in their organization. They are relatively easy to improve because they are easier to see.

Policy and paradigm constraints are harder to see and harder to improve. But they also hold the most opportunity for significant improvement in organizations today.

What is your organization trying to accomplish? What is its goal?

In a business, the financial goal will typically be to make more money, now and in the future. Of course, organizations have additional goals:to hire and retain the best people, to be competitive in their industry, and to bring innovative products to market. But operationally, the financial goal will be to make more money, now and in the future.

Lisa Scheinkopf explains that we connect our metrics to our goal by measuring the primary process of the organization—the inputs, the transformation of inputs into outputs, the outputs—in a way that connects to the goal.

A premise: organizations convert money into more money by adding value

The process begins when we buy inputs from vendors. This is money that we put into the system so that we can convert the money we spent into more money.

We convert this money by transforming the inputs and selling the result to customers at a higher price than the cost of our inputs.

This knowledge allows us to define “value added” as the difference between the amount that customers pay us for the outputs of our system and the amount we paid vendors for the inputs.

If we generate more “value added” than the amount we pay for the ongoing operation of the system, then we make a profit.

The Theory of Constraints defines the fundamental financial components of an organization in this way:

• Throughput: The rate at which the organization adds value to the system by generating money through sales.
• Inventory: The money spent on purchases from vendors that the organization will turn into throughput.
• Operating Expense: All of the money that the organization spends to do the work of turning inventory into throughput. This includes everything spent to run the enterprise—wages and salaries, buildings, equipment, insurance and taxes.

The constraint holds organizations back from increasing their throughput

When we look for the constraint in the system, we are searching for the element in the system that is holding us back from increasing the rate of throughput.

The constraint could be anything in the system that limits better performance in relation to the goal. Improvements that reduce the constraint are measured as increases in throughput.

If we talk to people in the different functions of the company, we are likely to be given a long list of possible constraints. Everyone has their perception of the important problems in the company.

But if we change our view to the system as a whole we can see that the organization is composed of an interdependent collection of resources. Throughput flows through this chain at a given rate. That rate is determined by the constraint in the system.

Like a chain that can be no stronger than its weakest link, the throughput in the system can be no faster than what the constraint permits.

Find the constraint, improve its throughput, and you improve the throughput of the entire system.

Imagine I am a new employee at your company and you want to show me around.

You would likely take me on a walking tour, first through reception then to the different departments: sales, customer service, legal, engineering, accounting, and production.

On this tour you would show me the different functions in your company.

Now let’s imagine that I ask people in the different functions about the problems they face each day.

What are the constraints that hold them back from doing their work?

People See Problems Through Their Functional Experience

People in each function will experience and describe a different set of problems. It’s likely that they will conclude that the problems in their function are also the problems that the entire company should be devoting itself to solve.

The receptionist will perceive that the problems lie in people’s unwillingness to answer the phone and return calls.

Sales thinks that the products are overpriced and that lead times to delivery are too long.

Customer service spends too much its time on expediting orders, not supporting buyers.

Purchasing doesn’t have enough lead time to make acquisitions.

Manufacturing is asked to meet impossible goals.

Which Are the Most Important Problems to Fix?

As you hear people in each function describe the problems in the company as they see them, how do you evaluate which problems take priority over others? Which ones are the most important to fix?

Lisa Scheinkopf says that this functional view puts you too close. You can see trees or branches on trees, but you can’t see the forest.

Instead of looking at functions, step back and look at the whole system of the organization.

Scheinkopf quotes her friend, John Covington. When asked how he approached complex problems, he replied “Make the box bigger!”

What did he mean by that comment? He was saying that sometimes we should look at functions in a system and other times we should look at the whole system. Too often we try to solve a systemic problem by looking at one function.

Step Back and Look at the Whole System

But the only way to decide which problem to solve is to find where the constraint is in the system. To do that we have to look at the whole.

We look at the whole because it enables our perspective to change. When we were looking at organizational functions, we saw a jumble of seemingly unrelated activities. But when we step back and look at the whole organization we can see a pattern. We can see a pattern of flow.

Raw materials and inventory flow into the organization. Inside the organization the raw materials and inventory are transformed into products and services. The products and services for customers flow out from the organization.

This output is the means by which your organization accomplishes its purpose.

The rate at which you generate output is the rate at which you accomplish your purpose.

All organizations want to improve the rate of their output because it improves the organization’s ability to accomplish its purpose.

And the way to improve your rate of output is to find the constraint. The constraint is the most important problem to fix.

Recently Justin Roff-Marsh interviewed Doug Voss, one of his clients, about the projects they have done together. Doug is a vice-president and founder at Sayfa Systems in Melbourne, Australia. Sayfa manufactures and distributes height safety and fault management systems for people who work on the roofs of buildings.

Sayfa went through a sharp growth period prior to working with Roff-Marsh. Sayfa’s future growth potential lies less in customer acquisition and more in the innovation of new products for existing customers.

The period of rapid growth revealed that Sayfa didn’t have a scalable business model. Consequently, most of the work with Roff-Marsh has been with new product development, technology, and engineering. They needed better technology to manage the teams and to make the work of the teams more visible.

Doug wanted a “scalable process that could cope with a high velocity of calls and customer service requirements that come through every day.” He wanted to get better control over the scalability of his systems so the company could cope with a high velocity of calls to sales and customer service.

At the solution design workshop with Roff-Marsh they mapped customer service and inside sales. Field sales activities were reduced from “small and spasmodic” to none.

Technology Migration for the Project Sales Team

Sayfa’s project sales team mostly does specification work. The inside sales team subscribes to a feed of specifications for new building projects in Australia. Inside Sales then chases specifications through the project workflow as projects move from architect to builder to installer.

They had been managing their projects in a spreadsheet, one project per row. But the spreadsheets became unwieldy—“teetering, freezing, crashing.” So they worked with Roff-Marsh to move the projects into a CRM application. The CRM application made it possible to associate one project with multiple stakeholders (architects, builders, and installers).

Everyone in project sales uses the CRM. This includes anyone from specification through tender through installation.

Improved Processes in Customer Service

Once Sayfa had moved project sales to the CRM application, they began a project to improve the process for customer service.

Many high demand requests come into the customer service team.  They found it difficult to manage the workloads of the team. Customer service staff struggled to find someone who would take a difficult call when it arrived.

Roff-Marsh helped them to manage the flow and velocity of work through the team.

They made two distinctions in the type of tasks the team performed. The first distinction was between product requests and technical requests. The second was between short lead-time calls and long lead-time calls. They divided the work among team members who specialized in different types of calls.

They also began to measure on-time case completion. When they started using this metric, they completed 60% of their cases on-time. Now they are at 90%. They hung a board on the wall so everyone could see the on-time case completion rate.

When everyone can see the results, Doug explains that  “you get a bit of banter among the team members.” This talk amongst the team causes people to look for ways to help each other to complete their cases and keep the on-time completion rate high.

When we can see the on-time case completion number coming down, we know we need to get some extra resources in or have a quick meeting to understand what is happening. What do we need to do to sort out or jump in and help the team. The team is aware of it, they haven’t become blind to it.

Identify and Exploit the Constraints, One at a Time

At their strategy session, Roff-Marsh and the Sayva team identified two possible constraints, project sales and customer service. They agreed that project sales was the most important constraint to focus on first.

Project sales faced problems with its ability to deliver full value to customers within a committed time-frame. However, this inability was not due to input constraints (they had plenty of new project specifications that they needed to respond to) nor was it due to a critical resource. The cause of the constraint was due to their sales production and planning practices.

In order to exploit the constraint in project sales, Sayva looked at the causes that might be restricting their output and how they could improve sales productivity. They identified problems with their technology, the “teetering, freezing, crashing” spreadsheets, and migrated to a CRM application.

Once they had addressed the constraint in project sales, the next constraint was in customer service. Once again, the cause of the constraint lay mainly with production and planning processes. Since the CRM application was in place by this time, they elevated the constraint mainly by examining the flow of calls into the customer service team. Better routing of calls and the measurement of on-time call completion made it possible to dramatically improve their quality and throughput.

Doug Voss at Sayva knew they had an opportunity to grow through innovation in new products. He also knew that he couldn’t pursue this opportunity until he had built a scalable machine in project sales and customer service. By identifying the constraints (in sales and service) and maximizing their productivity, Sayva was able to improve quality and throughput in both these functions. This improvement now makes it possible to pursue their growth opportunities in new product development.