The learning curve theory is a way to understand the improved performance of an employee or investment over time. The idea is that the more an employee does something, the better they will get at it, which translates to lower cost and higher output in the long term. It's a useful model for tracking progress, improving productivity and ensuring your company is hitting certain performance targets. In this article, we explain what the learning curve theory is, how it works and how you can apply it to improve your company's performance.
What is a learning curve?
A learning curve is a visual representation of the change in production efficiency over time. The basic theory behind the concept of a learning curve is that there will usually be an initial period in which the amount invested will be greater than the return, but after the learning curve has been overcome, the return should be greater than the investment.
For example, imagine you've just started a new job. At first, the company will need to spend time training you, and you'll need some time to get used to the new tasks you're responsible for. However, after this initial period of high cost and low output, you'll have the experience necessary to perform the same tasks faster and better.
Requirements for learning curve models
To develop an accurate learning curve model, you need to make sure you have all the variables that go into the formula. The data required to plot a learning curve and then interpret what it means for your company's performance includes the following:
- A measurable unit of output
- A defined unit of cost (in time, dollars, effort, etc.)
- Either a fixed time frame or a fixed productivity target
Why are learning curve models important?
Learning curve models are useful because they help you understand whether the processes and employees at your company are performing well enough to match the resources put into improving them. The shape of the learning curve can tell you whether performance is improving, declining, stagnating or fluctuating. This can then bring awareness to the potential influences on this performance so you can make decisions that help amplify the positive influences on your company and mitigate the negative influences.
The advantages and disadvantages of the learning curve model
The learning curve model is very useful for monitoring various aspects of company performance and identifying areas in need of improvement. It can provide a lot of great insights, but there are some limitations. In this section, we examine some of the key advantages and disadvantages of the learning curve model.
Advantages of the learning curve model
Using a learning curve model to track the progress of various aspects of your company can help you in the following tasks:
- Strategic planning to improve the output of employees or even whole departments
- Motivating company staff by creating a culture of ongoing learning and progress-tracking
- Identifying trends that can be used for more accurate forecasting and better business decisions
Disadvantages of the learning curve model
As valuable as it can be to a company, there are some important limitations to understand so the learning curve isn't misinterpreted or misused. Here are some of the key disadvantages and limitations:
- Learning progress is influenced by a number of variables, including time, previous experience, quality of training and so on. As a result, tracking only one of these variables might give you misleading data.
- Some performance or progress is difficult to quantify and measure. If there is no specific deliverable, such as a product or a sale, it can be hard to define a single unit of output for the purpose of measuring progress.
- It's incomplete on its own. Because there are so many variables that can impact performance, it's important that you use the learning curve model in combination with other methods of tracking company performance for a more complete picture.
How to use the learning curve theory to improve performance
The learning curve theory—which states that the more time spent doing something equates to a more efficient output—can be used to improve the performance of your company as a whole as well as specific departments and even individual employees. Here's a quick overview of how to use it for that purpose:
- Establish clear and precise terms. The learning curve theory only works if your data is defined and consistent.
- Measure multiple variables at the same time. For the most complete picture of how your company is performing, you want to apply the learning curve model to as many areas as possible and see how they compare and influence each other.
- Double-check your data. Before you make a decision based on the results, make sure you haven't overlooked any variables. This is especially true if the results are not what you expected to see. If it's unexpected, it may be because you're missing a key piece of data.
- Make informed decisions. If the data shows that your current training process is not very effective, for example, make some modifications and measure it again. In some cases, the best decision to make will be immediately clear. In other cases, it may take a few rounds of trial and error to find the change that actually improves performance.
- Continue monitoring. Don't just apply the learning curve theory during times of change or difficulties in your company. Monitor year-round. Continuous monitoring will alert you to problems as soon as they appear so you can correct and modify your approach quickly.
How to use the learning curve formula
The learning curve, expressed as an algebraic formula, is as follows:
Y = AX^B
The variable Y is the average time per unit of output. A is the time it took to complete the task the first time. X is the total number of attempts or units of output. B represents the slope of the function.
Simply put, the equation helps you understand the average cost, usually in time, of reaching maximum efficiency. The more attempts that are included in the formula, the lower the average time will be.
To use this formula, you'll need to track the time and output data for whatever you're trying to measure. Follow these steps to do so:
- Define your unit of output. If you are measuring a new employee's efficiency, this might be a certain deliverable. If there is a shortage of regular or useful deliverables for the purpose of measuring output, you will have to get creative about finding a way to quantify the employee's output.
- Define your cost. As mentioned above, in most applications, your cost is measured in time. In some cases, however, you might be measuring actual financial cost, such as the cost of a reorganization of a department's structure or the renovation of a new storefront.
- Establish your time frame. Depending on what kind of information you're looking for, you will need to establish a time frame. For example, if you're implementing a new training method, you might set your time frame as the average time it took to reach maximum efficiency with your older method. If the new method has achieved better results in the same amount of time, you know that it's an effective change.
- Establish your target. Other applications might be more open-ended in regard to time frame. Rather than trying to see what level of productivity you can achieve within a certain period, you might be trying to find out how long it takes to reach an ideal level of productivity. In this case, leave the time frame open-ended and then establish a specific, measurable target you're trying to achieve.
- Begin measuring. For the most accurate data, you will ideally measure the very first attempt, which will likely have the highest cost per unit of output. Having this first attempt measured will give you the most accurate depiction of progress and total cost.
- Keep your data organized. You can use a spreadsheet program to model your data as you go. However you choose to do it, make sure you're recording measurements in a single place and in a clear and consistent manner so it will be easy to translate those results into a visual representation.
Four examples of the learning curve model
There are four main types of learning curves you'll see when you begin to model your data. These are distinguished by the path of progress for whatever it is you're measuring. Below are some examples of each type and how they can impact company decision-making:
- The diminishing returns learning curve
- The increasing returns learning curve
- The S-curve
- The complex learning curve
The diminishing returns learning curve
This shape, which looks like a steep, downward slope quickly followed by a plateau or straight line, means that learning time is relatively fast and you are quickly achieving maximum efficiency. In terms of decision making, this result means that this aspect of the company is performing well but that you need to make sure you are keeping costs down after the plateau begins.
Another thing to look for here is where the plateau happens. The lower down the plateau occurs (toward the X-axis), the more efficient the performance is. If the plateau begins in the upper half of the graph, this indicates that the task you're measuring has a consistently high cost and you might want to look for ways to lower that cost.
The increasing returns learning curve
The slope in this type of learning curve will be shallow at first, followed by a steep drop off. This suggests that the task takes a long time to learn, but once learned, employees can quickly reach a point of high efficiency.
In terms of decision-making, you may not need to do anything here, as the initial cost of slow learning is quickly returned once the high-efficiency phase is reached. However, if the rest of the data suggests this task shouldn't take very long to learn, this shape could indicate your company may need to modify its training method.
The S-shape of this learning curve suggests that learning is slow at first, followed by a steep drop in cost per unit of output. Then, after the steep drop, there is a plateau effect.
Like the diminishing returns learning curve, if this plateau is happening close to the X-axis, this represents a highly efficient performance. If, however, the plateau is happening in the top half of the graph, performance may not be as efficient as it could be. In this case, there is cause to take a closer look both at the training method and at any variables that could be impacting the cost of ongoing performance in the plateau phase.
The complex learning curve
A complex learning curve is one that you will most often see for the more difficult tasks in your company. It begins with a shallow learning curve, followed by a plateau, followed by a drop off that might be steep or shallow.
This suggests that the task being measured is difficult to learn, and even after learning, it takes a certain amount of practice before an employee can really perform the task efficiently. For decision-making, a company might look for ways to improve training, or it might choose to reduce costs by hiring candidates who already have the experience required to reach peak efficiency for this task.