The learning curve theory proposes that a learner’s efficiency in a task improves over time the more the learner performs the task. Sounds simple? Let’s break it down.
The learning curve theory proposes that a learner’s efficiency in a task improves over time the more the learner performs the task. Sounds simple? Let’s break it down.
Understanding learning curves is essential in both personal development and professional growth.
Whether you’re picking up a new skill, adapting to a new role, or navigating the complexities of a project, learning curves provide insight into the time and effort required to achieve proficiency.
In this blog, we’ll explore what learning curves are, the different shapes they can take, how to calculate them, and how L&D teams can put the theory to practical use.
We’ll learn:
- How to actually apply this in your L&D strategy
- What a learning curve is
- The four common types of learning curve
- How the Ebbinghaus forgetting curve connects to it
- How to calculate the learning curve
- Real-world industry learning rates
- Benefits and drawbacks of using the model
💡 Pro Tip
An LMS directly impacts the learning curve by streamlining content delivery, personalising learning experiences, and providing real-time feedback. It helps flatten the learning curve by offering structured courses, adaptive learning paths, and on-demand resources, making complex topics easier to grasp.
Learn more about our LMS
What is the learning curve?
The learning curve is a graphical representation of the rate at which someone learns a new skill over time.
📊 Did you know? The concept traces back to psychologist Dr Hermann Ebbinghaus, who in 1885 ran a series of memory experiments on himself. He found that without reinforcement, we forget roughly 56% of new information within an hour, and around 75% within six days — a finding now known as the Ebbinghaus Forgetting Curve. It’s the flip side of the learning curve: one shows how fast we acquire a skill, the other shows how fast we lose it without practice.
This concept is vital across education, business, and psychology as a way to gauge the efficiency of learning processes.
Typically, the horizontal axis of the curve denotes time or experience, while the vertical axis represents proficiency or performance.
Initially, the curve often shows a steep incline, reflecting rapid skill acquisition, followed by a plateau where progress slows as the learner reaches a higher level of competence.
A learning curve is often expressed as a percentage indicating the rate of improvement. Visually, a steeper slope signifies rapid initial learning, leading to significant cost savings. As learning progresses, the slope flattens so further improvements are slower, and additional efficiency gains become harder to achieve.
The four common types of learning curve
Not every skill or process follows the same shape. L&D teams tend to see four recurring patterns:
| Curve type | What it looks like | Typical scenario |
|---|---|---|
| Increasing Returns | Starts slow, then accelerates upward | Learning is difficult at first with little visible payoff, then compounds quickly once the fundamentals click (e.g. learning a coding language) |
| Decreasing Returns | Steep early gains that level off | The classic “learning curve” shape — fast early progress, then diminishing returns as mastery approaches (e.g. onboarding into a new software tool) |
| S-Curve (Combination) | Slow start → steep rise → plateau | Most realistic model for complex skills: a warm-up phase, a breakthrough phase, then consolidation (e.g. becoming proficient in a new leadership role) |
| Complex / Plateau-Dip | Rises, plateaus, dips, then rises again | Common when a learner hits a skill ceiling and must unlearn old habits before progressing further (e.g. moving from “good enough” to expert-level performance) |
Why this matters for L&D: if you’re designing training assuming a simple decreasing-returns curve but your learners are actually on an S-curve, you’ll misjudge how long support and reinforcement are needed. So you could wind up pulling scaffolding away too early, right before the “steep rise” phase would have kicked in.
Phases of a learning curve
A learning curve typically comprises three key phases: the initial learning phase, the plateau, and the mastery phase.
In the initial learning phase, learners experience rapid progress as they acquire basic skills and become familiar with the task at hand — often depicted as a steep upward slope.
Following this, the plateau phase occurs, where the rate of improvement slows significantly as the learner consolidates their skills and knowledge.
Finally, in the mastery phase, performance levels off as the learner achieves a high level of proficiency, with only marginal gains from further practice.
These phases map neatly onto the four stages of competence, a model often used alongside the learning curve in L&D:
| Stage | Description |
|---|---|
| 🔴 Unconsciously incompetent | The learner has no idea what they’re doing or how to do it |
| 🟠 Consciously incompetent | The learner is aware of where they’re going wrong and what they need to work on |
| 🟡 Consciously competent | The learner has grasped the basics, but it isn’t yet natural to them |
| 🟢 Unconsciously competent | The learner has mastered the skill and can apply it without conscious effort |
How to calculate the learning curve
The learning curve formula models how the time required to complete a task decreases as a worker gains experience with it.
A commonly used form of this formula (based on T.P. Wright’s 1936 model, developed for aircraft manufacturing) is:
Y = A × X^b
Where:
- Y — Time required to produce the X-th unit (or cost, or another metric of interest)
- A — Time (or cost) required to produce the first unit
- X — Cumulative number of units produced
- b — The learning rate exponent, calculated as log(r) / log(2)
- r — The learning rate, often expressed as a percentage (e.g. 80%, 90%)
🧮 Worked example If the learning rate is 80% and the first unit takes 100 hours, then doubling production from 1 unit to 2 units means the second unit takes 100 × 0.8 = 80 hours. Double again to unit 4, and it drops to 64 hours. The pattern repeats every time cumulative output doubles.
Typical learning rates by industry
Learning rates vary a lot depending on how repetitive, manual, and standardised the task is. These are commonly cited industry benchmarks worth knowing as a reference point:
| Industry / task type | Typical learning rate |
|---|---|
| Aerospace manufacturing | ~85% |
| Shipbuilding | ~80–85% |
| Electronics assembly | ~90–95% |
| Machining | ~90% |
| Welding | ~90% |
| Repetitive clerical/admin tasks | ~85–90% |
⚠️ A lower percentage (e.g. 80%) means a steeper curve — bigger efficiency gains per doubling of experience. A higher percentage (e.g. 95%) means a flatter curve — the task was already close to optimal, so there’s less room to improve
Pros and cons of using the learning curve
| ✅ Pros | ❌ Cons |
|---|---|
| Reduced costs through increased efficiency — as employees and processes gain experience, tasks are completed more quickly and with greater precision, lowering labour costs and reducing waste over time. | Initial inefficiency and higher costs — at the start of the curve, productivity is low and error rates are higher, which can increase costs and slow output while employees get up to speed. |
| Enhanced quality of output — experience leads to a deeper understanding of tasks and workflows, reducing errors and defects as techniques are refined. | Dependence on time and repetition — the benefits rely heavily on repeated practice. In fast-paced environments expecting immediate results, there may not be enough time to realise real improvements. |
| Continuous skill growth — the model emphasises ongoing development, supporting continuous training so staff stay current with evolving tools and methods. | Resistance to change — as teams become comfortable with learned processes, they can develop resistance to new tools or methods, hindering agility and slowing adoption of better solutions. |
| Consistent and measurable performance — gives L&D teams a framework for predicting how performance improves over time, useful for tracking training effectiveness and forecasting time-to-proficiency. | Uneven progress across individuals or teams — not everyone learns at the same pace, creating inconsistent performance that’s harder to manage in large or diverse teams. |
| Reduced risk through experience — accumulated experience helps organisations get better at identifying and managing potential risks, informing more effective mitigation strategies. | Overreliance on past experience — leaning too heavily on the learning curve can make organisations favour existing methods over experimentation, risking missed opportunities to innovate or pivot. |
Putting the learning curve to work in your L&D strategy
Knowing the theory is one thing — here’s how to actually use it:
Use an LMS to flatten the curve. Structured courses, adaptive learning paths, and on-demand resources reduce the time spent in the “unconsciously incompetent” stage and get learners to competence faster.Wrapping up
Reinforce, don’t just deliver once. Because of the Ebbinghaus forgetting curve, a single training session isn’t enough. Build in spaced repetition, refreshers, and periodic testing so retention doesn’t decay as fast as the forgetting curve predicts.
Set realistic proficiency timelines. Use the formula (or just the shape of the curve) to set honest expectations with stakeholders about how long it’ll genuinely take new hires or new processes to reach target performance — rather than assuming a straight line.
Identify which curve shape you’re dealing with. A steep decreasing-returns task (like data entry) needs different support than an S-curve task (like leadership development). Match your coaching intensity and check-in cadence to the actual shape of the curve, not a generic onboarding template.
Watch for plateaus, and plan for them. A plateau isn’t a training failure — it’s consolidation. But if it goes on too long, it may signal the learner has hit a ceiling that needs a new approach (a mentor, a stretch project, different content format) rather than more of the same.
Wrapping up
Usually, people get better at doing something the more they do it. Time and resources spent doing something once will always be higher than the time and resources spent repeating that activity.
This idea of continual improvement through repeated learning should underpin your learning strategy. While you might convey an idea once within your learning content, you actually need to reference important information multiple times, offer refreshers, and test frequently.
That way, you can be sure your learners are getting — and staying — in the loop with particular job-related tasks.d upcoming events.
FAQs
What do you mean by a learning curve?
A learning curve is a graphical representation of the rate at which someone learns a new skill over time. It was first created by Dr Hermann Ebbinghaus, who in 1885, tested his own memory and knowledge retention.
What does it mean when someone says steep learning curve?
A steep learning curve is an expression that is used colloquially to describe the initial difficulty of learning something. It applies back to the learning curve principle, where it is overfacing to begin with.
What does the learning curve tell you?
The learning curve is the correlation between a learner’s performance on a task or activity and the number of attempts or time required to complete the activity.
