Learning Curve: Why Getting Better at Something Is the Most Predictable Thing in Business

In 1936, an aeronautical engineer named Theodore Paul Wright published a paper that would quietly reshape how entire industries think about cost, pricing, and competitive strategy. Wright noticed that every time an aircraft factory doubled its cumulative production, the labor hours required to build each plane dropped by a consistent percentage. Not sometimes. Not roughly. Every single time.

That finding, which Wright documented at Curtiss-Wright Corporation, became the foundation of what we now call the learning curve. And if you're a marketer who thinks this is just a manufacturing concept, I'd encourage you to keep reading, because some of the most aggressive pricing strategies in modern business history are built directly on top of this idea.

What Is the Learning Curve?

The learning curve describes a systematic reduction in per-unit costs as cumulative production experience increases. The original observation was specific to labor: workers get faster at tasks as they repeat them. But the concept extends well beyond the factory floor.

The basic principle works like this: if a company experiences an 80% learning curve, then every time cumulative production doubles, the per-unit cost drops to 80% of the previous level. So if the first unit costs $100 to produce, the second costs $80, the fourth costs $64, the eighth costs $51.20, and so on.

This isn't wishful thinking or motivational speak. It's an empirically observed phenomenon documented across dozens of industries over nearly a century. The aircraft industry saw it first, but semiconductors, consumer electronics, automotive, solar panels, and even healthcare have all demonstrated the same pattern.

Learning Curve vs. Experience Curve: What's the Difference?

This is where things get interesting for marketers. In 1966, Bruce Henderson at the Boston Consulting Group (BCG) took Wright's learning curve concept and expanded it into something more ambitious: the experience curve.

Where the learning curve focused primarily on labor efficiency, the experience curve captured total cost decline, including labor, materials, overhead, distribution, and marketing. Henderson's research showed that total costs declined by 20 to 30 percent every time cumulative production experience doubled.

The distinction matters because it means that market share itself becomes a cost advantage. The company that produces and sells the most units accumulates experience fastest, which drives costs down fastest, which creates a competitive advantage that's hard to dislodge.

Texas Instruments: The Case Study That Changed Pricing Strategy

The most famous application of learning curve theory in marketing history came from Texas Instruments in the early 1970s. Morris Chang, then a general manager at TI (later founder of TSMC), worked with BCG to develop what he called "learning curve pricing."

The strategy was audacious: price your product below your current cost of production, knowing that as volume increases, your costs will fall along the experience curve until you become profitable. TI applied this to calculators. When they entered the market in 1972, handheld calculators cost hundreds of dollars. TI priced aggressively, captured massive volume, rode the experience curve down, and within a decade, calculator costs had dropped from thousands of dollars per unit to under $10.

This wasn't just a pricing tactic. It was a market domination strategy. Competitors who couldn't match TI's volume couldn't match their cost position, and many were forced out of the market entirely. It's the same logic that later drove pricing in semiconductors, solar panels, and electric vehicles.

Why the Learning Curve Matters for Marketers

I think most marketers understand the learning curve intuitively, they just don't realize they're already making decisions based on it. Here's how it shows up:

Pricing Strategy. If your company is on a steep learning curve, aggressive pricing to capture volume can be rational even at short-term losses. This is the logic behind penetration pricing and why some SaaS companies burn cash for years before turning profitable. The bet is that cumulative experience will drive costs down faster than the market price erodes.

Market Share Investment. The experience curve is the theoretical foundation for why market share matters beyond revenue. Every incremental unit of share accelerates your movement down the cost curve. That's why BCG's growth-share matrix (the famous dogs, stars, cash cows, question marks framework) placed so much emphasis on relative market position.

Campaign Efficiency. Your marketing team has its own learning curve. The first campaign you run on a new platform or in a new channel is almost always the most expensive per result. By the tenth iteration, you've learned what works, what doesn't, which audiences convert, and which creative approaches resonate. This is why I always tell clients to budget for learning when entering a new channel, your early CPM and conversion rates will not reflect your long-term performance.

Content Production. The first blog post takes 20 hours. The fiftieth takes 4. The learning curve applies to creative processes too, which is why building content marketing systems that repeat and scale beats one-off creative efforts.

Real-World Examples Across Industries

The Limits of the Learning Curve

Here's where I need to add some caution, because the learning curve has traps. HBR published a now-classic article on this topic back in 1974, and the warnings still hold.

First, the learning curve assumes your product doesn't fundamentally change. If a disruptive technology resets the game, all your accumulated experience can become worthless overnight. Kodak had decades of experience in film photography. None of it mattered when digital took over.

Second, pricing below cost based on projected learning curve gains is a bet that demand will materialize at the volume you need. If the market doesn't grow as fast as you planned, you're just burning cash. TI's calculator strategy worked brilliantly, but the same company lost hundreds of millions when it tried the same approach with home computers in the early 1980s.

Third, competitors can sometimes leapfrog your cost position through innovation rather than volume. Economies of scope can beat economies of scale if a competitor finds a fundamentally different (and cheaper) way to produce.

What's Changed: 2020–2026

The learning curve concept has found new life in the AI era. The cost of AI inference has dropped on its own experience curve, with per-token costs falling roughly 10x every 18 months according to multiple industry analyses. Companies that invest early in AI-powered marketing automation are riding this curve, accumulating process knowledge and efficiency advantages that late adopters will struggle to match.

Solar energy provides perhaps the most dramatic modern example. The International Renewable Energy Agency (IRENA) reports that solar PV module prices have fallen 99% since 1976, almost perfectly tracking an 80% experience curve. This has turned solar from a niche environmental play into the cheapest form of new electricity generation in most of the world.

For marketers specifically, the learning curve now applies to data. The more customer data you collect and analyze, the better your targeting, personalization, and ROI predictions become. First-party data strategies are essentially learning curve strategies for marketing effectiveness.

How to Apply Learning Curve Thinking to Your Marketing

FAQs

What is a learning curve in simple terms?

A learning curve describes the predictable pattern where costs go down and efficiency goes up as you gain experience doing something. In business, it means the more units you produce, the less each one costs. The rate of improvement is remarkably consistent across industries.

How does the learning curve differ from economies of scale?

Economies of scale reduce costs by spreading fixed costs over more units at a point in time. The learning curve reduces costs through accumulated experience over time. You can have both simultaneously, but they're different mechanisms. A factory that doubles production in one year gets scale benefits immediately but learning curve benefits gradually.

What is learning curve pricing?

Learning curve pricing (also called experience curve pricing) means setting prices below current costs, anticipating that costs will decline as production volume grows. Texas Instruments pioneered this approach in the 1970s with calculators. It's a high-risk, high-reward strategy that only works in markets with steep experience curves and elastic demand.

How does the learning curve apply to digital marketing?

Every new campaign, channel, or creative approach has its own learning curve. Early performance is almost always worse than optimized performance. Budget for the learning phase, track improvement rates, and use that data to project when a channel will become profitable.

Can the learning curve work against you?

Absolutely. Over-investing in a technology that gets disrupted means your learning curve experience becomes worthless. Also, if you price aggressively based on projected learning curve gains and the volume doesn't materialize, you're just losing money with no payoff.

What industries show the steepest learning curves?

Semiconductors, solar energy, battery technology, and software development tend to show the steepest curves (70-80% rates). Service industries and custom manufacturing tend to show shallower curves (85-95%).

How did BCG use the learning curve?

BCG expanded the learning curve into the "experience curve" in the 1960s, showing that total costs (not just labor) decline with cumulative production. This insight became the foundation for their growth-share matrix and influenced decades of corporate strategy around market share acquisition.

Is the learning curve still relevant in the AI age?

More than ever. AI inference costs are declining on their own experience curve. Companies that adopt AI tools early accumulate knowledge and efficiency advantages. The learning curve is also relevant for AI model training: the more data and iterations, the better the performance.

Sources & References

1. Wright, T.P. "Factors Affecting the Cost of Airplanes." Journal of the Aeronautical Sciences, 1936.
2. Henderson, Bruce. "The Experience Curve." BCG, 1968. bcg.com
3. BCG, "BCG Classics Revisited: The Experience Curve." 2013. bcg.com
4. HBR, "Limits of the Learning Curve." 1974. hbr.org
5. Commoncog, "Texas Instruments: Inventing Learning Curve Pricing." commoncog.com
6. Corporate Finance Institute, "Experience Curve." corporatefinanceinstitute.com
7. Wikipedia, "Experience Curve Effects." wikipedia.org
8. StrategyU, "The Experience Curve: BCG's Framework That Changed Corporate Strategy." strategyu.co
9. Strategosinc, "Learning Curves in Manufacturing and Marketing Strategy." strategosinc.com
10. Wall Street Mojo, "Experience Curve: What Is It, Vs Learning Curve, Examples." wallstreetmojo.com

Written by Conan Pesci | April 4, 2026 | Markeview.com

Markeview is a subsidiary of Green Flag Digital LLC.