Cognitive Load: Why Simple Feels Effortless

Have you ever opened a screen, felt a small wave of "ugh," and closed it? That feeling has a name. It is cognitive load — the total amount of mental effort a task demands at one moment. When the load is too high, people slow down, make mistakes, and quit. When it is low, the same task feels effortless. This chapter explains where that effort comes from and how good design quietly removes it.

The idea comes from Cognitive Load Theory, developed by the educational psychologist John Sweller in the late 1980s. His starting point is a fact about the brain: working memory — the small mental "desk" where you hold information you are actively using right now — is tiny and easily overwhelmed. By contrast, long-term memory (everything you have stored over a lifetime) is effectively unlimited. So the whole design goal is simple to state: do not blow the working-memory budget.

The three types of load

Sweller split load into three kinds. You manage them differently.

Intrinsic load

How hard the thing genuinely is. Setting a tax rate is harder than typing a store name — that difficulty lives in the task, not your design. You cannot delete it. You can only manage it, usually by breaking a hard task into smaller steps.

Extraneous load

The extra hard work your interface piles on top — clutter, confusing layout, jargon, irrelevant detail. It helps the user with nothing. This is the load you create, so it is the load you can remove.

Germane load

The "good" effort the user spends actually understanding and getting their job done (forming lasting knowledge in long-term memory). You want to leave room for this.

The whole craft lives in one move: reduce extraneous load, and you free working memory for the real task. For a non-technical print-shop owner, every bit of clutter, every jargon word like slug or payload, every extra toggle is extraneous load stealing from the brain-budget they need to run their store.

The laws that govern load

Hick's Law — more choices, slower decisions

The time to make a decision grows as the number of choices grows. The precise nuance matters: it grows with the logarithm of the choices, not in a straight line. The classic form is RT = a + b·log₂(n), where RT is decision time and n is the number of options. Practically: cut options, group many options into chunked steps, and highlight a recommended default.

Miller's Law — "7±2," and the modern correction

In 1956, George Miller published "The Magical Number Seven, Plus or Minus Two." People can hold about 7 (range 5–9) items in working memory at once. Two nuances are essential:

• The unit is the chunk, not the item. A chunk is several items bound into one meaningful unit. Three words are easier to hold than nine letters. Chunking lets you beat the raw limit.
• "7" is overstated. Nelson Cowan's 2001 review put the true limit closer to ~4 chunks (about 3–5) in adults — the "magical number four." So the modern rule of thumb is roughly 3–5.

Recognition over recall

This is Jakob Nielsen's Usability Heuristic #6 (from his 10 heuristics, 1994). Recognition is spotting something as familiar when you see it. Recall is dragging it out of memory from scratch with no hint. Recognition is far easier because the interface hands you the cue.

Fitts's Law — bigger and closer targets are faster

From Paul Fitts (1954): the time to hit a target depends on its distance and its size. Big, near targets are fast and error-free; tiny, far ones are slow and frustrating. So make primary actions large, place related controls close together, and never ship tap targets that are tiny on a 375px phone. A neat trick: screen edges and corners act as "infinitely large" targets because the cursor stops there — which is why menu bars sit at the very edge.

Jakob's Law — people expect you to work like the sites they already know

Also from Nielsen, popularized in Jon Yablonski's Laws of UX (2020): users spend most of their time on other sites, so they expect yours to behave like those. The cart icon belongs top-right; the logo links home; blue underlined text is a link. Reinventing these conventions is novelty that costs the user extraneous load. This is exactly why this project's rule is "match existing patterns before inventing."

Three tools that lower load

Two vivid examples

Google's near-empty homepage is the textbook Hick's Law win. In the 1990s, Yahoo and AltaVista crammed pages with links, categories, news, and ads. Google offered essentially one choice: a search box. Decision time dropped to nearly zero.

The overloaded admin settings page is the opposite. One flat screen, dozens of ungrouped toggles, jargon labels, every option shouting at once — pure extraneous load. The user can't tell what matters, freezes (Hick's Law), and feels dumb. The fix is everything above: chunk into labeled sections, use plain labels (recognition, not recall), and progressively disclose the advanced controls.

  HIGH LOAD (flat)          LOW LOAD (chunked + disclosed)
  [ toggle ] [ toggle ]     Store Info
  [ toggle ] [ toggle ]       Name [______]
  [ toggle ] [ toggle ]     Payments
  [ toggle ] [ toggle ]       Gateway [v]
  [ toggle ] [ toggle ]     Shipping
  ...freeze, quit             [ More options ]  <- advanced

Key takeaways

• Cognitive load is mental effort; working memory is tiny, so your job is to not overspend it.
• You can't remove intrinsic difficulty, but you can — and must — remove extraneous load (clutter, jargon, needless choices).
• Apply the laws as empathy: fewer choices (Hick), chunk don't memorize (Miller), show don't make them remember (recognition over recall), big near buttons (Fitts), follow conventions (Jakob).
• Chunking, sensible grouping, and progressive disclosure are your three reliable levers — they cut load without cutting capability.
• For non-technical owners, every extra choice or thing-to-remember is a chance to feel stupid and quit; "simple feels effortless" because you spent the effort instead of them.