Dr. Wallace's Favorite Mental Models

Weber's Law (Just Noticeable Difference)

Why a $1 jump on a $10 item gets noticed immediately, and the same $1 jump on a $100 item barely registers at all.

130-Second Overview

The idea in one sentence: the smallest change in a stimulus that a person can reliably notice is not a fixed amount — it is a fixed percentage of whatever they started with, so the exact same absolute change feels huge against a small baseline and invisible against a large one.

Where it comes from

German physiologist Ernst Heinrich Weber measured it in the 1830s by having blindfolded subjects compare weights, finding that people could tell two weights apart only when they differed by a roughly constant proportion, not a constant amount. In 1860, Gustav Fechner built on Weber's data to propose that perceived intensity itself grows with the logarithm of the physical stimulus — the Weber-Fechner law.

Why it matters

Price increases, contract concessions, performance feedback, and interface changes are all judged against a baseline, not in absolute terms. Knowing roughly where that threshold sits tells you whether a change will land as negligible or as a shock — and whether staying under or clearing it is actually the right goal.

2Mental Model Compass

Type of thinking
Psychophysics / relative perception
Best for
Sizing a change so it lands exactly as intended
Difficulty
One formula, easy to forget to use it
Time to apply
2–5 minutes per decision

Reach for it any time you're about to raise a price, propose a concession, deliver a correction, or tweak something people already use — anywhere a change will be judged against what someone is already anchored to, rather than evaluated in a vacuum.

3Walk Through It

4Check Your Own Threshold

Enter a baseline value and the new value you're considering. This computes the percentage change and compares it to the rule-of-thumb range marketers and pricing teams use for when a change becomes reliably noticeable — roughly 5–10% for everyday price and quantity changes. Treat that range as a practical heuristic, not a law of nature: the real Weber fraction varies a great deal by sense and by how closely someone is paying attention (see Section 8).

5Same Percentage, Different Stakes

Pick a situation below to see how the same roughly 5% change plays out completely differently depending on what it's being compared against.

The baseline

The change

6Diverse Examples

Retail pricing
A grocery chain raises a $3.49 item to $3.59 (under 3%) without complaint, but a $3.49 to $3.99 jump (14%) draws comments at the register — the same thirty-cent gap in other categories would barely register against a $30 baseline.
Packaging & shrinkflation
A snack brand quietly trims a bag from 200g to 190g (5%) rather than raising the printed price, because the weight drop sits closer to most shoppers' noticing threshold than an equivalent price increase would.
Product & UX design
Volume and screen-brightness sliders are built on logarithmic, not linear, scales, because a jump from 10% to 20% brightness is obvious while an identical ten-point jump from 80% to 90% is barely visible.
Negotiation
A buyer who opened at $40,000 barely reacts to a seller's move to $39,000, but reacts strongly to a move from $31,000 to $30,000 — the second concession is smaller in dollars but larger as a share of where the conversation currently sits.
Performance management
A manager nudges a direct report's rating from "meets" to a slightly stronger "meets" and gets no reaction; the same manager moves someone from "meets" to "below expectations" and triggers an immediate conversation — large perceptual categories move the needle, small numeric tweaks inside one category don't.
Subscription pricing
A streaming service raises its price from $9.99 to $10.49 (5%) with few cancellations, but a jump from $9.99 to $12.99 (30%) in one step triggers a visible churn spike — many of the same companies now raise prices in several small steps rather than one large one for exactly this reason.

7Worked Application

Situation: a small software company charges $24/month for its core plan and needs more revenue to fund a hiring plan. The founder is deciding whether to raise the price to $26, $29, or $36.

  • Step 1 — Find the baseline everyone is anchored to: existing customers are anchored to $24, the number on their last twelve invoices — not to the value of the product, which may have grown substantially since they signed up.
  • Step 2 — Compute the percentage for each option: $24 → $26 is about 8%, $24 → $29 is about 21%, $24 → $36 is 50%. All three are "a couple of dollars more" in the founder's head; they are not remotely equivalent to the customer doing the math against their old invoice.
  • Step 3 — Decide which side of the threshold serves the goal: if the goal is minimal churn and minimal attention, $26 sits close to the noticeable-but-tolerable range companies commonly use for routine increases. If the goal is to reset expectations and justify it with real new features, $36 is large enough that it needs a visible value story to go with it — a bare number that size, with no accompanying change, reads as a shock rather than a reasonable adjustment.
  • Step 4 — Test and watch the actual reaction, not the planned one: the founder ships $26 with a short, honest note about rising infrastructure costs, watches the cancellation rate for two weeks, and treats a quiet response as confirmation the move stayed under the threshold — not as proof the number could have gone much higher next time without revisiting the same math.

8Common Mistakes

Thinking in dollars instead of ratios. "It's only five dollars" is meaningless on its own — five dollars on a $20 bill and five dollars on a $2,000 invoice are different universes of noticeability.
Assuming the threshold is a fixed, universal number. The 5–10% figure used in pricing is a practical rule of thumb built on top of Weber's Law, not the precise, lab-measured Weber fraction itself — that fraction differs by sense and by exactly what's being judged.
Forgetting that attention changes the threshold. Someone who is actively comparing prices, rereading a contract line by line, or watching for a change will notice far smaller shifts than the passive, half-attentive perceiver the lab constant describes.
Using it to justify deception. Sizing a change to slip under someone's notice is a different act depending on intent — tactfully phasing in a necessary price increase is not the same as deliberately shrinking a package to avoid an honest conversation about cost.
Applying it to one-off, highly salient items. A single item someone is specifically watching — the one SKU they buy every week, the one clause they flagged in the last contract — gets noticed at almost any change, because it's already outside the "background noise" the law describes.
Stacking small changes and assuming they stay invisible forever. Several below-threshold moves in a row can still add up to something the person notices in total, even if no single step crossed their threshold on its own.

9When Not to Use It

  • When the stimulus is near the top or bottom of its range. The constant-ratio relationship holds in the middle of the perceptual range; it breaks down near absolute thresholds (barely perceptible at all) and near saturation (already maxed out), where the math stops predicting behavior well.
  • For qualitative, not quantitative, changes. Weber's Law describes "more or less of the same thing" — brighter, heavier, louder, pricier. It doesn't describe a change in kind, like a product category switch or a change in color hue, which people register completely differently.
  • As a license to hide bad news. A threshold tells you what people are statistically likely to notice, not what you owe them in a transparent relationship. Treat it as a planning tool for timing and framing, not a justification for concealment.
  • When something else dominates attention. Modern signal-detection research shows detection also depends on motivation, expectation, and context, not stimulus intensity alone — a distracted, low-stakes audience and a vigilant, high-stakes one will not share the same threshold even for an identical change.
Common myth to retire: the "boiling frog" story — that a frog dropped in slowly heated water won't notice and will stay until it's too late — is often used as a folksy stand-in for Weber's Law. It is not a real finding; frogs placed in gradually warmed water detect the change and climb out well before any danger. Use Weber's Law itself, which has over a century of replicated psychophysics behind it — not the frog story, which has none.

10Related Models

Anchoring Bias

Sets the baseline this model measures changes against in the first place. Anchoring explains where the reference point comes from; Weber's Law explains how big a move away from it has to be before anyone notices.

The Contrast Effect

A close cousin: people judge a stimulus against whatever they encountered immediately before it, not in isolation. Weber's Law is the more general, quantitative version of the same underlying truth — perception is comparative, not absolute.

Prospect Theory

Describes how people value gains and losses relative to a reference point, not in absolute terms — the same reference-dependence Weber's Law describes for raw sensory stimuli, extended into decisions about money and risk.

11Try It Yourself

12Save as Skill

Turn this mental model into a reusable instruction you can hand to any AI assistant — or to yourself — before sizing any price change, concession, correction, or revision that someone else will judge against a baseline.

Copied!