Equal-weight foci fork the flow

What it is

A landing page puts a value proposition at the same type size as the title in the left column and a product photo of matching dimensions on the right; both fill the first screen. Gaze bounces between title and photo two or three times before it reaches “Get started” below. Two equal foci split the flow. This is not purchase hesitation. There is no unique centre of interest, so the selector cannot return a first place. What forks waste are the first seconds after entry. The behaviour is back-and-forth, not a miss.

Why it happens

Attentional selection is close to winner-take-all: the peak has to stand clearly above the runner-up or first place will not stay put. When two objects tie on size, contrast and isolation, the peak gap sits in noise and the system oscillates until one side is “chosen off” by comprehension or fatigue. During the oscillation, third place (the actual action) cannot join the queue. Designers often read “visual balance” as equal weight left and right. Balance can be true of the composition and still be a manufactured double peak in the flow. Forking is not a gentle “both matter, so look at both.” It delays every later object.

Studying it

Watch the first seconds of the scanpath: round-trips between two areas of interest above a “visit once and leave” baseline, plus delayed arrival at a third area, counts as a fork. The independent variable is the weight gap between the two peaks (shrink one block’s size or contrast in steps); dependents are round-trip count and time to the third object. Past some gap, round-trips should collapse to a single-peak sequence. Two tied firsts and no third in five-second writing can mark the same behaviour, less directly than gaze.

Where it stops holding

When side-by-side comparison is the task (two plans, two candidate photos), forking is the mechanism, not a defect — people need to oscillate to decide. Then keep the comparands equal and make the real next step (“pick this”) a clear third peak, rather than crushing one side. A carousel or autoplaying video can move the peak from one side to the other in time, so a still double peak becomes a switching single peak. On small screens, left/right becomes stacked, the fork may vanish, and the problem turns into discovery below the fold.

Applying it

• Allow one centre of interest on the first screen. If a second block must stay, make it clearly lighter on at least one channel: lower photo contrast, drop title size, pull whitespace off one of them.
• “Left-right symmetry looks premium” is not a constraint on a first screen. Use symmetry in decoration; the action region needs a measurable first place.
• On a page whose job is comparison, make the pair equal and give a heavier action of its own, instead of three blocks shouting together.
• Check: record pointer or gaze for three seconds after entry. More than two round-trips between the two large blocks before a third dock is a fork. Shrink one block until round-trips die, then see whether “Get started” enters those three seconds. Do not add motion to both large blocks — that turns a fork into two equal captures.

Related

• Same group: F3.05.1 Flow follows the weight sequence, not positional order · F3.05.3 Flow should match the order of the task
• Nearby: F3.08.1 Emphasising everything emphasises nothing · F3.01.2 A difference on one dimension is easily cancelled by another · F5.11.3 The number of accent points on a screen needs a budget
• Search terms: centre of interest · competing foci · visual flow · winner-take-all