The lower the comparability between options, the higher the load the same option count produces

What it is

Choice overload isn't determined by option count alone — the comparability between options plays an independent role. Ten options that all carry values on the same set of dimensions (price, rating, capacity) and can be lined up side by side produce far less load than ten options that are structurally different, don't fit into one comparison table, and where some options simply lack an attribute the others have. That second, apples-to-oranges situation raises cognitive load noticeably even when the count hasn't changed — count and comparability are two independent variables, and looking at count alone misses half the picture.

Why it happens

Comparable options let evaluation take an attribute-alignment shortcut: every option goes into the same virtual table, gets compared column by column on the same dimensions, and the decision reduces to "which is better on each dimension" followed by weighting — a structured process whose rules can be reused across every option. Incomparable options offer no such shared table. The user is forced to build a bespoke basis for comparing each pair individually, sometimes even having to first work out whether two options are even addressing the same thing. Constructing an ad hoc comparison framework like this costs far more than applying a ready-made structure, and the cost doesn't carry over between options — every additional incomparable option has to pay that construction cost again from scratch, unlike the comparable case, which just adds one more column to the same table.

Studying it

The standard way to manipulate comparability is to control what share of attributes options share: in a high-comparability condition, every option carries a value on the same set of dimensions; in a low-comparability condition, some options are missing values on certain dimensions, or the options themselves come from different subcategories (putting a suit jacket and a pair of running shoes in the same "clothing" choice set). Holding option count constant and comparing decision time, satisfaction, and abandonment rate between the two conditions, lower comparability typically comes with markedly longer decision time and a higher abandonment rate. This kind of study is common in multi-attribute decision-making and consumer-choice research; methodologically, count and comparability need to be manipulated separately — conflating them risks misattributing a comparability effect to option count.

Where it stops holding

This boundary only matters when options genuinely differ on objective dimensions — if a set of options is inherently homogeneous (different colors of the same product), comparability is already high by default and there's little a design change can add. Also, artificially forcing comparability has a cost of its own: cramming genuinely different options into the same dimensional framework can mask the differences that actually matter, so the user appears to compare more easily while actually being misled by information that got flattened out. The benefit of raising comparability only holds if it doesn't distort the real differences — otherwise it just trades "hard to compare" for "misleading judgment."

Applying it

• When presenting multiple options, first identify the core dimensions shared across all of them, and present those dimensions aligned in a uniform table or card structure, rather than designing a separate information layout for each option.
• For a dimension that genuinely exists but only some options have, mark it explicitly as "not applicable" rather than leaving it blank, so users don't mistake it for unknown or something requiring extra lookup.
• When options come from genuinely different subcategories with no natural shared dimensions, don't force them into one choice list — filter by category first and route users into a homogeneous subset before showing a comparison table.
• How to verify it: count how often users switch back and forth between option detail views to compare them, and compare that count before and after introducing a unified comparison structure. A clear drop confirms the aligned structure actually reduced load.

Related

• Same group: A9.13.1 choice overload is not a universal effect · A9.13.3 staged filtering carries less load than listing everything at once · A9.13.4 a curated small set can produce higher satisfaction than a fuller list
• Nearby: E3.10 Option count and control-type matching
• Search terms: choice overload · comparability · attribute alignment