An IKEA complexity index scores 21,290 SKUs on steps times parts, which is mostly size squared

Multiply two numbers that both grow with the size of the furniture and you get size, squared. A BILLY bookshelf carries 48 parts across 13 steps for a score of 624 ^[3][4]. The HEMNES that tops the single-manual list carries 290 parts across 74 steps for 21,460 ^[8]. On parts the dresser is 6.0 times the bookshelf; on steps, 5.7 times ^[1][2]. The published score puts it 34 times higher, because 6.0 times 5.7 is 34 ^[3]. There is no assembly fact in that gap. It is what happens when two correlated inputs are multiplied rather than averaged.

The BILLY variant makes the same point on a smaller board. Adding the foldable table takes the score from 624 to 6,586, a factor of 10.6 ^[5][4]. Gizmodo went to the manual and found the addition is a smaller BILLY on wheels with a hinged top that rides a curved rail into position ^[6], then asked, reasonably, whether that is ten times the work ^[7]. Because the score is a product of two integers, 6,586 also reveals something the site does not print: it factors as 2 x 37 x 89, so the counts behind it can only be 74 steps by 89 parts, 37 by 178, or 2 by 3,293 ^[5]. Take the first pair. Not quite double the parts, and the score rises more than tenfold.

Then the doorless PAX. Doors add hardware and steps, so a configuration without them outscoring one with them ^[11] says something about which manuals were bundled into which SKU, not about the afternoon. When a metric inverts at the top of its own ranking, the ordering below it is a guess with four digits attached.

What the counts miss is visible in the HEMNES itself. Per Gizmodo, the step people get stuck on is the anti-tip interlock that stops two drawers opening at once ^[12], and that is one mechanism, which a tally of 290 parts cannot distinguish from 290 dowels. Nor does the score see the 120-plus pounds spread across three boxes that has to be moved before step one ^[9][6].

The useful artifact sits underneath the ranking: step and part counts for 21,290 SKUs ^[1], plus a second sort that borrows IKEA's own ease-of-assembly rating, which IKEA collects separately from the overall product score ^[13][14]. That field is a measurement of the experience rather than a proxy for its size, carrying the complaint-skew Gizmodo flags ^[15]. If the index showed both inputs beside every score and let buyers weight them, it would lose the headline number and gain an answer to the question people actually have, which is not how big the furniture is but how long the evening will be.