Research
What 42 public assembly manuals have in common
Published
We read 37 public assembly manuals in full and counted the pages of 5 more. Here is what they have in common, how we measured it and what we did not measure.
Key numbers
| Measure | Result |
|---|---|
| Pages per manual (n = 42) | median 28 (27.5), mean 45, range 2 to 180 |
| 50 pages or more | 18 of 42 |
| 20 pages or fewer | 15 of 42 |
| Print 2+ languages in one document (where stated, n = 33) | 18 of 33 |
| One language only (n = 33) | 15 of 33 |
| Languages in one document | median 2, range 1 to 14 or more |
| Say assembly needs 2 or more people (n = 37) | at least 18 of 37 (others do not say or say "an adult") |
| State an assembly time (n = 37) | 6 of 37 (40 minutes to one day) |
| Point to a QR code or video (n = 37) | at least 9 of 37 |
| Tell the customer not to return the product to the store / contact the maker first (n = 37) | at least 6 of 37 |
| Words per manual (13 manuals counted) | median about 4,000; range 328 to 38,757 |
"At least" = counted only where the reader noted it.
What stands out
- Size varies enormously: half of the manuals are 28 pages or fewer, and 18 of 42 run to 50 pages or more.
- In 18 of 33 manuals the customer reads a document printed in two or more languages; the longest have 14 or more, so most of the pages are not in the customer’s language.
- At least 18 of 37 say the job needs two people; only 6 of 37 say how long it takes.
- At least 9 of 37 already send the customer to a QR code or a video, and at least 6 of 37 tell the customer to contact the maker rather than return the product: the manuals themselves admit that the paper is not always enough.
- We make UnboxStep, an interactive assembly guide; read these findings with that in mind.
Method and limits
We read 37 public assembly manuals in full in October and November 2026 and recorded the page count of 5 more: 42 in all. Products: sheds, gazebos, greenhouses, playsets, trampolines, grills, bunk and loft beds, bookcases and wardrobes, TV stands, high chairs and strollers, e-bikes, kids' bikes, rowing machines and similar. Sold in North America and Europe. Published by the manufacturer or hosted by a retailer.
This is a convenience sample chosen for our research, not a random sample of all manuals. Fields are recorded as each manual states them; empty fields are not counted. Brand names are not published.
What we did not measure
- The number of steps (not reliably countable from the text).
- The quality of the drawings.
- Whether customers succeed in assembling the product.
How to cite
UnboxStep, Assembly manuals study, November 2026. unboxstep.com/research/assembly-manuals-2026
Published under CC BY 4.0: you may use the numbers with a link to this page.
Update note
v2, November 2026. We plan to add more manuals, in particular large outdoor products.
Frequently asked questions
How were the manuals chosen?
It is a convenience sample chosen for our research, not a random sample of all manuals: public assembly manuals of products sold in North America and Europe, published by the manufacturer or hosted by a retailer.
Are brand names published?
No. The study publishes aggregate numbers only; brand names are not published.
Can I use these numbers?
Yes, with a link to this page. Cite it as "UnboxStep, Assembly manuals study, November 2026". The data is published under CC BY 4.0.
What is missing?
We did not measure the number of steps, the quality of the drawings or whether customers succeed. We plan to add more manuals, in particular large outdoor products.