Industries / Consumer goods / Textiles and apparel / Footwear
Fit and sizing
Fit and sizing is the work of measuring feet, labelling shoes with sizes, and matching the two, whether in a shop, at home or online.
7
Milestones achieved, from 1688 to 2019.
2
To watch for: announced, not yet confirmed outside.
3
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1688
Sizes first written down
SATRA, the British footwear technology centre, names Randle Holme III's The Academy of Armory (1688) as the first recorded description of a shoe-size system, with "each size being the fourth part of an Inch": a child's shoe of size one or two was 5½ inches long, and men's and women's shoes started at 8¼ inches and ran to 15. The step of one third of an inch (one barleycorn, 8.47 mm) that UK and US sizes use today is next documented in Robert Gardiner's Illustrated Handbook of the Foot (London, 1856), and the New York businessman Edwin Simpson introduced a detailed system with widths in 1880, which the US trade generally adopted about a decade later, by SATRA's account.
LessonThe sources disagree on the barleycorn story: SATRA says Edward II decreed the barleycorn inch in 1324; Wikipedia dates the statute that says "3 grains of barley dry and round do make an inch" to an estimated 1266–1303, and its text does not mention shoes; Today I Found Out says "there is no evidence" that Edward set a shoe standard. The oldest sizes on record are quarter inches, not barleycorns.
rholme.taths.org.uksatra.comen.wikipedia.orgtodayifoundout.com
Credit
Randle Holme III (first record); Robert Gardiner; Edwin Simpson
1928
Length, width and arch
Charles F. Brannock of Syracuse, New York, where his family ran the Park-Brannock Shoe Co., filed in November 1925 for a foot-measuring instrument that sized the shoe "both as to length and width", granted in August 1928 as US patent 1,682,366. A second patent, filed in September 1927 and granted in August 1929 (US 1,725,334), also measured "the length of the foot from heel to ball joint", the arch length. Lemelson-MIT, MIT's invention programme, reports more than one million devices sold, says the US Army hired Brannock in the Second World War to make sure boots fitted enlisted men, and says the device "remains the standard for the footwear industry".
LessonBefore it, by Lemelson's account, "the available option was a primitive block of measured wood".
Credit
Charles F. Brannock
1957
X-ray fitting banned
The shoe-fitting fluoroscope was an X-ray machine that showed the bones of the feet inside a shoe. The historians Duffin and Hayter give the strongest claim to its invention to Jacob Lowe, a Boston physician, whose Foot-O-Scope appeared in shoe-shop advertising in 1920 and whose patent (1,614,988) was granted on 18 January 1927. By the early 1950s about 10,000 machines were in use in the United States, 3,000 in the United Kingdom and 1,000 in Canada, and measurements from the late 1940s found doses to the feet of 7 to 14 R (roentgen, an old unit of X-ray exposure) for a 20-second look. In 1957 Pennsylvania became the first jurisdiction to ban them; state rules, warnings from professional bodies and pressure from insurers ended their use in the US by about 1960, but they stayed in limited use in Canada and the UK until at least 1970.
LessonNo injuries to customers were reported, but operators were hurt: a shoe model's radiation burn led to the amputation of her leg, and one survey found salesmen saw the machine as "a sales gimmick rather than a useful tool", by the account of Oak Ridge Associated Universities' radiation museum.
Credit
Jacob Lowe (invention); the State of Pennsylvania (ban)
1970s
Sizes in millimetres
Mondopoint, devised in the early 1970s, gives a size as the length of the foot in millimetres, in steps of 5 mm, where the UK and US count barleycorns and most of mainland Europe counts Paris points of two thirds of a centimetre (6.67 mm). The current standard, ISO 9407:2019 (second edition), bases Mondopoint on "defined measurements of the foot that the footwear is intended to fit" and applies it to all types of footwear. SATRA reports that it is used extensively in China, Japan and other parts of Asia, and was used in the former East Germany.
Credit
the International Organization for Standardization (ISO 9407)
2015
Conversion tables, as a guide
In 2015 ISO published ISO/TS 19407, three tables for converting between the main shoe-size systems, replaced in June 2023 by the full standard ISO 19407, which covers Mondopoint, European and UK sizes and those of China, Japan, the Republic of Korea and the United States, all based on foot length. Its scope says each system developed "without being formalized", so two of its tables only "provide a guideline which can benefit the consumer".
LessonThe standards body does not claim an exact conversion, and Ales Jurca of the scanner maker Volumental wrote in 2021 that it is not clear whether size standards refer to foot or shoe length, and that "Almost every footwear brand has its own interpretation of the sizing standards.".
Credit
ISO
2019
A million feet scanned
Ales Jurca, Jure Žabkar (University of Ljubljana) and Sašo Džeroski (Jožef Stefan Institute) analysed 1,200,847 barefoot scans taken from November 2017 to August 2018 in 712 shops using Volumental's 3D scanners in North America, Europe and Asia, and reported in Scientific Reports in December 2019 that within a single length, foot width varied so much that "at least 3 shoe widths per length class are required to provide a proper fit for 90% of customers". They also found that grading tables, the rules for how much a shoe grows from one size to the next, "haven't changed for decades". Fleet Feet, a US chain of running shops, says it has offered Volumental scans in its shops since 2017, and Carlos Tamayo and John Smith of Texas A&M University–San Antonio found in 2025 that the scanner's measures were consistent (reliability above 0.90) across repeat scans of 20 adults.
LessonJurca declared a financial interest in Volumental, and the raw scans "are not publicly available, because the data are owned by third parties".
Credit
Volumental; Jurca, Žabkar and Džeroski
2019
Returns counted
The University of Bamberg's returns research group, led by Björn Asdecker, surveyed 68 German online retailers in late 2018 and found that shoes and clothing came back most often, about 46% of parcels, against about one parcel in six across all online goods. Zalando, the Berlin online fashion retailer, reported in 2021 on a 2017 test of size advice for shoes with 720,000 customers given advice and as many given none: the advice cut size-related returns by 3.8% relative to the group without it, and telling customers a shoe ran small or large raised orders of two sizes of the same shoe at once (called bracketing) by 11.1% and 19.0%.
LessonBy Zalando's own measure, size advice reduced returns by a few per cent, and some advice led customers to order a second size.
Credit
Björn Asdecker and the Forschungsgruppe Retourenmanagement, University of Bamberg (survey); Andrea Nestler, Nour Karessli, Karl Hajjar, Rodrigo Weffer and Reza Shirvany, Zalando (test)
Watch for
One width fits under half
Ales Jurca of Volumental, the Stockholm scanner maker, wrote that the company has scanned more than 80 million feet in more than 3,000 shops in over 60 countries, that most lasts still in use "share a basic geometry that has not fundamentally changed in decades", and that a single width per size "leaves more than 50% of the population without a well-fitting option".
Achieved whena group with no stake in Volumental reaches the same share from scans it can publish, paired with the sizes people bought and kept.
Claimed by
Ales Jurca, Volumental, June 2026
Running shoes with no size
FitX says its FitX 01 running shoe ($299) is built from a phone scan taken in a web browser, with the buyer standing on a sheet of A4 or Letter paper as a size reference, so the buyer chooses no size or width; the scan and the buyer's fit preferences shape the upper and the midsole, a 3D-printed lattice over a foam base, with each foot treated separately.
Achieved whenshoes reach buyers and an outside tester reports, for at least a few dozen pairs, how many fitted without remaking or exchange.
Claimed by
FitX, as reported by Marko Maslakovic in Gadgets & Wearables, November 2026; his pair was due in January 2027 and he had not yet worn it
Open
01
Inside length, same label
A public table of the inside length and width of at least 40 adult shoe models from at least 15 brands, in walking, dress, work and casual shoes (running shoes are already measured by others), each in three labelled sizes (for example UK 6, 8 and 10) with a photo of each size label, so readers can see how much room each label gives and whether each size step adds a barleycorn (8.47 mm), a Paris point (6.67 mm) or something else.
The briefCredit · unclaimed02
Open scans with kept sizes
An openly licensed set of foot scans from at least 500 consenting adults, each paired with the brand, model and labelled size of shoes the person buys during the study, whether they kept or returned each pair, and a fit rating after two weeks of wear, with a baseline model that predicts keep or return from foot length, width and labelled size.
The briefCredit · unclaimed03
Retail size apps compared
A public test of the foot-measuring and size apps that shoe brands and retailers offer shoppers. Recruit 50 adults; measure each foot with calipers by two people and once with a Brannock Device; then measure each foot three times with each app. Report each app's mean error and limits of agreement (the range that holds 95% of the differences) in millimetres, and how often its recommended size differs from the Brannock size.
The briefCredit · unclaimed
Open 01 · the brief
Inside length, same label
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public table of the inside length and width of at least 40 adult shoe models from at least 15 brands, in walking, dress, work and casual shoes (running shoes are already measured by others), each in three labelled sizes (for example UK 6, 8 and 10) with a photo of each size label, so readers can see how much room each label gives and whether each size step adds a barleycorn (8.47 mm), a Paris point (6.67 mm) or something else.
Why it is still open
RunRepeat casts the inside of running shoes in gel, says "We always test in men's US 9", and records inner width and height. Shoefitr, a Pittsburgh start-up that scanned shoes "both inside and out" for retailers such as Nordstrom, Cole Haan and REI, was reported bought by Amazon in 2015, and its data was never public. In Austria a health insurer and kidsfeet.info checked 1,898 pairs of children's shoes in 2018 and found 86% shorter inside than advertised. Volumental found that two women's running shoes labelled US 9 medium suited feet 251.7 mm and 245.0 mm long. Searches for public tables of shoe interiors found none for adult shoes across a size run.
runrepeat.comrunrepeat.comtechcrunch.comeuronews.comproc.3dbody.tech
Built with what exists
A sliding inside-length gauge, used in shops with the manager's permission; a two-part silicone or alginate casting compound for casts of a subset, as RunRepeat does with gel; digital calipers; a photo stand; a vision-language model such as Claude Fable 5.2 to transcribe the label photos into the table (see the Note "Three working devices"); a spreadsheet.
First step
Measure three sizes of two models in one shop with the gauge, cast one pair of each, and write the method down.
Done looks like
40 models in three sizes, two pairs per size (240 pairs), each measured twice by two people with the repeat error reported, and the spread of inside length within each label and the average size step per brand published, with all data and photos under CC BY.
Open 02 · the brief
Open scans with kept sizes
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
An openly licensed set of foot scans from at least 500 consenting adults, each paired with the brand, model and labelled size of shoes the person buys during the study, whether they kept or returned each pair, and a fit rating after two weeks of wear, with a baseline model that predicts keep or return from foot length, width and labelled size.
Why it is still open
The 1.2 million scans in the 2019 study are not public, and Volumental's 2021 match of 90,720 scans to unreturned purchases of 841 women's running shoes is private too. Cambridge's Foot3D release offers 118 high-resolution foot scans but no shoe sizes. The MORFO3D database (2004) holds scans of 316 women of EU size 38 who tried shoes and filled in fit questionnaires, "available upon request for non-commercial use". A search for an open dataset of foot scans paired with shoe sizes and returns found none.
Built with what exists
A phone-photo or depth-camera method of rebuilding the foot in 3D, checked against Foot3D's scans; a Brannock Device and calipers to check each scan; a consent form approved by a university ethics committee; an online form for purchases and returns; Gemini Decision 1 to sort free-text fit comments into fixed answers such as too short, too narrow or heel slips (see the Note "A thousand questions, one pass"). Publish with no names, faces, exact ages or locations.
First step
Write the consent form and data plan, then scan and hand-measure 20 volunteers.
Done looks like
500 people and at least 1,000 purchases with kept or returned outcomes, published under CC BY, with the baseline model's accuracy on held-out people reported beside the simple rule "buy your Brannock size".
Open 03 · the brief
Retail size apps compared
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public test of the foot-measuring and size apps that shoe brands and retailers offer shoppers. Recruit 50 adults; measure each foot with calipers by two people and once with a Brannock Device; then measure each foot three times with each app. Report each app's mean error and limits of agreement (the range that holds 95% of the differences) in millimetres, and how often its recommended size differs from the Brannock size.
Why it is still open
A 2021 review of 26 foot-measuring apps by Muhammad Ashad Kabir and colleagues found "little information regarding the evidence-based quality of these apps", and that they "do not follow any specific guidelines for measurement purposes". The accuracy figures shoppers hear are the makers' own: Nike said in 2019 that its phone scan recommended sizes "down to a millimeter preciseness", and Cambridge said in 2022 that the Snapfeet app "knows where the phone is to within 2mm accuracy". The Texas A&M–San Antonio study of Volumental's shop scanner measured repeatability, not accuracy. A search found studies of single research apps, but no side-by-side test of the size apps that shoe brands and retailers offer.
Built with what exists
Digital calipers and a foot-measuring board; a Brannock Device; the brand and retailer apps on offer at the time; one iPhone and one Android phone; Python for the agreement statistics.
First step
Measure ten volunteers twice each with calipers and the Brannock, to learn the hand-measurement error before testing any app.
Done looks like
100 feet from 50 adults and at least four apps, all raw readings published without names, each app with a mean error and limits of agreement in millimetres for length and width, and the share of feet where its size differs from the Brannock size by half a size or more.