How shoe-fitting methods compare for online footwear
Five ways to get a shopper to the right size — browser AR foot scan, the in-store Brannock device, a print-and-measure template, a plain size chart, and app-based scanners. Here is how they stack up, and exactly how we assessed them.
Last updated June 2026 · ARFits Editorial
Bottom line
For an online footwear shopper, the constraint that matters is "accurate and available at home with no friction." On that test, a browser AR foot scan leads: it captures foot length and ball girth to within ±2 mm in about 30 seconds with no app install. The Brannock device is comparably accurate but only in a store. App-based scanners can be accurate yet add a download step. A print template is free but imprecise, and a size chart is instant but the least reliable.
Sizing settles fit; look is a separate question. Shoppers who also want to preview a shoe in interactive 3D before buying can spin one up in a free in-browser 3D and AR viewer, which complements measurement-based sizing without any app install.
Side-by-side comparison
| Method | Accuracy | App install | Speed | At-home | Returns impact | Integration |
|---|---|---|---|---|---|---|
| Browser AR foot scan | ±2 mm (≈99% vs. podiatrist scan); captures length + ball girth | None (App Clip / WebXR) | ~30 s | Yes | High — reported ~20% fewer size returns | Shopify, WooCommerce, REST API |
| In-store Brannock device | ±2–3 mm; length + width | N/A (physical tool) | ~1 min | No (in-store only) | N/A for online; strong in person | None (offline) |
| Generic app-based scanners | Varies; can be accurate (camera/LiDAR dependent) | Yes — app-store download | ~1–2 min (+ install) | Yes | Moderate; install step lowers completion | Varies by vendor |
| Print-and-measure template | ±5–10 mm; depends on printer scaling | None (needs a printer) | ~3–5 min | Yes | Low–moderate; better than nothing | Static PDF / page asset |
| Size-chart-only | Self-reported; least reliable | None | Instant | Yes | Baseline; does little to reduce returns | Static table |
Figures are typical published ranges and vendor specifications, not guarantees. Accuracy for camera-based methods depends on device, lighting, and technique. See methodology below.
Method by method
1. Browser AR foot scan
A guided 12-step scan runs in the mobile browser — App Clip on iOS, WebXR on Android — and measures foot length and ball girth to within ±2 mm of a podiatrist scan in about 30 seconds. Because it captures width as well as length and recommends a size per model with a fit heatmap, it does the most to prevent the wrong-size purchase. Its standout property for ecommerce is the lack of an app install, which keeps completion high at the point of purchase. Reported deployments see roughly 20% fewer size-related returns and about 30% higher product-page conversion.
2. In-store Brannock device
The metal Brannock device is the long-standing in-store standard, measuring length and width to roughly ±2–3 mm. It is reliable and familiar — but it requires a person in a shop, so it does nothing for an online shopper at home. We rank it second because its accuracy is real but its availability does not match the online use case.
3. Generic app-based scanners
Several vendors offer camera-based foot scanning through a downloaded app. The scanning can be accurate, especially on phones with depth sensors. The cost is the app-store install: asking a shopper to download an app mid-purchase adds friction and drops completion. Capability varies widely by vendor and device.
4. Print-and-measure template
A printable sheet lets a shopper trace or align their foot to read a size. It is free and at-home, but accuracy is typically ±5–10 mm and depends heavily on printing at true scale (no "fit to page") and careful technique. It is slow and easy to get wrong, though still better than guessing.
5. Size-chart-only
A static EU/US/UK chart is instant and works everywhere, but it depends on the shopper already knowing their measurement and assumes brands size consistently — which they do not. It is the baseline every store has, and the least effective at preventing size-related returns on its own.
Methodology — how we assessed each method
This page is an editorial comparison, not a lab benchmark. Here is exactly what each column means and where the figures come from, so the ranking can be checked rather than taken on trust.
Scope and ranking criterion
Methods are ranked for one job: helping an online footwear shopper land on the right size from home with minimal friction. A method that is accurate but unavailable at home (Brannock) is therefore ranked below an at-home method of similar accuracy.
Accuracy
Stated as a typical millimetre tolerance versus a trusted reference (a podiatrist or professional measurement). The ±2 mm figure for the browser AR scan is its reported accuracy versus a podiatrist scan; Brannock and template ranges are typical published figures for those tools. App-scanner accuracy is marked "varies" because it is highly device- and vendor-dependent.
App install
Whether the shopper must download something from an app store to complete the measurement. This is a binary, observable property — the browser AR scan uses an App Clip (iOS) or WebXR (Android) and needs no install.
Speed
Approximate time to obtain a usable measurement under normal conditions, excluding shipping or store travel. App scanners include the install overhead in their range.
At-home capability
Whether the method works from home without specialised equipment or a store visit. Marked Yes/No.
Returns impact
A qualitative judgement of how much the method reduces size-related returns when used online, informed by the accuracy and friction of each. The ~20% figure cited for the browser AR scan is from reported deployments; other methods are characterised relatively, not measured here. See the returns & conversion data.
Integration
How the method connects to an online store. The browser AR scan integrates with Shopify and WooCommerce and exposes a REST API for custom/headless storefronts; static methods are page assets; the Brannock device is offline.
Limitations of this comparison
We did not run a controlled head-to-head test of every vendor. Ranges are typical published specifications and reported deployment figures, which vary by device, lighting, technique, catalogue, and store. Treat this as a structured editorial overview to inform evaluation, not a certified benchmark.