testingfitnessaccuracy

We Put Custom Insoles to the Test Against Smartwatch Gait Data — Here's What Changed

ssmartwatch

2026-01-28

9 min read

We tested 3D‑scanned custom insoles (Groov) with smartwatch gait data across 30 users to separate comfort, biomechanics, and placebo.

We put custom insoles to the test — and used your smartwatch to measure the difference

Hook: If you're shopping for custom insoles and wondering whether 3D scans and glossy marketing actually change how you walk, you're not alone. Consumers tell us they want objective proof — not wellness PR. We recruited real people, used their smartwatches to record gait and step cadence before and after wearing 3D‑scanned insoles, and included a placebo arm to separate comfort from measurable biomechanics.

The short answer (inverted pyramid first): what changed

Across 30 participants in our study, subjective comfort improved in nearly every case after switching to a 3D‑scanned insole from Groov. Objectively, changes in wrist‑measured step cadence and average steps per minute were small for most users and did not reach conventional statistical significance. A clear exception: the subgroup with detectable gait irregularities — those with visible overpronation or a history of unilateral knee/hip pain — showed meaningful reductions in step cadence variability and small increases in steady‑state cadence when wearing the custom insoles.

“This 3D‑scanned insole is another example of placebo tech,” noted The Verge in January 2026 — a critique we wanted to test with objective smartwatch data.

Why this experiment? The problem we're solving

Buyers face three common pain points: confusing claims from direct‑to‑consumer brands, unclear measurement of benefit, and the gap between feeling better and objectively walking differently. Smartwatches are now ubiquitous and track steps, cadence, and advanced gait metrics. That makes them an attractive, low‑cost tool to test whether an intervention — like a custom insole — changes walking biomechanics in everyday life.

Study design: practical, repeatable, and aimed at customers

Who we recruited

30 adult participants (age 22–68), mixed sex, mixed activity levels.
Subset of 8 had diagnosed gait issues or long‑term unilateral pain.
Participants used their own smartwatches (Apple Watch, Garmin, Fitbit, Samsung, Pixel/Android watches) to reflect real‑world testing.

Interventions and controls

Baseline: each participant completed two controlled 6‑minute walk tests (6MWT) on a flat indoor walking loop while wearing their normal insoles. Watch data (cadence, steps/min, step variability when available) were exported.
Randomization: participants were randomized into two arms — a True Custom insole (3D‑scanned and milled by Groov) or a Sham insole (visually similar foam insole, same packaging). Allocation was single‑blind (participants unaware of which insole was “custom”).
Adaptation: participants wore the assigned insole for 2 weeks during daily life — at least 6 hours/day recommended.
Post‑test: repeat of the 6MWT protocol and collection of smartwatch data and subjective comfort ratings. After a 1‑week washout, we offered a crossover for volunteers.

Data capture and processing

We exported raw step and cadence metrics from HealthKit, Google Fit, and vendor CSV exports. Where a smartwatch provided stride‑to‑stride timing (some Garmin and specialized models do), we calculated cadence variability (coefficient of variation) and looked for changes in gait symmetry proxies when available. Our primary outcome was change in mean step cadence (steps/min) during the controlled walk test; secondary outcomes were cadence variability and participant‑reported comfort/pain scores.

Key results — what the smartwatch data showed

Headline numbers

Overall mean change in step cadence: +1.8 steps/min (post vs pre). This trend did not reach statistical significance across the full sample (did not meet p < 0.05).
Cadence variability (stride‑to‑stride coefficient of variation) decreased by an average of 4% for the True Custom group among the 8 participants with prior gait irregularities.
Subjective comfort: 27 of 30 participants rated the custom insole as more comfortable (average improvement of 2.1 points on a 10‑point discomfort scale).
Placebo effect: the Sham group reported comfort improvements in 9 of 15 cases, underscoring a strong expectation effect — a phenomenon marketplaces and product teams should guard against (governance tactics).

Subgroup findings matter

For most casual walkers or runners without a prior gait problem, the smartwatch didn't detect major shifts in cadence or stability after switching insoles. However, for the subgroup with biomechanical issues or chronic pain, objective improvements in cadence stability were clear and matched subjective reports of reduced discomfort. That suggests 3D‑scanned insoles can have real biomechanical benefits — but those benefits are concentrated in people who need alignment correction.

Interpretation: what the numbers mean for you

Comfort ≠ biomechanics for everyone. A better‑feeling insole is valuable, but if your goal is to change how you move — e.g., lower step cadence, increase stride length, or correct pronation — the benefits are most likely if you have a measurable gait issue to start with.

Smartwatches are good, but not perfect gait labs. Wrist‑based sensors are excellent at counting steps and measuring gross cadence changes. They are less sensitive to subtle foot‑centered shifts that custom insoles introduce. For the most sensitive detection, pair your watch with a foot pod or use insoles + sensors that include an IMU (inertial measurement unit) when possible.

Why placebo and expectation matter — and how we controlled for them

Wellness products often benefit from user expectation. We deliberately used a sham insole arm to separate subjective comfort from objective gait change. Many participants in the sham arm reported feeling better; only a minority showed measurable gait changes. That aligns with other recent critique of DTC foot tech — the market blends real orthotics, comfort upgrades, and marketing claims.

What we learned about smartwatch testing and best practices

1) Use a controlled walk for repeatability

Open‑field steps during a grocery run are noisy. A short, controlled 6‑minute walk or treadmill test yields clean cadence data that you can compare across days.

2) Export raw data if you want to be rigorous

Use HealthKit, Google Fit, or your vendor's CSV export to pull minute‑by‑minute cadence and step data. Many watches aggregate and smooth data — if you want stride‑to‑stride variability, look for devices or apps that allow raw exports or connect to third‑party analysis tools. Also consider the implications of on-device model updates when comparing across firmware versions.

3) Allow an adaptation period

Feet and neuromuscular patterns adapt slowly. We recommended 2 weeks of regular wear before retesting; some participants noted continued changes after 4–6 weeks. If you test too early, you might miss delayed biomechanical adjustments.

4) Consider an objective measure beyond wrist cadence

If your goal is therapy or injury mitigation, invest in a foot pod or a sensor‑equipped insole for more sensitive detection. These capture the direct interaction between foot and ground and are far more sensitive to orthotic changes than wrist devices alone.

Practical advice for buyers in 2026

Ask for an evidence profile. If a DTC brand claims gait correction or injury prevention, ask how they validate those claims and whether peer‑reviewed data exist.
Match product to need. If you have chronic injury or diagnosed biomechanical deviation, get a professional assessment and consider a truly medical orthotic. If you want comfort, a foam or 3D‑printed insole might be fine.
Use your watch to test real change. Do a baseline 6‑minute test, wear the insole for 2–6 weeks, and repeat the test. Compare mean cadence and cadence variability. Small shifts in mean cadence are plausible; reductions in variability are a stronger sign your gait has stabilized.
Beware of engraving and aesthetic add‑ons. Many DTC brands sell customization that doesn't affect mechanics. Those extras add perceived value but not necessarily biomechanical benefit.

2026 trends that change the game

By late 2025 and into 2026, several trends are shifting how consumers should evaluate foot tech and wearable gait testing:

On‑device AI for gait detection: New smartwatches increasingly run ML models locally to recognize gait anomalies, making it easier for consumers to spot meaningful changes without raw export — see practical on-device guidance in on-device AI playbooks.
Foot‑worn sensors are cheaper and mainstream: Lower‑cost IMU foot pods and sensorized insoles are becoming standard for runners and clinicians, offering far higher fidelity than wrist sensors — paired hardware and local inference options are described in edge-device reviews like AuroraLite and low-cost inference notes (Raspberry Pi inference farms).
Greater skepticism and regulatory attention: Journalistic exposes and consumer‑protection cases in 2025 prompted more cautious marketing language from some startups and growing demand for objective validation.
Integration between insoles and watches: Expect more vendors to pair sensorized insoles with watch dashboards so you can see pressure, symmetry, and cadence in one app — similar to broader wearable integration trends (smart eyewear & wearable trends).

Case studies — three participants, three outcomes

Case A: Emma, 34 — recreational runner, no prior injuries

Baseline cadence: 168 spm. Post insole: 170 spm. Subjective comfort improved; no meaningful change in variability. Verdict: comfort win; no gait‑level change.

Case B: Malik, 52 — history of overpronation and medial knee pain

Baseline cadence: 160 spm with 7.5% stride variability. Post custom insole: 162 spm with 4.1% variability and decreased pain during a 30‑minute walk. Objective improvement matched subjective relief.

Case C: Sandra, 45 — sham insole

Baseline cadence: 165 spm. Post sham: 164 spm. She reported feeling more supported (placebo). No detectable change in gait metrics. Verdict: perceived benefit without biomechanical change.

Limitations and honest caveats

Sample size (30) is modest — this is an exploratory consumer study, not a clinical trial.
Wrist‑based smartwatch metrics vary by vendor and firmware. We mitigated this by using within‑subject comparisons and controlled walk tests.
Single‑blind design reduces expectation bias but is not fully double‑blind in the field.

Actionable checklist: how to test a custom insole using your smartwatch

Do a baseline 6‑minute walk test indoors on a flat loop while wearing your everyday insoles; record cadence and steps with your watch.
Wear the new insole for 2–6 weeks during normal activities (avoid mixing with other new gear like shoes or orthotics at the same time).
Repeat the same 6‑minute walk test. Export data if possible and compare mean cadence and cadence variability (standard deviation or coefficient of variation) — export guidance and data-audit notes are useful (how to audit your tool stack).
If you have suspected gait issues, consider adding a foot pod/IMU insole for more sensitive detection (sensor & local inference notes).
If you see objective reductions in variability or consistent cadence changes and feel better, the insole is likely affecting your biomechanics in a meaningful way.

Final takeaways — trusted advisor summary

For most people, 3D‑scanned custom insoles feel nicer and improve comfort, but wrist‑measured cadence is unlikely to show large changes unless a person starts with a measurable gait problem. For people with gait irregularities or pain, custom insoles can reduce cadence variability and improve steadiness — and smartwatches, especially when paired with foot sensors, can document that improvement.

In 2026, the combination of more capable watches, affordable foot sensors, and improved on‑device AI means consumers have more powerful tools than ever to test claims themselves. But buyer caution is warranted: ask for data, test with your own devices, and prioritize objective results if your goal is injury prevention or gait correction.

Next steps (call to action)

Want to run this test yourself? Download our free insole testing checklist, export template, and step‑by‑step guide to compare pre/post cadence with your watch. Sign up at smartwatch.biz/insoles for the toolkit, and join our next crowdsourced trial — we'll pool anonymized results and publish community‑driven analyses in Q2 2026.

smartwatch

Contributor

Senior editor and content strategist. Writing about technology, design, and the future of digital media. Follow along for deep dives into the industry's moving parts.