Working with the third largest global retailer, Lidl, we built a custom machine learning algorithm to deliver the first virtual workwear fitting app for their workforce.
Understanding The Problem
Having over 300,000 employees with a 150% turnover rate presents its own challenges. One of which is that all new staff members had to report to an off-site location to try on new uniforms at everyone’s inconvenience. Lidl wanted to create a better first impression with their new staff and automate the hassle of this process.
We were requested to make a virtual fitting workwear app for Lidl employees with a 90% fitting accuracy. Traditionally people need to measure themselves and input all their body data points to achieve this accuracy. This user experience is simply not acceptable, so we had to reimagine this online fitting method.
We faced a common trade off between data collection methods & UX, where the more data the better but less is more with users. UX should always take priority, otherwise people simply give up, leaving you with no data for your algorithm.
What we did
We built a workwear fitting app that took under 1 minute to order a new uniform with a 90% accuracy! This was a very user friendly process where employees selected their basic biometric data, body shapes and we created a one of a kind feature that asked ‘what brand size currently fits you well?’.
To create this feature and remove this major pain point of measuring yourself. We conducted an initial ‘pilot test’ to gather sample data and downloaded sizing charts from all fashion brands to train our machine learning algorithm. This feature streamlined the UX to reduce the number of steps to make the process of ordering workwear much easier & faster for Lidl staff.
Our Design Process
Feel & Look
Body data is a sensitive area that we addressed with our UI design. We employed the use of body animations to lighten the process. We decided to go with just three body shape options to reduce body categorization of people with extreme sizes.