Shopping for clothes online has skyrocketed in recent years. A third of all clothing purchases now take place online, representing a global worth of around $403 billion. However, 41% of customers order multiple sizes and send back the ones that don’t fit. Ultimately up to 70% of all purchases are returned.
It’s annoying for customers and costly for retailers, and there’s also an environmental impact from transporting all those unwanted clothes back and forth.
Shavatar could see the huge potential in helping not only customers get their right fit from the get-go, but also retailers by looking after their bottom line.
The tech start-up Shavatar (a portmanteau of Shape and Avatar), came to us with a statistical shape model which could create an accurate 3D avatar with just three measurements and a handful of other metrics. The user can then get sizing advice for over 50 popular brands. However, it wasn’t yet accessible by anyone except its creators.
Shavatar challenged us to take this model and bring it to the cloud as a minimum viable product so that users could access it.
Their product needed to be built with flexibility in mind to be future-proof for their B2B iterations, and with an eye on infrastructure to keep the start-up budget in line.
We first needed to test the viability of bringing the model into the cloud for users to access. We conducted a ‘spike’ to explore whether it was technically possible to run the model and check feasibility in a short time frame.
As part of this, we started rudimentary development to check for performance and scalability.
Once feasibility was established, we introduced our functional workshopping process to shape the user story backlog. All the functionalities were defined, allowing us to sketch the User Experience (UX) and create the User Interface (UI) designs; laying out exactly how each page would lead to the creation of a 3D avatar.
Shavatar’s MVP communicated the user’s request via an API, tapped into the statistical shape model in order to tell it to turn the user’s measurements into a 3D avatar.
Traditionally this would all have been hosted on a dedicated server with starting and infrastructure costs of around $1000.
As part of our efforts to optimise the budget for this start-up, we used serverless technology to eliminate the infrastructure costs and bring the model and user profiles to the cloud for a price up to 10 times lower.
As serverless only charges for the computing power you use, starting costs are low. The flexibility of serverless infrastructure also gives us the ability to expand capacity alongside Shavatar’s business needs as they develop.
Using our scrum methodology and weekly check-ins with Shavatar, it was possible to deliver a working MVP in just five weeks. This speed is enabled by November Five’s all in-house expertise, from defining the MVP right through to build and delivery.
Eight weeks after launch Shavatar has already attracted nearly 3000 registered users who can return each time they need a fitting and created a total of 11.000 shavatars.
We were able to optimise the loading time of the creation of the avatar to just three seconds. Serverless technology enabled us to do all this for up to 10 times less cost than a traditional server.
Shavatar is now looking towards their next B2B phase using the feedback generated from users. Eventually, they will be able to integrate the model into webshops so that customers can check their sizes live while shopping.