Case Study: Super Socks

  • AI & machine learning
  • Applications
  • Data integration
  • Data visualisation
  • Digital transformation
  • Marketing campaigns
  • Workflow automation

The Challenge

Supersocks are a Birmingham based retail start up offering a unique line of quirky garments. Their assortment of clothing and novelty gift products are fully customisable, featuring images and colours selected by the buyer. A growing number of retail businesses are becoming digitally native, making it more important than ever to remain competitive. Supersocks were looking for a way to distinguish themselves in the market by meeting consumers’ digital expectations with a memorable and enjoyable user experience.

With a deadline approaching, the company challenged Shoothill to develop a software platform allowing their products to be personalised upon purchase in a live retail environment. This solution would greatly speed up the purchasing process, with tools to edit images on site, all accessed by consumers through an intuitive and easy to use front-end UI. Supersocks also required an integrated payment gateway and back-end customer databasing.

Shoothill saw an opportunity to take this project to the next level using AI. Our customer’s were excited to see what we came up with to transform their customer interaction. Brimming with ideas, we began developing this sophisticated system in a rapidly shrinking window of time.

“It’s great to work with a company that not only does good work, but actually cares.”

Charlie Rodman

Managing director at Nerdy Banana

Our Solution

Given our expertise in developing intelligent applications and leveraging disruptive technologies, Supersocks were interested in any informed suggestions we could make to modify the project. Workshopping with the company, we determined that an out of the box implementation of AI (Artificial Intelligence) would offer an innovative solution to accelerate the sales process.

Meeting the challenge, we delivered all of the software required to operate the touch-screen point-of-sale kiosks which Supersocks would be installing in their storefronts. By keeping to the deadline and completing the project within six weeks, the system was ready for deployment in time for Supersocks’ scheduled appearance in Selfridges department stores, in locations across the country.

Accessed on these POS kiosks, the system guides a user through the entire personalisation process. The customer first selects their choice of template from the wide array, before adjusting the design to their taste. Their images then need to be provided, either by being uploaded from an external device, such as a smartphone, or taken by the cameras mounted on the kiosks. This picture can then be edited on the software.

Shoothill have a proven track record with AI and we recognised here an opportunity for its application in this procedure. Buyers can opt to trace the desired region of the image with their finger on the touch screen, and apply that selection to their item, as per Supersocks original specification. However, we developed an AI alternative which automates this step. The tool automatically identifies and isolates the faces of both people and pets in seconds, allowing for quick and effortless image manipulation.

This functionality relied on a machine learning model. Machine learning is a branch of cognitive technologies in which the AI ingests sample data in order to construct mathematical algorithms which inform its behaviour. Shoothill developers built the AI model from the ground up in Python, and ‘trained’ it to detect and crop faces by feeding it a library of demonstrative images. The neural network, the core structure of the AI which takes the form of an immensely complex flowchart modelled after the layout of neurons in a human mind, was designed to interpret this data and ‘learn’ what faces look like by recognising and modelling patterns.

Throughout development on this project, user experience and customer satisfaction were held paramount. To this end our design team ensured the software had a simple but effective user interface, function married with visual appeal. This UI was then coded with Typescript, React, and MobX for sleek and fluid performance.

Using Microsoft’s Azure platform, we hosted this application on the cloud. As well as the diverse selection of high-powered tools available to us as Silver Partner’s we as a company have created a portfolio of tool, functions and routines we can access and apply swiftly to complex projects. Cloud computing in this project allowed us to directly employ our wealth of experience, whilst also providing a means to make the software and data available to Supersocks remotely.

Once the product is fully designed and personalised, the user can interact with a payment gateway to finalise the purchase. Both payment and invoicing are handled directly through the kiosks as a result of an API integration with Shopify, the third-party e-commerce platform Supersocks use to sell their products. Once the payment is processed, the socks are printed on site.

Data manipulation was handled using SQL servers, providing Supersocks with data feedback from the POS units for analytic, CRM, and marketing purposes. This information is uploaded directly to their customer database. In its entirety the application’s functionality was provided by back-end software written in C# using the latest .NET Core version.

The final product was delivered on budget and on time, to achieve its full impact. Enabling full personalisation of socks in a live environment by condensing the process into an incredibly narrow timescale was made possible through the pioneering and innovative use of new technologies. The unique nature of the sales platform and the software driving it captures its audience’s imagination, serving as a USP to increase catchment and boost brand value.

Conclusion

This digital solution was incredibly successful, performing flawlessly from its launch, which allowed Supersocks to seamlessly open and run their point of sale kiosks on time. The result was a boom in business with their presence in a face to face capacity driving conversions. The service this software enabled proved very popular with consumers, receiving a great many positive reviews and reactions. Supersocks have received a lot of exposure in the time since, further exhibiting the touchscreens and their functionality. The machine learning tool works smoothly and accurately, leaving many unaware they are even interacting with an artificial intelligence.

Supersocks were delighted with the project, which fully met and exceeded their expectations. The addition of the contemporary use of AI wowed them and the software operated at a quality and speed beyond what they had hoped. Implementing the solution to their business has assisted them in continuing along their steep upward trajectory, marking them as a digitally native and unique business in their competitive market.

Having had previous difficulties with other software development firms attempting to achieve this result, the company were looking for a reliable business to work with. Charlie Rodman, the managing director of Nerdy Banana, parent company of Supersocks and a number of similar brands, was both exhilarated to see the full potential of the application including its cognitive ability, and tremendously relived that at the swift deployment saw it in action after such a short period. Mr Rodman commented that he had “not worked before with a company that doesn’t just do good work, but actually cares”

Following the conclusion of this project, Shoothill are still in business with Nerdy Banana. As well as migrating the app to the mobile platform, with the personalisation tools and shopify integration available to users on demand, our two business are continuing to innovate together, searching for new and inspired applications of technology to their novel enterprise. At Shoothill we believe in empowering people to do their best work, so we always like to see our customers businesses doing well. We expect more great things, and further pioneering projects, to be borne from this partnership.

This project was complex and challenging. A unique and enjoyable UX, swiftly guiding consumers through the personalisation, image manipulation and payment process, completed on a tight timeline. We are especially proud of how Artificial Intelligence was used in the software, breaking new ground with the unprecedented tool. Especially given the narrow deadline, we take pride that we were able to build our own cognitive system from the ground up, utilising cutting-edge machine learning skills to develop this state-of-the-art solution. We look forward to discovering new and innovative uses of AI technology in future projects.


Other projects you may be interested in