How FLO Cut Customer Refund Time from Weeks to Seconds

It’s the 1960s, and you live in Turkey. You just bought a pair of beautiful shoes — the latest pair from one of the finest shoemakers in Turkey. But for some reason, the shoes are not quite right. Such things happen even with the best products.

A problem with one of the heels, you think. So, you return to the Gaziantep workshop, where you’re greeted by owner Ahmet Ziylan. He inspects the shoes and agrees there’s an issue. Then, he accepts your return and provides you with a refund, which you use to purchase another pair. The whole transaction takes only a short time. You leave happy with your replacement and with Ahmet’s service.

Fast forward to the 21^st century. That 50-square-meter workshop in Gaziantep has grown into an international retail business, FLO. Today, FLO is the largest footwear retailer in Turkey. Not only does the company continue to make high-quality shoes, but it also sells footwear from other manufacturers. Brands include Lumberjack, Polaris, and Kinetix, Turkey’s No. 1 sports shoe brand in terms of sales volume. FLO has more than 650 stores in 21 countries, producing around 3.5 million pairs of shoes and selling around 35 million pairs every year.

Ensuring happy returns

Product returns are still part of the business. With millions of shoes purchased from FLO stores each and every year, even the small percentage of returns represents a significant number — around 500,000 annually.

As the business grew, handling the returns became a challenge, affecting FLO’s customer satisfaction, operational efficiency, and bottom line. Until recently, the return process involved shoes being shipped from the stores to a FLO facility, where the company’s team of quality control experts inspected the shoes by hand. If a defect was noted, a customer was given a refund — but not immediately. 

Because of the return volume and centralized inspection, the wait time to get an answer about a refund was up to three weeks. Needless to say, this created some undelighted customers.

It became clearly evident to FLO that they needed to find a way to satisfy their customers and contain costs by providing its patrons with an instant solution at the stores in case of a defective product. But how?

The shoe (application) fits

To meet the challenge, the company took “a leap into the future,” turning to digital transformation and a machine learning (ML) solution. FLO created an application powered by SAP HANA Cloud Services in the SAP Business Technology Platform, leveraging the advantages of cloud computing, big data analytics, and real-time replications.

The application can predict if a returned item is defective by taking into account attributes such as suppliers, product family, customer behavior, and season. A defective product can be identified by ML in 80% of the cases. As a result, only complex issues are now handled by human inspectors in the centralized facility.

With the application predicting the inspection result at the time of return, customers have only to wait seconds, not weeks, at checkout for an answer, increasing their satisfaction and ongoing loyalty. The application also helps FLO avoid a related challenge of the past: products being accepted incorrectly.

After walking that proverbial mile in their customers’ shoes, FLO reduced human errors related to the return process by 35% — equal to €500,000 annually — and lowered inspection workload by 50%. At the same time, the footwear maker has been able to bring down costs such as cargo and SMS by 50% and increase brand value.

All those wins have given FLO a leg or foot up in the market, being even more successful and a customer favorite, as well as being named a Winner at the 2022 SAP Innovation Awards.

To get the technical details behind Flo’s amazing accomplishment, check out their Innovation Awards pitch deck.

Artificial Intelligence, Machine Learning