Where does your return go? A dusty warehouse? Your neighbor\u2019s doorstep? Refurbishment center? More likely – to waste.<\/p>\n
5 billion pounds of returned items end up in a landfill every year, according to Optoro, a company specializing in technology, analysis, and distribution strategies for major retailers to manage returns.2<\/sup> In addition to the environmental consequences of excess returns, retailer bottom lines are also suffering due to the inefficiencies and costs associated.<\/p>\n With the fast rise of e-commerce, retailers rose to the occasion to meet the growing demand of customers, but secondary processes, such as returns have been handled inconsistently and inefficiently across retailers. Further compounding the complexities of e-commerce returns, return rates can be as much as four times higher than traditional retail return rates and often require additional transport for evaluation and re-distribution.3<\/sup> Handling returns is a complex and nuanced process, factors such as product type, current location, weight, damage, and market demand, factor into each product\u2019s value and optimal secondary distribution channel. Returns logistics prove costly and there is a delicate balance in optimizing for greatest recovery price of a good and the cost to re-distribute that returned good.<\/p>\n Retailers across the country have started using Optoro\u2019s predictive analytics solution and distribution services to create value and savings from their excess and returned inventory and turning their inventory into assets and focusing on the recovery of costs.3<\/sup><\/p>\n Why Machine Learning<\/strong><\/p>\n Optoro uses machine learning to create data-based recommendations in a space where spreadsheets and email dominate internal distribution and external secondary seller channels are difficult to evaluate and compare on a per-product basis.3<\/sup> The speed and accuracy of these recommendations are especially valuable for products that lose value quickly, such as new technologies.2<\/sup><\/p>\n Optoro\u2019s algorithm is looking to predict which distribution channel will recover the most cost of a given returned good for a given retailer. Optoro\u2019s machine learning searches for an informed recommendation based their wide data collection of large proprietary data-sets and real-time secondary market information.4<\/sup> The optimal recommendation will be created based on the reinforcement of previous patterns and trends from the data relevant to the specific returned product.<\/p>\n Machine learning is particularly suitable due to the availability of cross-validation in continuing to hone the algorithm to detect further optimizations in the recommendation of a returned good\u2019s path. This is possible as there are quantifiable outcomes, such as the dollars recovered or the time to return to the market, that provide a basis to compare and improve the algorithm.4<\/sup> These algorithms not only enable the organization to select the optimal re-distribution channel, but they also enable Optoro to provide retailers with robust reporting on product performance and inventory management that can be used to optimize inventory planning, warehouse design, and other functions.3<\/sup><\/p>\n Current Plans for the Future<\/strong><\/p>\n With over $380 billion in merchandise returned annually, Optoro raised $75 million in new funding July 2018 in order to further grow the organization to meet the growing demands of the marketplace.5 ,6<\/sup> With this funding, Optoro plans accelerate the expansion of their recommendation solution, partner with the biggest retail companies in the world, and grow their team to fulfill demand.5<\/sup> This funding is expected to continue to bring an innovative approach to distribution and liquidation of returned goods, while leveraging marketplaces and channels to sell goods across digital platforms ranging from Amazon, eBay, and their own proprietary secondary market sites.7<\/sup><\/p>\n In this last round of funding, Series Funding E, Optoro has accepted additional funding from the United Parcel Service (UPS). Further solidifying a relationship founded on principles of reducing waste and creating efficiencies, such as label creation and other logistics.7<\/sup><\/p>\n Recommendations for the Future<\/strong><\/p>\n Optoro\u2019s success hinges on its ability to deliver value to its retailer customers, by recovering as much as possible. Optoro\u2019s prioritization of expanding to new markets and customers is aligned with further optimizing the recommendation engine and predictive analytics that fuel their bottom line. The more data available for the machine learning reinforcement algorithm, the more weight and confidence can be placed in the recommendations of the distribution recommendation. With more data, Optoro will be able to contextual patterns in a way that may unveil seemingly idiosyncratic data in one context to be an example of a larger trend within other data sets or retailers with similar properties.\u00a0 In the near-term, Optoro should continue expanding prediction abilities to new retailers and in the long-term evaluate ways to insert themselves into the supply chain to reduce chances of substitute products.<\/p>\n Open Questions on Optoro\u2019s Decisions and Future<\/strong><\/p>\n How do you evaluate the benefits of reducing waste with the excess cost (monetary and environmental) of transportation for a given good? Is it worth reducing savings to reduce waste in certain cases?<\/p>\n Is the UPS relationship beneficial or harmful in the near-term and long-term?<\/p>\n <\/p>\n The intersection of big-box retailers and saving the planet – Optoro uses machine learning to optimize return logistics both for organizations' bottom line and the environment.<\/p>\n","protected":false},"author":11331,"featured_media":33676,"comment_status":"open","ping_status":"closed","template":"","categories":[826,346],"class_list":["post-33630","hck-submission","type-hck-submission","status-publish","has-post-thumbnail","hentry","category-big-box-retailers","category-machine-learning","hck-taxonomy-organization-optoro","hck-taxonomy-industry-retail","hck-taxonomy-country-united-states"],"connected_submission_link":"https:\/\/d3.harvard.edu\/platform-rctom\/assignment\/rc-tom-challenge-2018\/","yoast_head":"\n\n