Influencer Marketing Demand forecasting and Optimisation

Online commerce is growing rapidly, however the biggest challenge for online retailers is dwindling margins and conversions. In recent years, the industry has embraced influencer marketing as brands leverage trustworthy recommendations from social media influencers to drive better ROI and sales. The very nature of influencer marketing campaigns causes the demand for products to sour, resulting in out-of-stock situations. Most online retailers experience out of stock levels of around 10%, but during promotions the high demand for products can cause the out of stock levels to reach as high as 60%.

In the UK 83% of the eCommerce stores run influencer marketing campaigns but they face three key challenges i) tracking sales and performance ii) finding the right influencer for their campaign and iii) predicting the stock levels for the campaign to plan and replenish stock. We address the first two issues by enabling brands to connect with the right influencers, as well as providing analytics regarding performance and conversions through our existing platform. With this project, GAL aims to address the third problem of demand prediction. The problem will be resolved in two ways: first predict the likelihood of a product going out of stock based on its historic performance with the influencers in the campaign; second, if it still goes out of stock, suggest a similar or substitute product from the same retailer.

A new out-of-stock module will be built on our current Women in Innovation Award project by adding an innovative layer on top of our existing influencer marketing platform to enhance user experience and sales for online stores. It will be a completely new concept to recommend substitute products dynamically in real time on the product listing page.

GAL is a fast-growing machine learning startup focused on online retail. It has developed a sophisticated marketing platform for online retailers. Hundreds of global retail stores use the platform, which serves millions of users around the world. Areas of research and development include use of machine learning, advanced data processing and event processing which is at the leading edge of R&D in eCommerce systems. With its project, GAL aims to develop state-of-the- art machine learning processes to support eCommerce infrastructure. Through this technology development, online shoppers will be able to better manage their stock or discover similar or substitute products at the online retail sites.