/ future of retail

Leading technology for leading results

Massive impact of AI for retail

The importance of AI in marketing has grown immensely in recent years:

  • according to Forbes Magazine, 86% of marketing leaders believe that artificial intelligence will revolutionize marketing by 2020.
  • McKinsey recently analyzed the potential impact of AI per industry. The estimated aggregated US-dollar impact of AI is the largest for the retail industry, up to $777 billion, of which $470 billion is generated by marketing and sales solutions – mainly in the business area of pricing and promotion with $210 billion.

One of the biggest challenges remains targeting the right consumers for ads and promotions and identifying the right discount level to induce a consumer to buy. This is especially true for grocery and other FMCG retailers – an industry which is very much based on traditions and established processes and ideas.

Combining substantial internal research with external academic cooperation for a next-level AI

SO1 (Segment of One) is a driving force in the sector of pricing and promotion with fresh and revolutionary ideas on how to efficiently target consumers.


We combine internal and external research for the best possible results. This means we put great effort into hiring talent from all around the globe, and enable them to cooperate closely with top-tier universities on machine-learning research projects – to stay close to academic research and their latest findings. This continuous learning approach accelerates our success and forms the foundation of our shared knowledge within the company.

Let's have a look at some examples. The first three describe our latest external academic cooperation projects, the fourth an internal research project.

(1) In our joint work with researchers from the Massachusetts Institute of Technology we are developing scalable deep-neural networks that use rich market basket data and loyalty card data to predict consumers' purchases across all categories in a retailer's product assortment.

(2) The collaboration with the Humboldt University Berlin has the goal of improving coupon recommender systems in grocery retail with the help of deep reinforcement learning.

(3) Research projects with ETH Zurich focus on the design of loyalty reward programs and their link to price promotions in grocery retail.

(4) Along with external research projects, we also tackle specific retail and AI challenges, such as “learning non-decomposable objectives”, internally within our R&D team. Let's look at this example in more detail: There are various KPIs that retailers are interested in, such as redemption rate, promotion feed diversity or revenue uplift. But these KPI's are non-trivial (i.e. difficult) to optimize. One possible way of tackling the problem of optimizing multiple KPIs is to try to improve the target objective indirectly. For example, by optimizing cross-entropy in the case of highly imbalanced data and then choosing a bias with cross validation. This is a standard method to achieve a better recall. The approach that we take is to create a surrogate objective that bounds the original one and makes it possible to solve the optimization problem with stochastic gradient descent. This makes the training of our recommendation engine scalable to millions of customers and helps us to outperform standard methods.


Knowledge is key to always staying one step ahead of the competition

At SO1, we channel all internal and external research to focus on building the best version of our product every day. Our artificial intelligence has been applied by leading solution providers and retailers in the CPG and grocery retail industry in the US and in Europe. These partnerships have shown that AI targeting leads to significant positive financial impact for retailers while at the same time improving customer satisfaction and redemption rates.

During our most recent project, SO1 competed against rule-based targeting using demographics and other metadata, and outperformed it by an average of 90% with regard to coupon redemption rates. Additionally, retailers can save up to 60% on the budgets that are spent on discounts when these are allocated to the right consumers.

This is a huge step towards a new and improved version of CPG retail marketing and will allow retailers to realize the AI dollar impact predicted by McKinsey. This is the future of retail.

PS: SO1 (Segment of One) is one of the driving forces of retailer digitalization. We have created a very powerful AI for retail which is capable of personalizing promotions for users in real-time and across devices. The SO1 Engine (combining targeting, recommendation, and optimization features) sources the entire portfolio of the retailer and automatically selects the right products for each individual consumer and adjusts discounts such that revenue, profit, or consumer satisfaction are maximized. To learn more, reach out to:
US Sales: Patricia A. Cucinelli, cucinelli@so1.ai, +1 917 757 6221
EMEA & APAC Sales: Stephan Visarius, visarius@so1.ai, +49 160 93 59 69 95

SO1 is a registered trademark of SO1 Incorporated.

Personalization | Future of Retail | Precision Marketing | AdTech | Marketing Automation | Marketing Science | Hyperpersonalization | Recommendation Engine | 1:1 Marketing | one-to-one Marketing | Retail AI | Retail Technology | Personalization Technology | Personalization Solutions | Coupon Optimization | Offer Optimization | Promotion Optimization


Stephan Visarius

Stephan Visarius

Heading the Customer Acquisition & Success team at SO1, I take care of sales, biz dev & marketing. Before I worked for PAYBACK (sales to FMCG brands) & MARS (Ice Cream Trade Mktg & strategic projects)

Read More