How Machine Learning is Helping a Middle Eastern Giant Protect Its Market Share from Amazon?

Overview of the E-Commerce Market in the Middle East (ME):

The ME is one the fastest growing e-commerce markets in the world, growing from USD 26.9bn in 2018 to USD 48.6bn by 2022 [1]. This growth in online sales is driven by the growth in the smartphone penetration; for instance, 88% of the UAE’s population are internet users with smartphone penetration of 78% [2].
However, only 15% of the businesses in the ME have online presence, and 90% of online shopping import products from outside the region [2]. This gap between supply and demand presents a significant opportunity for online retailers to take-over existing “Brick and Mortar” businesses.
Focusing on the retail market, Carrefour has captured a significant market share in the ME, operating in 38 countries and serving more than 200,000 customers per day [3]. However, in 2017, Amazon announced its acquisition of (biggest online marketplace in ME), which marks Amazon’s first move into serving the large untapped ME region [4]. This move presents a significant risk not only to Carrefour, but also to all Brick-and-Mortar retailers in the ME.

Carrefour’s Plan to Protect its Market Positioning:

Majid Al Futtaim CEO, Alain Bejjani, had announced in early 2017 his plan to transform the shopping experience in the ME by leveraging data analytics and machine learning. As a first step, the organization hired 200 employees from Silicon Valley as part of its ambitious strategy to expand its digital and e-commerce operations [5].
In my view, machine learning will help Carrefour protect its market positioning through three levers:
1. Providing product recommendations: As an online retailer, Carrefour can gather customer data and provide personalized recommendations, which enhances the shopping experience and increases cross-selling. This can be achieved by linking customer IDs to historical purchases and assessing the basket of SKUs. By leveraging machine learning, Carrefour can rely on historical data to understand customer behavior and preferences, and based on look-a-like modeling, predict what the customer “will need”.
2. Optimizing promo-effectiveness: By clustering customers and identifying their interests and needs, Carrefour can offer more targeted adds and increase its promo-effectiveness. In addition, it leverages machine learning to predict the effectiveness of future campaigns and the key features that contributed to this effectiveness.
3. Enabling effective business decisions across functions: Machine learning in combination with visualization dashboards offer managers across business units the needed data to make close to real-time data driven decisions. This will help Carrefour not only improve customer engagement and retention, but also enhance operational efficiency.
In an interview with The National [5], Mr. Bejjani highlighted how Machine Learning and Big Data has helped them optimize their assortment: "We discovered that there was one specific biscuit brand that actually had a big impact of people walking out of the store if they don’t find (that brand)” [5].
In addition, Mr. Bejjani launched the first School of Analytics and Technology in the ME [5]. The school aims at covering a wide range of digital and analytics lessons to employees across the organization (front-line to senior executives). These courses will help educate employees on the importance of collecting data (relevant to front-line) and the benefits of leveraging the data to drive decisions across different functions. In addition, the school aims at developing a culture in the organization that enforces their position as a digital retail player.

What’s Next?

With this, Carrefour ME is building their advanced analytics unit in the short / medium term to protect their market from Amazon. However, these initiatives, though crucial, are not enough to protect them in the long-term.
Carrefour should think of effective ways in transforming the shopping experience across its channels by considering innovative initiatives such as:
1. Ultra-fast Delivery: The norm is 2-day delivery nowadays. However, Amazon now is talking about a 30 min drone delivery. Pioneering such a technology could position Carrefour in the forefront of innovation and technology. This will also impact the customer journey in the store, as Carrefour can eliminate the hassle of juggling shopping bags and send the goods to the customers home immediately.
2. Robot Customer Service: Robots offer an innovative form of self-service that customers might appreciate. Robots can speak multiple languages and help customers find and learn about items [6]. The use of robots can vary from robot security guards all the way to robot cashiers. A significant number of start-ups are spending significant time on how robotics can disrupt the retail market, and Carrefour can partner with one of these start-ups.

1. How can Carrefour further differentiate its product offering to compete with Amazon?
2. What will the role of the retail store be once digital channels take over?

[1] The National,, accessed November 2018
[2] The Next Web,, accessed November 2018
[3] Carrefour, “About Majid Al Futtaim Retail”,, accessed November 2018
[4] Tech Crunch,, accessed November 2018
[5] The National,, accessed November 2018
[6] Forbes,, accessed November 2018




ADIDAS – Additive Manufacturing for the Masses


Can machine learning provide a solution to CNP frauds?

Student comments on How Machine Learning is Helping a Middle Eastern Giant Protect Its Market Share from Amazon?

  1. Interesting read after the Alibaba case! I don’t believe e-commerce will totally eliminate the offline channel. There will always be the desire to inspect certain products before a purchase, especially with higher ticket items. Retail stores do need to invest in a better and augmented in-person experiential experience. How Amazon is doing so with their technology-enabled human free checking system in Seattle is a great example.

    From a competitive perspective, I do think Carrefour needs to create a better “ordering experience” (one of the advantages Alibaba gained in the Chinese market), for the local community. Amazon is too well-funded and because of the acquisition of a local player, less likely to make errors similar to eBay when they entered a foreign economy. Building trust as one of the local companies can perhaps help in the overall MENA region.

  2. Jad, it’s interesting to read about the digitalization change in ME! I remember when we held a E-commerce workshop in China in 2016, our Dubai manager shared that the on-line business was not yet there at the time. Thoughts about how to compete with Amazon: I think being local would be a great competitive advantage. That means finding out where ME market is different from other global markets, which could be consumer’s trust level, preference on delivery speed, influence of purchase decision, etc. Then Carrefour should focus on those areas. Amazon is a great global business and no doubt it has power of using big data, but localization is a different skillset. For example, during the Alibaba case we mentioned ebay but not touched on Amazon. Amazon China offers similar on-line purchase service as it’s US biz does, but the service was not preferred by the Chinese consumers. In my view, an important reason that JD got it’s advantage position over Amazon in B2C area is its decision to strategically focus on logistics and make claim on “delivery the next day”, which meets the needs of the consumers.

  3. Thank you for your local insight Jad, I agree that the prospect for e-commerce grocery in the Middle-East is immense. While Amazon did a great strategic acquisition with the local e-commerce player, I still believe Carrefour has a key competitive advantadge that has not been fully realized yet – it has years of experience in personal face-to-face customer interactions which lead to building the much needed trust in grocery shopping. In my mind the challenge and chance for Carrefour going forward is to find ways to bring that warm, human interaction element online, so that their service is clearly differentiated from Amazon’s.

Leave a comment