KI Handel
Insights

AI: A welcomed boost for retailers

By Melanie Tschugmall & Nadja Keidel &

The gradual decline of brick-and-mortar retail sales could mean fewer investments and innovations in future. This is where artificial intelligence (AI) can provide a welcomed boost and open up numerous opportunities for retailers.

Insight in brief

  • Thanks to the use of AI solutions, the retail industry is offered a multitude of new possibilities. 
  • More and more retailers are finding ways to make data available to customers - often in exchange for incentive schemes such as loyalty programs.
  • In the future, AI will play a prominent role in the form of diverse, innovative AI-supported hardware and software applications in the areas of logistics, store management and personal customer experiences.

It goes without saying that it’s tougher for brick-and-mortar retail stores to implement AI solutions than it is for e-commerce. Somewhat surprisingly, though, the use of AI is still in its infancy in both of these sectors.

AI solutions can offer a wealth of new possibilities for retailers – not just to improve the customer experience, but also to enhance operations’ efficiency and boost productivity. Marketing, sales and supply chain management are thus among the main areas where AI is being put to use. Whether the aim is to create targeted marketing campaigns, improve the customer journey or optimise inventory management and spatial planning, AI gives retailers a better understanding of shoppers and their behaviour so they can offer them unique and personalised experiences. In addition, brick-and-mortar retail and consumer goods companies can implement AI applications to improve the product range and inventory management for each store and continuously optimise their supply chains.

2 diagonalestriche lightgray

Personalised product recommendations

Customers want personalised products and services, which is why retailers should tailor their offerings to individuals based on their past purchases. And this is where AI’s strengths come to the fore. Many retailers view self-learning algorithms that can make accurate predictions by combining data from various sources as being the way forward. Amazon has been leading the way in this field: the online giant’s recommendation engine analyses users’ past purchases, items currently in their shopping basket, products they’ve reviewed and many other factors to find out what products are most relevant to them. This ability to gather and sort large volumes of structured or unstructured data and use it to increase relevance for the customer and thereby gain a decisive competitive advantage (a practice known as ‘hyper-personalisation’) wouldn’t be possible without AI.

Dynamic price optimisation and loyalty programmes

AI is a particularly hot topic when it comes to price, advertising and markdown optimisation. With price optimisation, there are two main objectives: firstly, to tailor prices to the customer, and secondly, to predict when a discount should (or shouldn’t) be offered. But despite the obvious benefits, retailers should carefully weigh up the pros and cons of introducing such a model. With dynamic price optimisation, prices rise and fall depending on the time of day. However, customers in brick-and-mortar retail are not as accustomed to dynamic pricing as in other areas (e.g. petrol stations or flight booking platforms), so caution is advised here. To avoid disgruntled customers, dynamic price changes must be made in a way that appears fair, not indiscriminate. In other words, customer-focused retailers shouldn’t just do what they can legally get away with but, rather, what makes sense and generates real added value for the customer.

More and more retailers are finding ways of collecting data from customers. They frequently involve offering something in return through incentive schemes such as loyalty programmes. The more information the customer provides voluntarily, the easier it is for retailers to increase loyalty through personalised marketing. But here, too, transparency is the watchword: retailers must be able to show exactly how the data will be used. By doing so, they create trust and increase acceptance of digital innovations. Nevertheless, retailers can only increase customer satisfaction and loyalty in the long run by offering real added value.

Graphic Element - Orange Z outline

Behavioural and geospatial analyses

Geospatial analysis gives retailers a better understanding of the relationship between their brick-and-mortar and online business. Deep learning programs can now analyse footfall, buying behaviour and customer preferences from footage recorded by security cameras already installed inside a shop. The insights gained from this analysis can then be used to help make decisions about store design or marketing measures.

At the same time, more retailers are testing direct interactions between customers and AI solutions inside the store itself. For example, a French cosmetics chain is using AI applications to advise customers when purchasing beauty products. A virtual makeup assistant allows the customer to try out thousands of lipsticks, eyeshadows, false eyelashes and many other similar items. The virtual assistant also offers interactive beauty tutorials that teach users how to create certain looks by practising on a digital version of their own face. Another new feature helps customers to find the right colour tone for their skin by uploading a selfie. Innovations like these are bridging the gap between in-store business and digital AI-based solutions.

Inventory management and optimisation

The aim of good inventory management is to match supply and demand, i.e. deliver products to the right place at the right time to meet customers’ needs and preferences. This is important because overstocking may result in additional markdowns, while understocking can lead to lost sales or disgruntled customers. Both of these scenarios can weigh heavily on inventory productivity and margins. Stock levels can be reduced if the retailer can accurately identify what is being sold, when and in which stores. Large brick-and-mortar retailers achieve this with in-store robots that use RFID technology to continuously track inventory levels and forward the information to a central system. Another option might be to digitally link the shelves themselves to count the number of items directly. This would also make dynamic ad-hoc replenishment easier.

The fashion industry is another area where optimum inventory management is very important. Here, AI is used to analyse cash receipts and returns to gain insight into purchases at individual stores. The algorithm helps the business to know which items to promote and which products to stock more of in certain locations. For example, the data shows that some items tend to sell more in stores in the city centre. By having this information available to them, companies can accurately match their inventory with customer demand.

The future of ai in retail

Use of AI in retail will continue to increase. In its study entitled ‘AI in Retail’, market research firm Juniper Research predicts that global spending on AI in the retail sector will reach USD 7.3 billion per year by 2022, compared to the estimate of USD 2 billion for 2018.4 In the increasingly digital world of retail, it is essential for companies to make the shopping experience as personalised as possible. The ability to do so will largely depend on the degree to which artificial intelligence is adopted. AI is therefore likely to play a prominent role in the future with innovative AI-assisted hardware and software applications for improving logistics processes, branch-specific procedures and personalised customer experiences. The latter in particular is a highly emotive area in which AI-based solutions can offer the customer real added value. AI can also enable brick-and-mortar retailers to scale up their core competencies and become more competitive.

Melanie Tschugmall Zühlke

Melanie Tschugmall

Business Development Manager
Contact person for Switzerland melanie.tschugmall@zuehlke.com +41 43 216 6414

Melanie Tschugmall joined Zühlke in 2016 and has a Master in Strategic Marketing with a focus on Innovation. Before joining Zühlke, she worked in different service companies. In order to stay ahead with new ideas and cross-industry impulses, Melanie is involved in various networks and continuous education, eg. Digital Ethics & Behavioral Economics. This makes her a creative and energetic sparring partner. Melanie is fascinated by digitalisation and continuously challenges status quo. 

Nadja Keidel is an expert in Data Science and works as Lead Consultant at Zühlke Engineering

Nadja Keidel

Lead Consultant
Contact person for Switzerland nadja.keidel@zuehlke.com +41 43 216 6852

Nadja Keidel joined Zühlke in February 2015 as a Data Scientist. She holds a MSc ETH in Mathematics and a Certificate of Advanced Studies ETH in Computer Science with a focus on Data Science. She is a strong analytical thinker and has several years of experience in conducting machine learning projects in all stages: from vision and scope shaping, evaluating and prototyping to operationalizing machine learning algorithms – always with a focus on real business value creation.