Guest article: At K-Ruoka online store, data serves to assist and delight the customer

Grocery shopping differs from most other purchasing experiences. Buying a new TV requires consideration, as it’s not exactly a weekly occurrence. Food, however, is a daily necessity. Customers at the K-Ruoka online store fill their carts with dozens of food and household items, and they typically don’t want to spend time pondering over bags of flour. Shoppers expect speed and ease, whether it's in-store or online. 

Developing an exceptional online customer experience begins with the mindful application of customer data. In this article, I will delve into what kinds of measures we've taken to enhance the customer experience and expand our online grocery store's customer base. 

Turning data into meaningful interaction 

Digital commerce provides many sources of customer data. Some of it comes directly from customers, like their names or food preferences. However, people rarely give answers to optional queries, which means that this type of information is usually fragmented and difficult to utilise as such. 

However, behavioural data, such as K-Plussa customers' purchase history, is much more detailed and generally available in large quantities. This kind of information is ideal for personalising the user experience. For example, we can display tailored product recommendations in the K-Ruoka app based on individual purchase data and buy rates. 

A big challenge in online grocery shopping is product substitution when an item in the customer’s order goes out of stock before the collection phase. We believe it is crucial to find the best replacement. With machine learning, we can identify interchangeable products and select the most suitable alternative for each customer. 

While behavioural data enables accurate recommendations for each customer, its detailed nature creates challenges for strategic analysis and decision-making. With highly granular data, it is possible to generate hundreds of thousands of scenarios, which makes it difficult to see the bigger picture. 

However, combining data from multiple sources allows us to identify and explore various customer segments. While segment data is based on generalities and thus less precise than individual-level data, it supports strategic thinking by illustrating how customer groups operate and identifying repeated behaviour patterns. These discussions benefit from utilising customer stories based on identified segments to visualise findings through a narrative form. 

Our considerations at K Group might include pondering what kind of content would resonate with a young, childless couple who enjoys convenience in cooking and rarely spends time browsing our recipes. This makes it easier to meet this group's needs and deliver more targeted messaging. 

The renaissance of customer profiles via large language models 

I believe that large language models will bring an age of renaissance to the use of customer profiles as internal tools, as models trained with natural language require data in a human-friendly form. We must be able to describe customers concisely to effectively use AI in tasks like creating personalised communications. 

AI can also fill in data gaps when the information doesn't need to be pinpoint accurate, but suitably detailed for the context of personalisation. For instance, most product information on bananas lacks an explicit vegan label. While obvious to humans, systems may not recognise it as such without a specific label. If we want to recommend only vegan items to vegan customers, we need comprehensive product data. 

On the other hand, language models can infer such self-evident facts and readily answer questions like whether or not a Pirkka banana is vegan. They also improve data quality, which in turn enables a better, more personalised customer experience. 

This simple action will surprise your customers 

Customer data doesn't have to stay within your organisation to provide tangible value. One surprisingly underutilised idea is feeding data back to the customer. Naturally, we have legislation in place that requires companies to provide customers with access to their data, but why stop at the bare minimum? 

The most useful data often reflects customer behaviour. Reviewing one's habits and preferences can be rewarding and fun. With the Ruokavuotesi food year summary feature in the K-ruoka app, customers receive a personalised story based on their grocery purchases over the past year. 

Rather than just listing the top products, we chose to highlight unique, socially shareable insights from each customer’s shopping basket. In 2022, the Ruokavuotesi summary increased new app downloads by 800% – and brought home a victory at the 2023 Grand One gala. 

These ideas can also be combined with third-party expert knowledge. Our K-Hyvinvointi service, centered on wellbeing, became another Grand One award winner in 2024. The service compares purchases made by our K-Plussa customers with expert nutritional guidelines and helps our customers make better choices, such as increasing their vegetable or fibre intake. 

At K Group, data literacy creates real customer value 

The K Group has a long history of using data to enhance customer experiences in both digital services and in-store locations. For instance, our beloved OmaPlussa benefits – based on individual shopping histories – have been around in some form for over ten years. 

As opportunities to personalise digital services grow, so does the utilisation of data, both in terms of volume and versatility. This is driven by our company culture that values customer-centricity, hands-on action, and a shared commitment to building better retail. 

In my view, good data literacy across the organisation is key to turning data into meaningful customer benefits. At K Group, data isn’t just for specialists to lord over. Employees can explore data widely via shared reporting tools – within a framework of carefully defined access rights and privacy guidelines. 

Responsible use of data is a cornerstone of our operations: customer data is always brought to an appropriate level. This helps build a culture where we can generate real customer value from data – in an agile and meaningful way.