“Use as easy tools as you can” – ICA Gruppen has a shared data toolbox but distributed project ownership

ICA Gruppen’s operating businesses range from grocery retail to pharmacy and from financial services to real estate. How does their group-level data team support the business lines with their varying needs while the cross-company functions – such as the loyalty program – play a remarkable role?

“We have quite a wide toolbox,” says Olof Granberg, Director of Data and Advanced Analytics Technology at ICA Gruppen. “And we have to be smart on what tools to use. In some cases, we get a request for a new machine learning model to do something – say, to predict the assortment. You may really need a new machine learning model to do that. But sometimes, it’s just about getting the right data in the right place and having a dashboard with a good user experience rather than building a new AI model. It’s always good to use the easiest solution possible to each problem.”

The abundant but sensitive customer data

“If you look at ICA from the outside, it would be extremely easy just to say ‘let’s just pull all of the data from all of the companies into one place, and form a picture of the customer,” Granberg illustrates.

But there are weighty reasons for a more granular approach. Firstly, the data collected from different kinds of business – like banking activities and grocery shopping – is not always consistent. Secondly, the business lines may operate under different regulations, regarding privacy and data protection. And last but not least: all data use activities, and their benefits must be crystal clear to consumers. After all, ICA Gruppen’s vision is to make every day a little easier.

“The technical side of pulling data from several sources is not the biggest challenge these days. It’s more about how we work with the data.”

On a journey to make data governance a natural part of the business

There’s one thing over the others that Granberg emphasized when he visited the Data Insiders podcast: “Even though you centralize a data team and resources, you cannot isolate that from the business. The business use cases need to be run by the business, where also the delivery side is.”

Granberg also shared his tips on linking data government and business development together. In a nutshell, the steps are as follows:

1. Map out the whole data management organization and processes. Who are the information owners? Who works actively with data now?
2. Create a company-level data strategy. How do you want to work with the whole lifecycle of data? Expand the view from the mere analytics use.
3. Avoid the trap of putting everything on IT – start pushing data management into the business.
4. When you look for new data owners, make sure to find the people who have the time to dig into numbers and use cases. They need to be on a high enough level to make decisions – but they also need sufficient hours for their new role.

ICA Gruppen is on the journey right now, bringing data closer to businesses. “Data governance isn’t a side quest. It should be part of natural business development,” says Granberg.

Olof Granberg works as Director of Data and Advanced Analytics Technology at ICA Gruppen AB, a leading retail company in the Nordic market. The Group includes ICA Sweden and Rimi Baltic, which mainly conduct grocery retail, ICA Real Estate, which owns and manages properties, ICA Bank and Apotek Hjärtat.