We’ve been looking at how we handle our data. Over the years we’ve had recommendations from both in-house and consultancy teams about how to do this, but now we want to break the cycle and finish what we started.
Above all we’ve been thinking about how we can take a more ‘Co-op’ approach to our data. We pulled together a multidisciplinary team from across the business to look into this and they’ve become know internally as the ‘data layer’ team (explained in more detail by Rob in his being trusted with data post).
So, where to start?
We want to be trusted with data, and use data to inform what we do. The purpose of the discovery was to explore how we should go about creating the right conditions, both online and offline, to support this; and where to start.
We wanted to understand:
- How we deal with data now.
- The options for doing things differently.
- What we should test, explore and do with the data we have.
- What we can, and should, do with data in the longer term.
The purpose of this discovery was to stay at a high level when thinking about data rather than focusing on any one area in particular. We knew that would come later so didn’t want to get into specifics at this stage.
What we did
We needed to understand the problems and opportunities we faced: not just in the Co-op, but in the wider world too. This meant referring to a mix of research, from reading previous data reports to holding interviews and workshops.
The data science team
We started by speaking to the data science team here at the Co-op to get an idea of the challenges the business faces when it comes to data. They identified 47 points that needed to be addressed. We discussed each one and prioritised them in terms of urgency.
Inside the Co-op
We also spoke to different Co-op groups and the Executive team. We wanted to understand how people around the business approach data, how they’re working with it and how we can meet our Co-op ambition of being more trusted with data.
People outside the Co-op
To get a better external perspective we met 4 private companies that work closely with data. We found examples of how easily accessible data can improve the way a business runs, and creates an environment for identifying new product and service opportunities.
Speaking to the public gave the team a different view of data and consent. We like to speak to our members regularly about data in order to be better informed. Understandably people feel strongly about how their personal information is used, and approaches change depending on how this usage is explained and how customers feel about the business in question.
Where we ended up
Three main themes came out of our research:
1. We have multiple versions of the truth
Teams work from different databases that don’t necessarily stay in sync, or use consistent definitions. This makes it hard for users (colleagues) to find the best source for information, and be sure they are interpreting data correctly.
2. We under-leverage our data for analysis and insight purposes
Individual teams own lots of data and use it only within their team. We could be much better at sharing data and insights across the Group so it could be helpful for everyone. The problem is partly down to technical constraints, and partly a reflection of our how widespread we are and how differently we work: even within the Digital Group itself.
3. Consent and preferences aren’t understood holistically
Customers opt in to marketing for individual parts of the business, but we don’t have a clear central understanding of the consent an individual has granted across the group. This means that we may not be giving the best customer service that we could.
None of these themes came as a surprise – but the value in the work came from prioritising how we should move forward.
We identified 4 alphas which will be the first step in our vision for a Co-op that is built on, and led by, relevant, transparent and trusted data.
First we need stronger data governance across the whole of the Co-op group, so that our data can become more consistent, more joined up, and easier to find and understand.
Secondly, we should have a single view of each individual so that all of our information on members and colleagues matches up. We’ll look at membership to help with this.
Thirdly we want to look at what we mean by ‘consent’ when we ask members to trust us with their data, and we want to be sure this meaning is exactly the same across all Co-op businesses: from Food to Funeralcare.
Finally, we should develop a data science lab so that our data scientists and analysts have the right environment, tools and processes to develop and improve the work they are doing. This will also help us to attract more of the kind of people that will make the Co-op a centre of excellence.
These 4 alphas will start from September, and we will share our progress as we go. It’d be great to hear your thoughts.
Interaction / service designer
Head of Data Engineering