How Coop delivers for its customers using AI


Learn how Coop, a large retail organization in Denmark, augmented their customer service offering with artificial intelligence (AI) with our use case.



  • Challenge to maintain customer satisfaction at scale

  • Hiring more and more staff to meet demand

  • Chatbots were too much work


  • Implemented raffle Web on Coop Shopping and Members' section

  • Generates own answers and re-learns

  • 24/7 service


  • Phenomenal self-serve rate

  • 60% of answers given "thumbs up"

  • More time for support team to answer more complex questions.

How does a large supermarket chain scale up customer satisfaction?

As the biggest retailer of consumer goods in Denmark, with around 1,100 supermarkets, discount stores and hypermarkets in the country and a market share of 40%, Coop has a large challenge to maintain customer satisfaction at scale.

In this case study you’ll learn how a large organization augmented their customer service offering with artificial intelligence to face this challenge.

You’ll also learn how Coop Denmark not only improved the quality of their service, but also their bottom line, through raffle Web, a state-of-the-art AI first line support solution.

Efficiency challenges

Having a large amount of customers means, of course, that Coop receives a large amount of customer requests. This, in turn, puts pressure on their customer service team to resolve their requests speedily and effectively.

And when your customer service team is under more pressure, your first instinct is to give them more help. Traditionally, this means recruiting more people.

Palle Esbensen, Head of eLogitistics and Development at Coop, spoke about raffle at the ComputerWorld conference:

“We were hiring more and more people in customer support. They came in every day from the customer center and said we need to bring in more agents.”

With a glint in his eyes, he said that if they continued along this path, eventually there’d be more people in Coop’s customer service team, than in the whole rest of the group.

READ MORE: 6 things to do before you hire more customer support staff

Before raffle: time-consuming chatbot technologies

And so Coop looked into ways they could automate their customer service processes where possible. Palle Esbensen explains:

“We decided to invest time and money into new technology. One of the things that was important was that we should find a system that didn’t require a large amount of effort for us.”

However, he found that this was not as simple as he had hoped. He spoke to many different vendors of chatbots and discovered just how much work was involved:

“Many of those we had a dialogue with needed an insane amount of time from us to define questions and answers, answer trees, and how we should build it.”

raffle Web: problem solved.

Coop eventually came across raffle Web. Palle recalls:

“After many conversations with different vendors, we found out about

What characterizes in relation to many of the others is the way they have built algorithms that more or less run by themselves. It generates its own answers that a robot couldn’t answer.”

Coop added raffle Web onto the “Shopping” section of the website, as well as on the members' area. A 24/7 service, it answers customer questions as they come in. Instead of long laborious chatbot-style conversations, it simply delivers the answers quickly, giving three options.


raffle Web on Coop Shopping

The results speak for themselves. Let’s take the Coop Shopping channel as an example:

Over the last 90 days, the tool has solved 65% of all questions — that’s a phenomenal self-serve rate when compared to the industry standard.

The remaining 35% were routed onto their customer service team, either via a contact form or phone.

With 38% of the customer requests, the customer actively opted to give the answer from raffle Web a big thumbs up — positive feedback for the answer given.‍

Coop’s data on raffle Analytics

Better service, greater data, stronger ROI

Emil Kolind, Project Manager at, told us:

“raffle Web has changed the way we work. Now, we allow our customers to get answers to their questions in a convenient way on our website. Plus, the quality of our service has improved now we have more time to answer more complex questions.”

And the benefits don’t end there:

“We also get comprehensive data from raffle Analytics on what our customers are asking and where we need to improve our customer-facing knowledge to improve the experience.”

It’s helped their bottom line too. Initial estimates from when they first adopted the technology found that, based on just a 20% reduction in simple questions coming through to customer service, they would save more than 120,000 DKK a month. With a greater reduction in calls, that saving is even higher.

What’s more, that figure is just for Coop Shopping, showing the huge potential for the technology across the organization as a whole.

For Coop, adopting a simple yet powerful AI solution has changed their perspectives on customer support. They no longer have to keep hiring more agents, but instead have found a way to achieve customer satisfaction using technology.

Want to achieve improvements in customer service like Coop?

Discover like many others how raffle Web cuts the backlog before it becomes a backlog connect with us below.

I want a demo