Skofabriken Kavat AB has lots of loyal customers. They mainly sell children's shoes, although consumers also include many young adults and residents of large cities.
"We create shoes with a minimal impact on the environment, which is popular with a lot of young people. It's simple, stylish Nordic design, in which we make use of nature's own technology – based on leather," says Fredrik Widman.
Identify buying patterns
Even if Kavat has collected order information over the years, customer analysis has been neglected.
"Kavat hasn't had any automated communication with its customers. It's been based on campaigns and newsletters instead," says Hans Stenberg, Global Account Manager at PostNord Stralfors.
A new system is now emerging in a pilot project. With the aid of predictive analytics, Kavat aims to identify buying patterns among those who visit the company's website. This will make it possible to design individual, targeted special offers. First of all, the company's sales data from the last year was analysed in order to produce base statistics.
"The analysis took us a couple of steps along the way. After a number of algorithms had processed our statistics mechanically, the result formed the basis of how PostNord Stralfors then created the programs," says Fredrik Widman.
Based on individuals
There is currently a lot of information about consumer groups, often based on where and how we live. A young person who lives in the centre of a big city can be assumed to have one kind of buying behaviour, a person from a residential area another.
"That kind of information is based on how groups behave. Here instead we go down to the individual level. This provides Kavat with proactive assistance in analysing data, so that they can then communicate with the customer in the best way," says Hans Stenberg.
When shoes are sold via Kavat's website, the program analyses customers' behaviour and groups them into what are known as segments. Which ones are loyal customers, which are new and which risk being lost? When purchases are made, information is created in real time, around the clock. Every morning you can have an overview of what has happened.
"The great thing about mechanical analysis is that the more data is processed, the better it gets at producing good supporting data for decisions. In modern terms this is known as 'machine learning'. Our aim is to be able to experiment with targeted messages and see which message will probably achieve the best conversion rate for different customer groups before we've created the campaign. This is the future," says Fredrik Widman.
Behaviour generates information
Many types of customer behaviour can be captured to provide information. Customers are then segmented into clusters, and the behaviour results in targeted special offers. A loyal customer might be given a bonus scheme, a hesitant customer a reminder in the form of a postcard.There are structures for how, where, when and in which channels this is to take place. Everything has to work together.
"When we see where and to what the customer responds, we can better understand the customer."
When it is finally well segmented, Kavat will have gained a totally new insight and an amazing tool. When they are using our system, follow-up is possible in many channels. The parts of the system are like Lego bricks that we use to build familiar components, such as sales letters, emails, postcards and text messages. This makes it possible for Kavat to link up and use all our channels in their marketing communications and to combine quick digital responses with slower ones.
"The challenge is to get all the pieces into place. The element that's untested is the predictive part, but pretty soon we'll be able to identify exciting insights. It's then a matter of doing the right thing with the material the computer generates," says Hans Stenberg.
It is important for Kavat to be heard above the advertising noise without nagging the customer. Individual special offers are the key to success, believes Fredrik Widman. Good special offers and greater accuracy through individual analysis make it more interesting for the customer.
"Of course it's an advantage if some of our online sales go through our own web shop. It's important to be able to combine internal and external channels. If we only used external channels, our shoes would have cost more. A balanced mix means that the shoes offer more value for money. With greater knowledge of our customers, we can also streamline our offering in all distribution channels."