Demo
Could you help us understand KIA’s CRM approach in Turkey operations?
There are two fundamental situations that we can perceive as both advantages and disadvantages in the sector we are in. One is the financial value of the product. That is, as you know, financial costs are quite high in Turkey.
How did you create the scorecard and what does this card include?
We needed to evolve every point of contact with the customer into an evaluative structure with scorecards. In the meantime, we started looking for answers to how we could effectively manage the process that leads from the customer saying “I want to buy a car” to entering a showroom and meeting that car. We have a large data pool inside. We know which vehicles the customers are interested in. We integrated the machine learning tool here.
In the second part, some customers have bonded with their vehicles. These customers occasionally visit our authorized services. They should also create value for us. We have created a scorecard by blending this into a more complex structure, looking at when they last interacted with us, and how often they interact. We created two different scorecards covering both the pre-sales process and the post-sales period.
We determine the value of the customer inside, but if we don’t touch the customer, they don’t have any value in the eyes of the customer. Therefore, we trigger our customers with different content both through mass communications and automation campaigns.
What does dynamic segmentation mean for you? How do you apply this in CRM?
This morning, I overheard a gentleman’s phone conversation on the way here. The gentleman was discussing with his friend how to get from Kadıköy to Beşiktaş. First, he decides to come by ferry. 20-25 seconds later, he said, “I’ll come by car.” Then his friend said, “You won’t find a parking slot.” Somehow he convinced him. Then he said, “Then I’ll come by taxi.” In other words, he changed his mind three times within 1 minute. If we start from here, we can see that the customer’s preferences change very quickly. We must also adapt to these changing customer preferences at the same speed, even beyond. In that sense, we positioned dynamic segmentation here. When creating segmentation, what data do we have at our disposal; first of all, we have a database. We know their past behaviors and their interactions with us.
On one hand, we run a machine learning algorithm. We also create various footprints both in our offline network, that is, in our dealerships, and digitally, and follow them. Based on these, we create dynamic segmentation. This is a matter of preference; you can do mass communication, it won’t be different from advertising on a radio or putting up a billboard because you may have given a message to maybe a million people at once. This message may represent a crowd worth 10,000 or 20,000 for you. The remaining people may have been exposed to this ad in vain. By applying dynamic segmentation, we try to offer the right product to the right customer at the right time. While doing this, we also try to communicate through the right channel. You may present the right product to the right customer in some way, but perhaps the channel you present it through is wrong. Not everyone may read emails or may not have a high interaction with pushes, but when you communicate through the right channel, you both turn the crowd into satisfied customers and reach the outputs you want to achieve more effectively.
