In the past, the customers relied on a salesperson, for instance the owner of a convenience store, to help them find what they were looking for. Based on their experience, the salespersons were able to quickly find the perfect product for the customer, and often suggested additional items that the customer wouldn’t have even thought of. In the past, the owner of a local store had a few dozen, maybe a few hundred customers, so knowing their needs and treating them in a personalized way was feasible for them. Today, the brands have tens of thousands or even millions of customers who, like customers in the past, expect a personalized approach.
The advances in data collection and analytics make it possible to deliver a similar experience. Using increasingly detailed data, the companies are beginning to create highly personalized offers that direct consumers to the right products or services, at the right price, at the right time and in the right channel. However, treating each individual customer in a personalized way, when you have tens of thousands of customers, is very costly and at the end of the day may not be profitable. The consumer segmentation may help and it will likely provide the greatest value in terms of return on investment.
Nowadays, the machine learning models accurately create groups of the most similar customers, with each group being significantly different from the others. They should be similar in size so that they can be handled as efficiently as possible. The segmentation can take different forms and focus on different aspects due to the needs of the company and the business goals pursued. The article presents three ways to group customers relevant to a loyalty program.
RFM Analysis – The First Step To Customer Segmentation
When you are considering a loyalty program, it is worth including a RFM analysis in pre-implementation activities. Based on the parameters of Recency (the period since the last conversion), Frequency (the frequency of transactions) and Monetary Value (the value of the transactions), you will learn about the behavior of consumers and what their flow looks like over time. The advantage of this approach is, first of all, the simplicity and ease of data interpretation, so you can get to know what percentage of customers generate the turnover, at what time, and what percentage of turnover is generated.
We had the opportunity to lead an interesting project for a certain retail company. The conclusions of the RFM analysis for that company clearly showed that the group of occasional customers, who rarely visited the store and spent medium or low amounts, was massive and thus they represented a significant share of the store’s turnover. The need to engage them in a loyalty program challenged us to develop a creative mechanism that would be attractive from the point of view of both regular and occasional customers, while being feasible for the retailer in terms of the cost and its business. If we hadn’t done the RFM analysis, the proposed standard points and cashback mechanism would actually be unattractive to most customers, i.e. the potential program participants.
We started looking for an alternative. Something to complement the traditional mechanics. A program similar to a family account turned out to be the perfect solution. The elements such as reminders of important events (birthdays, name-days), the option to create gift lists by the account administrator, product recommendations based on the transactions of all account users, and special price offers on recommended products made the occasional customer see that it made sense to join the program. In addition, all transactions credit the joint account with points, so even occasional customers can add their contribution for the benefit of the family. With such benefits, the participation in the program is not just about commercial value, but consists of many emotional connections between the customer and the brand.
Based on customer segmentation, you can develop the most effective loyalty program mechanism ensuring that the customer value will increase and will be maintained over time. After all, it is not difficult to come up with innovative mechanics, attractive for a moment to a few percent of the customer base generating a small portion of sales. The trick is to get to know your customers, understand their needs and engage them for as long as possible.
Not Everyone Is Just Waiting For Promotions
We were faced with another challenge from a fashion company that was building a loyalty program based on the belief that customers only count on discounts and promotions. As the size of the customer base increased, the cost of promotion and direct communication increased faster than the ROI. Moreover, the company collected a very selective set of data on its consumers, concerning only the type of category of the product purchased and the offer it came from (new offer/seasonal promotion). Nonetheless, we were able to divide customers into groups, which drastically changed the perceptions and showed how the consumers differ from one another.
The segmentation proposed was very simple. What was unique was the fact that we were able to create attractive segments from the point of view of the communication, marketing and company business based on such a small range of data. After all, in addition to the customers who were genuinely interested only in the products on promotion, there were several other, equally numerous groups of consumers who had completely different needs. For some customers, a new offer, the highest quality or a specific category often not associated with a discount at all, were more important factors. Among the consumers, we identified the segments that purchased the following product categories, among other things:
- children’s products
- products from the new offer,
- fashion accessories (excluding core products),
- winter products on sale,
- high-quality products without regard to promotion.
Thanks to such segmentation, the client knew what products to talk about, when to talk about them and what aspects of the offer should be highlighted in the loyalty program communications, so that the messages were contextual, and tailored to the needs of the program participants.
The sales are not always driven by the promotions and price offers, but unfortunately some companies mistakenly identify the loyalty program mainly with discounts. A price offer can easily be outbid by competitors, and contextual, inspirational communication and tailored offers are already much harder to copy.
In Search Of The Perfect Customer
When working for a client in the consumer goods industry, we were looking for the holy grail of a segment or set of features representing the ideal customer — willing and able to spend a lot of money. However, the analysis of a very wide spectrum of data: website visits, products viewed, communication interactions, NPS survey results, transactions, returns, complaints and many, many other data, did not have the desired effect. The search for the holy grail has been unsuccessful. What is the reason for this?
In some industries, the customers are activated periodically, which makes finding the ideal customer and their characteristics less important. A customer who behaves inconspicuously can turn into a super-customer. Their interest in a brand changes depending on the consumer’s situation. In this case, it is important to recognize the customer life stages and skillfully anticipate the transition from one stage to another. However, this is not easy to do based on transaction data, as they don’t take into account the social aspect of the customer’s life.
Sociological quantitative and qualitative research came to the rescue, helping to identify behavioral changes related to customers’ lives. This helped us to identify four main patterns of customer behavior:
- hibernation (no transactions),
- small projects (small, occasional purchases),
- medium-sized projects,
- large projects (a big change in life, requiring large expenses).
It was difficult to assign customers to the various segments. In this situation, the loyalty program came to the rescue. Thanks to the meticulously collected data, we translated the insights from the analysis into specific customers, which made it possible to identify the customer life stages and forecast the changes and the movement of customers between various stages.
Understanding the customer life stages helps create a personalized approach and find answers on how to treat a particular customer and what to offer them. Knowing that a customer’s behavior indicates that they are likely to spend 50x more than the average customer, you should seize the opportunity and offer them a special customer service process so that they fulfill most of their needs in your store.
Monitoring variables such as website visits, the products viewed, and correlating them with NPS survey results makes it easier to recognize which life stage the customer is moving from or moving to. It also enables proactive action and allows capitalizing on the opportunities the customer brings to the company. Good segmentation provides a basis for prioritizing business strategy: what is the condition of the customer base by specific store, to what extent does an outlet, with its sales goals, need to focus on attracting new customers, and to what extent on improving the customer service?
The customer segmentation models presented in the article are only selected examples from the array of marketing and analytical capabilities available on the market today. While meticulously collected data should be the basis of any segmentation to make the identified segments useful to the business and so that they meet the company’s marketing goals.