Analyze customer data and design strategies to maximize customer lifetime value (LTV).
── CRM advancement
Customer data is not just a record of the past; it's the starting point for future actions.
We analyze data to understand why customers act the way they do.
The same challenges appear across different industries.
There's a common complaint I hear repeatedly in the CRM field.
"We're competing with rivals for a limited number of customers. We can't see any way to expand our market."
"Even if we distribute coupons, it's always the same customers who take action."
"The evaluation of the initiative was based solely on the sales figures for that day. We haven't been able to track what happened afterward."
"We want to increase our loyal customer base, but we don't know what to do to achieve that."
"You only realize you've left after it's all over."
These are not separate issues...
The amount of data has increased,
Has your understanding of customers changed?
The root of these five challenges is one: we still don't understand what motivates our customers.
You can record what you bought.
The reason why they keep buying isn't clear from the records alone.
We're often told to "make use of the data."
However, those on the ground are overwhelmed with data collection, leaving no time to formulate policies.
Classify by attribute and distribute coupons.
Repeating the same process will not produce reproducible results.
Customer behavior has a structure.
We interpret data not merely as facts, but as a systematic framework of mechanisms that motivate customers. Our analysis is organized around three questions: "Who are the customers?", "How do they behave?", and "What determines customer retention?".
Segment optimization
This process involves defining customers as belonging to multiple stages, from light/casual to loyal, and understanding the size, characteristics, and membership requirements of each stage.
We define who we consider "royal" and in what contexts they use the service, creating a foundation for understanding subsequent transitions and designing KPIs.
Understanding the transition
This process involves understanding the user funnel from registration to retention, and the retention rate at each stage, based on the inflow channel and occasion.
Currently, we have a general understanding but lack the details. Our goal is to move from this state to one where we can understand user behavior in detail for each segment defined in the previous step.
Understanding KPIs
This process involves identifying intermediate KPIs that determine adoption and continuation, and structuring them into a KPI tree.
We derive action-based indicators from facts that are convincing to the field staff, demonstrating that "aiming for this will lead to sustained growth," and then establish a KPI system.
The quality of CRM data analysis
Change it to a management decision
Instead of competing with rivals for market share, we uncover untapped demand. There is room for cross-selling and market expansion in areas where products are related but haven't yet been purchased together.
We identify customers who will buy because of coupons, those who will buy even without coupons, and those who won't buy no matter what. We focus our resources only on customers who are truly receptive to discounts, thus protecting our profit margins.
Did the initiative merely generate short-term sales, or did it fundamentally change purchasing habits? We track repeat purchases without coupons to make investment decisions.
What needs to be done, and to what extent, to move middle management into the loyal customer base? We will identify behavior-based intermediate KPIs and the most effective measures to achieve them.
Which initial experience determines customer retention? We eliminate apparent correlations and identify the pure effect. We detect early signs of churn and identify customers who can be recaptured.
The examples listed here are only a part of our analysis menu. We design the optimal data science approach by combining the most suitable methods based on the client's challenges and existing CRM and ID-POS data.
The same data can be used in two different ways.
Customer data can be used to record what happened, or it can be used as material for designing actions. In many situations, the former is the primary use, with room left for the latter.
Attributes (age, gender, place of residence)
Context of the behavior (occasion)
Record what happened.
Deciphering the reasons why it happened
Rule-based system for amounts and attributes
Behavioral clustering
Sales during the policy period
Continued purchases after the implementation of the measures
