The Potential of "ID-POS" Utilization: The Key to Understanding Customer Shopping Behavior (Part 1)

Have you heard of "ID-POS"? Many people might say, "I've heard of POS, but I'm not sure what ID-POS is." In retail, ID-POS is extremely important data for achieving data-driven marketing. By analyzing it, you can grasp the reality of customer shopping behavior, connect it to the next action, and consider various initiatives.

This time, we interviewed three experts: Mr. Atsushi Konuma of DENTSU PROMOTION PLUS INC. (formerly Dentsu Tech Inc.), an ID-POS analysis specialist with extensive experience consulting for retail and distribution companies as well as supporting manufacturers' promotions; Mr. Hideki Sugimoto of Dentsu Retail Marketing Inc.; and Mr. Shingo Asano. In Part 1, we first asked Mr. Sugimoto about the fundamentals of ID-POS, its analytical perspectives, and methods of utilization. This should be particularly valuable for those involved in the retail industry.

It reveals "when," "where," "who," "what," "how much," and "at what price" a purchase was made.

Q. This is a basic question, but what exactly is "ID-POS"?

Sugimoto: First, are you familiar with "POS data"? When you shop at a store, your receipt prints details like "when," "where," "what," and "how much" you bought. That data, in its raw form, is "POS data." ID-POS data takes this a step further by adding information about "who" made the purchase. When you use a membership card or points card while shopping, it identifies "who" made the purchase, right? By assigning the purchaser's ID, we can determine "when," "where," "who," "what," "how much," and "at what price" the purchase was made.

In my impression, around 2006, so-called "point cards" started circulating widely, and the style of shopping to accumulate points became established. Along with this, "ID-attributed POS data analysis" began to be performed. Furthermore, in the last few years, it's become mainstream to use point cards based on other companies' platforms, like Rakuten Points or d Points, or app points, rather than just a company's own house card.

Q. You mentioned that ID-POS reveals "who" made the purchase. How detailed is this "who" information?

Sugimoto: POS data shows how much was sold and whether it exceeded the previous year's figures, but it doesn't reveal "who" made the purchase. By attaching an ID, depending on the level of information collected, we can determine basic details like age, gender, occupation, and annual income. Additionally, you can see how many are new customers buying that item for the first time, how many are repeat customers buying it again. Or whether they are high-value customers who buy a lot, or churn customers with no purchase history this year. It allows you to comprehensively grasp shopping behavior information.

Identifying challenges, their causes, and key areas for intervention from customer shopping behavior

Q. I see. So, what kind of analysis do you perform using ID-POS?

Sugimoto: The fundamental premise is that the core purpose of analysis is "understanding customer purchasing patterns." The basics are per-customer metrics like visit frequency (how many times they visit in a year) and purchase amount (how much they spend). Logically breaking this down further leads to identifying issues. For example, if sales are declining, is it because the average purchase amount per customer is decreasing, or because the number of purchasing customers is decreasing? If the number of purchasing customers is decreasing, is it because fewer new customers are coming in, or because repeat customers are no longer returning? If fewer new customers are coming in, what exactly are they no longer buying? We logically drill down the analysis like this to identify issues.

The other is "zero-level analysis." This involves starting with a flat, initial analysis to thoroughly identify, through data, what kind of customers are coming and what they are buying. Even if you think you intuitively know your best-selling items, looking at the actual data often reveals surprising discoveries or exposes misconceptions. When analyzing a restaurant business, we found that not just the main menu items, but the so-called "topping items" were a major driver of sales. This insight changes how you structure your menu and how you promote items in-store.

Crucially, this analysis should connect to CRM (Customer Relationship Management). We categorize customers into groups like "Loyal Users" and "Light Users," then examine their distinct purchasing behaviors. Among them, we might find someone who was a "Light User" last year but became a "Loyal User" this year. So, what purchasing actions led that customer to become a Loyal User? Focusing on that reveals pathways to help more customers become Loyal Users. Identifying such opportunities is the job of us data analysts.

Q. You've analyzed various types of data. Do the perspectives or approaches to analysis differ by industry or business type? Or do you apply the same analytical perspective across all industries?

Sugimoto: The methods and concepts for customer analysis don't differ significantly across industries or business types. The analysis itself is largely the same. However, there are cases where the analytical perspective differs.

Even within retail, the target area changes between "trade area analysis" for large commercial facilities and drugstores. For drugstores, the trade area is typically 1-2 km, focusing on whether nearby residents are visiting. For large commercial facilities, we look at how much of the trade area they capture and how much foot traffic they attract from beyond 5 km.

Or, when analyzing convenience stores, we break down customer movements by specific days of the week and times of day. For drugstores, there's usually not a huge difference in what sells well across days or times. But for convenience stores, what sells can be completely different on weekdays versus weekends. Furthermore, depending on the store's location, the customer base and sales patterns can be entirely different. Therefore, analysis needs to be conducted while considering the characteristics of each time slot and scenario.

Analyzing ID-POS data reveals why customers buy certain products, why they don't buy others, and what situations make them more likely to purchase. Crucially, it also highlights differences between "high-value customers" and "non-high-value customers." This distinction may offer clues and angles for cultivating high-value customers. Doing so opens possibilities beyond simply attracting customers through "advertising" or "special sales," revealing diverse approaches. In the second part, we'll hear from Mr. Konuma and Mr. Asano about key points to consider in ID-POS analysis and the further potential for data utilization.

• The Potential of ID-POS: The Key to Understanding Customer Shopping Behavior (Part 2)

*Dentsu Tech Inc. changed its name to DENTSU PROMOTION PLUS INC. in April 2022.

※Affiliations and titles are as of the time of publication.