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

• The Potential of ID-POS: The Key to Grasping Customer Shopping Behavior (Part 1)

ID-POS enables the identification of challenges by capturing and analyzing the reality of customer shopping behavior. This data is extremely important for those involved in retail, as it allows for the consideration of various initiatives. However, it is said that if the analysis method is incorrect, it may not be utilized effectively.

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 consulting experience in retail and distribution and promotional support for manufacturers; Mr. Hideki Sugimoto of Dentsu Retail Marketing Inc.; and Mr. Shingo Asano. In Part 2, we discussed the precautions for ID-POS analysis and its potential applications.

Failing to analyze data holistically risks missing the true issues

Q. Are there any common pitfalls to avoid when conducting data analysis?

Sugimoto: I believe the most crucial point to be mindful of is "viewing the whole picture." For example, when analyzing why a particular brand's sales are declining, we tend to dig deep into trends like shifts in the purchasing demographic to pinpoint the cause or key issues. However, stepping back and looking at the store's overall sales might reveal that it's simply a case of a decrease in the total number of customers. When you delve into brand-related matters—like whether a commercial aired, its quality, or competitor actions—various hypotheses emerge. But the fundamental issue of fewer customers visiting the store occurs before these factors come into play. In other words, the key is not to jump straight to examining "points" (specific details), but to break things down progressively from the top, like an inverted triangle.

Another common pitfall is analyzing data without a clear purpose, just to say, "We ran the numbers." You get results, but then it's like, "So, what do we actually do now?" You end up staring at the analysis, feeling vaguely frustrated, satisfied that you've done some analysis, and that's it. Unless you approach analysis with the intent to make a decision, the results won't lead anywhere. We've often been asked by client companies during our reports, "So, what should we do about it?"

Integrating various data sources strengthens personalized approaches.

Q. ID-POS data captures in-store purchases. Can we integrate and analyze online purchase data alongside it?

Sugimoto: Integrating offline and online data has become increasingly common in recent years. It's straightforward if there's a common key across the data sets. Integration reveals customers' patterns of using online versus offline channels. Adding location data and app information beyond just purchase data allows for a more detailed view of customer behavior. However, when there's no common key between online and offline data, integration sometimes involves some degree of inference.

Asano: There are two main approaches to data integration: "definite" and "estimated." The "definite" approach uses a key, like an email address. If the email addresses match, we can say they belong to the same ID. When no such common key exists, we attempt integration using the "estimated" approach. Developing the logic for determining "same person" in the estimation-based approach can be quite challenging. To improve accuracy, we must refine the underlying data, which inevitably requires significant effort.

Q. How are the analysis results actually utilized, and how do they inform subsequent actions?

Sugimoto: The areas where analysis results are applied are diverse. Examples include "underperforming store countermeasures," "product development," "targeting," "competitor research," and "price revisions." Ultimately, they feed into management decision-making. For instance, in retail, we first provide management with a preliminary analysis visualizing the current state. This analysis is conducted across four axes: "entire chain," "store-based," "product-based," and "customer-based," informing subsequent actions.

For one restaurant company, we conducted an effectiveness verification of "seasonal products." We examined limited-time items launched at several stores to determine if their pricing and presentation were effective, providing insights for planning future renovations.

Analyzing sales rankings, purchase rates among visiting customers, repeat rates, and average spend per customer revealed that customers were willing to pay a premium for these items. This meant there was no need to discount them. Furthermore, purchasers spanned a wide range of ages and genders. This indicated that presentation targeting all ages and genders, rather than specifically "women" or "young people," was effective. Additionally, we found that a high proportion of purchasers were loyal customers. Furthermore, many customers purchased it alongside drinks or other staple items.

These analyses revealed several possibilities: expanding extensions likely to please loyal customers, offering more value-packed sets with drinks and staple items, or increasing toppings while raising prices accordingly. Implementing the renewal based on these insights resulted in achieving four times the planned sales.

Q. Finally, could you share your outlook for future ID-POS utilization?

Sugimoto: ID-POS data represents actual purchase history and is extremely valuable real-time data. The trend of integrating this data with other sources will only accelerate. Thinking in terms of IDs fundamentally means considering the individual, which will inevitably change marketing approaches. Rather than targeting broad demographics with ads, we'll likely see more personalized approaches focused on stimulating existing customers.

Konuma: Moving forward, we'll see not only data connecting with other data, but also data connecting with devices and tools. Linking data to smartphones enables approaches like "targeting people who bought this product." Or, connecting not just to smartphones but also to digital signage allows for real-world approaches. From our perspective, we'll implement and combine various strategies within the viewpoint of "maximizing the overall effectiveness of promotions." This ID-POS analysis is a crucial factor within that framework. There are various customer journeys leading to a purchase and continued buying. We have extensive experience and strengths in capturing these journeys and responding by pinpointing key moments. Within this, we aim to be a group that can maximize promotional effectiveness by deploying various tactics—utilizing SNS, using LINE, and employing every possible approach—to continuously connect with customers across the entire funnel. We want to be of service to you all.

As data utilization advances further—such as integrating ID-POS data from physical stores with online purchase data—the possibilities for approach methods will diversify. However, the most crucial element remains the mindset of "seeing each customer as an individual." The expansion of ID-POS utilization will likely lead to a greater focus on the individual customer. The key to a company's growth always lies within its customers. Why not start by accurately grasping the profile and needs of your own customers, and then re-evaluate your company's approach and behavior?

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

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