A rich product master that supported the Rakuten Shop of the Year award. Lazuli PDP is used for SKU expansion, SEO enhancement, and customer data analysis.

――Congratulations on winning the "Rakuten Shop of the Year 2024." How has this award been received internally and externally?

Thank you. We received the Grand Prize for the first time in the "Department Store and General Mail Order category."

Two years ago, when I attended the awards ceremony, I looked up at the podium and felt regret. I told myself that merely looking up from below wouldn’t do; I would definitely get up on that podium.

While working in e-commerce, I have acutely felt that Beisia is not a nationwide brand. For customers in Hokkaido, Tohoku, and western Japan, even saying "Beisia Rakuten Ichiba Store" is just a mysterious company name. I definitely wanted to put the ribbon of the category grand prize on our site.

Actually, prior to my appointment, we had received a category award in a different category, but this time in the "Department Store and General Mail Order" category feels more like Beisia.

Our e-commerce strategy is to expand SKUs. A strategy of trying to sell anything and everything haphazardly is considered a poor strategy in supermarkets. However, I have believed that expanding SKUs is essential for increasing sales in e-commerce. Since June 2022, before introducing the Lazuli PDP, the number of SKUs has increased by tens to hundreds of times. As a result, even just isolating net shopping sales, they have increased fivefold.

Naturally, registering products for sale in the online shop requires product registration work. Lazuli PDP, which can automatically collect and manage product data, was a strategically essential solution.

(Mr. Tomozo Tode, Head of EC Department, Beisia Marketing Division)

SEO Enhancement and Increased Organic Traffic Through AI Tagging

――What other initiatives have you undertaken over the past three years besides increasing SKUs?

We were able to purchase products and register product masters using the PDP, but beyond that, how to attract customers was lacking in our marketing efforts. We hired specialized personnel and devised ways to attract customers tailored to each platform.

We also focused on SEO. In SEO, enriching product text is vital, and Lazuli PDP enabled us to enhance the information.

The category that particularly stood out for me is the snack category. It was a genre we hadn’t handled before, and since the unit price per item is low, I didn’t have high expectations. We didn’t particularly promote it, but once we organized the product information and tagged it, organic access began to increase naturally. Now it occupies a certain portion of sales in the food division.

What was particularly interesting is that the large set of "Umaibo" sold more than expected, and comparing it to when we hadn’t promoted it, sales increased by five to six times.

When tagging was done using AI, the tag "Barbecue" was assigned. Tags such as "Festival" and "Party" for "Umaibo" could likely come to mind for humans, but "Barbecue" would be difficult without AI. It certainly sounds fun to have various flavors of Umaibo at a barbecue.

We are also making detailed adjustments to our pricing strategy.

Products that sell well in e-commerce are also carried by other stores, so we compete by keeping prices down as much as possible. On the other hand, for products that other stores do not carry, we can take a certain margin. While balancing low-profit products that expand scale and products that secure profit despite lower sales volume, we have constructed it to be viable as a business.

For example, pet supplies, which we have targeted as a key area, are now in intense price competition. They are products that sell well but have a small margin.

Conversely, surprisingly few competitors exist in frozen foods. Many products cannot be handled without a freezer, so they do not compete with those of other stores.

Improving Accuracy of Product Information and Trustworthiness of the Team

――At the beginning of the implementation, we received feedback that the accuracy of the product information extracted by Lazuli PDP was "not 100%." It is designed so that AI learns from information corrected by humans to improve accuracy; has there been any change in accuracy in your use-up to this point?

Product information is ultimately checked visually by the person in charge. Recently, the number of times we request corrections has significantly decreased.

I feel that Lazuli PDP’s performance has improved, and we have begun to trust it.

New Business Development Arising from the Product Master and Establishing an Internal Data Utilization Base

――Thank you. Have you noticed any changes in awareness within the company regarding the management of the product master?

Starting from April 2025, we began handling Beisia’s private brand (PB) products at "Rakuten Mart," which operates an online supermarket. Although this is essentially a wholesale operation for us, outside of EC, I took on the project management role. There were many points where the know-how accumulated over the last three years in EC could be utilized, and regarding the product master, what we built with Lazuli seemed usable.

While it is easy to wholesale PB products, that alone would not suffice for Rakuten Mart. Even when products exist on the internet, if there is no product information, they cannot be displayed in the sales area.

Three years ago (before implementing Lazuli PDP), it would have been impossible to handle it at Rakuten Mart. It was because we had a well-organized product master that we could advance the project in an extremely short time.

When we implemented Lazuli PDP to increase the number of SKUs in EC, I never imagined it would expand into such a business. I have come to reaffirm that, regardless of what we do, the product master will always be a key factor, starting from the expansion of sales channels. If there are deficiencies in the product master, it will inevitably cause a stack somewhere.

In this project, employees from various departments must have understood the importance of the product master. We processed data from Lazuli PDP into Beisia’s unique product master and constructed PIM (Product Information Management) in-house. I believe we have mostly built a base of data created for EC that will be used across various departments for various purposes.

――Can you tell us about your future outlook?

Within the EC team, we say, "Let’s do CSV (Creating Shared Value)." It’s a management approach that benefits the seller, benefits the buyer, and benefits society.

I joined in 2022, having come from sales of several hundred million yen, and recently we have come to the point where we can set a sales target of 10 billion yen. However, even if we achieve that target, within a company with total sales of 300 billion yen, it is just a small part. I believe it is important not to chase numbers but to do good for our customers and society, leading to a state where profits naturally arise.

For instance, from purchase data, we have slowly begun to see the customer profile of what they seek. We will be able to provide tailored services, such as recommending products for each customer.

However, just because customers seek an individually optimized environment doesn’t mean it’s the only thing they want. The smell of fish in real stores and the sounds of children crying and throwing tantrums in the candy section represent real-life experiences and the joy of shopping. While it’s important to merge online and offline, I want to maintain the uniqueness of both real stores and e-commerce. I believe this embodies Beisia’s unique OMO (Online Merged with Offline).

――Finally, please share your requests and expectations for Lazuli.

There are other tools that can be leveraged, like Google’s Vertex AI. What distinguishes Lazuli is that it resonates well with Beisia’s specific circumstances. As I mentioned earlier, the groundwork for analyzing customer profiles from data has been laid. I would like to work together on how to implement this in future initiatives.

———————————————————————————————-

Beisia Rakuten Ichiba Store: https://www.rakuten.ne.jp/gold/beisia/
Beisia Net Supermarket: https://netsuper.rakuten.co.jp/beisia/
Past efforts of Beisia’s Lazuli PDP: https://corporate.lazuli.ninja/useful/beisia/
What is Lazuli PDP: https://corporate.lazuli.ninja/features/