The Future Created by Data Integration Between Retailers and Manufacturers Vol. 3 ~The Value of Data Integration~

In Chapter 3, we will examine the purpose of using data and what data collaboration should ideally look like, and consider what value data collaboration ultimately brings and what kind of future it creates.

What Are We Trying to Achieve by Using Data?

Big data, data analytics, data science, data scientists, data mining, data... Right now, the world is overflowing with terms like "data something," but what exactly are we trying to achieve by using data? In retail, it is often said that "sales are a measure of customer approval." If you use data to increase sales, you could also rephrase that as using data to improve approval from customers.

However, the expressions "using data to increase sales" and "using data to improve approval from customers" create slightly different impressions. I would like readers of this article to use the latter expression, "using data to improve approval from customers." Why? Because if you think only in terms of increasing sales, the existence of the "customer" tends to be neglected.

For example, when sales of a certain product do not increase, many people tend to think, "How can I increase sales of this product?" When that happens, they often resort to superficial measures, such as price discounts or other simplistic sales promotions. On the other hand, if you ask, "Why isn’t this product earning customer approval?" you will first try to think about the issues with the "product" itself. If the product itself has problems, spending money on sales promotions is absurd, yet that is actually happening quite a lot in reality. Sadly.

By using data to improve customer approval and, as a result, increase sales, the author believes it is extremely important to aim for a state that is good for all three parties, or as the old saying goes, a "win for seller, buyer, and society."

What Retailer-Manufacturer Data Collaboration Should Ideally Look Like

As explained in Chapter 1, the current data at retailers and manufacturers is optimized in isolation, and to call it data collaboration would be flattering. In other words, on the retail side, each company builds and operates its own product master, and as the data that can be collected expands, retailers with room for investment are, under the banner of DX, adding data in ways that make data scientists cry. On the manufacturer side as well, through building and operating DMPs (short for Data Management Platforms) via direct sales on their own e-commerce sites and web advertising, they collect various marketing data, conduct different surveys on products and brands, purchase anonymized POS data from retailers, and so on—again making use of a wide variety of data.

As described above, although data collaboration has begun—for example, by providing anonymized purchase data from retailers to manufacturers—it is still only a small part of the whole.

Of course, the author is not denying that retailers and manufacturers should hold individually optimized data to increase their own sales. Rather, I support each company having its own individually optimized data, and I think this is necessary from a corporate competitive strategy perspective.

On the other hand, I see the fact that data collection currently costs too much as a major problem. For example, in the case of information related to product evaluation, manufacturers confirm this through research. If a research company is used to survey customers who actually purchased a certain product, and a certain minimum sample size is to be secured, costs of several hundred thousand to several million yen will arise. Of course, this cost is added to the product price and ultimately paid by the customer, the consumer.

What would happen if, instead of conducting such surveys, retailers’ purchase data could be linked and used? Wouldn’t it be possible to obtain data that can serve as a substitute for research to some extent—who is buying it (gender, age group, values, etc.) / when they buy it / what they buy it with / how frequently they buy it / what competing products there are / whether they are repeat buyers (repeat rate, etc.) ...?

Also, what if the data used in the product master provided by manufacturers to retailers (the immutable information described in Chapter 1) could be fixed to some extent? Wouldn’t that save the effort of processing data prepared by the manufacturer side for each individual retailer, and reduce labor costs?

If data collaboration between retailers and manufacturers progresses, there is no doubt that the costs associated with data collection, as described above, will be greatly reduced. If that happens, the reduced costs will naturally be reflected in product improvements and product prices, which in turn will lead to "improving customer approval."

In the end, returning to the original purpose of "using data to improve customer approval" is, in the author’s view, the ideal form of data collaboration.

What Is the "Value" of Data Collaboration?

The author believes that the value brought by data collaboration is: 1) improved product value, 2) cost reduction, and 3) improved consumer experience. Below, I will explain each of these values.

Improved Product Value
I can say from experience that data expanded through retailer-manufacturer collaboration has a major impact on product development and improvement. For example, convenience stores develop and sell many original products. In a sense, this is a state where retailer and manufacturer data are integrated within a single organization.
Therefore, when developing original products, various integrated data are used to determine the product’s target audience, and then the materials, flavoring, packaging, volume, price, and even naming are considered to match that target. The developed product is then test-sold in certain areas, fine-tuned, and finally launched.
I believe that using this kind of data efficiently leads to improved product value and is one of the major reasons why convenience stores produce hit products year after year.

Cost Reduction
Regarding cost reduction, I mentioned in the previous section the costs associated with product research and the costs manufacturers incur when providing product information to retailers. Beyond that, for example, there is also the possibility of reducing the costs of new product development. As mentioned in ① regarding convenience store product development, if expanded data is available when considering a product concept, the cost of preliminary research can be reduced. More fundamentally, I believe that the sheer number of new products is itself a problem. This problem is influenced by a business custom that could be called "new-product supremacism." In particular, convenience stores have long continued a method in which new products are launched every week to keep their limited shelf space constantly fresh and continue increasing sales, and that tendency still exists today. However, in reality, looking at various data, sales growth per store has plateaued since around 2000. If you take cigarette price increases into account, it is quite possible that sales excluding cigarettes are actually declining. When developing and launching new products, manufacturers incur enormous costs such as equipment investment and advertising expenses like TV commercials, and needless to say, all of these are passed on to the sales price. If data collaboration progresses, increasing the probability of developing products that will sell, and if a system can be built that identifies from data products that are likely to sell with a bit of refinement and then improves and nurtures them, the author believes even greater cost reduction can be achieved.

Improved Consumer Experience
The content discussed in ① and ② above directly leads to improved consumer experience. In other words, consumers will be able to buy products with higher value, or products at more appropriate prices than before.
Even more important, the author believes, is that the probability of "encountering a better product for oneself" increases. In other words, as retailer-manufacturer data collaboration progresses and individual product information is expanded, consumers will be able to reach products that match their preferences from a variety of keywords. From the retailer’s and manufacturer’s perspective, the various values of those products become visible, and by leveraging them for so-called one-to-one marketing and similar efforts, they can present products to customers who have a strong need for them. As a result, consumer experience improves.

Epilogue

With that, this article concludes with "Chapter 3: The Value Brought by Data Collaboration." How was it? Needless to say, data collaboration is not the goal but a means. The point is to collaborate data in order to improve customer approval, analyze more highly detailed data, build high-accuracy hypotheses, and take action.
However, there are various obstacles that must be overcome in order to collaborate data. As one means of solving them, services such as "Lazuli PDP" exist. I think using such services as a hub function for data collaboration is very reasonable.

If I were to express my personal hope, it would be for a data-hub-like service that can be jointly operated by retailers, manufacturers, and wholesalers, and that also includes a product master capable of providing part of its product search functionality to consumers for the purpose of collecting additional data. Wouldn’t that be extremely convenient!? That is what I think.

Last but not least, I sincerely hope that in the future, retailer-manufacturer data collaboration will advance, and as a result, many consumers, including myself, will be able to encounter wonderful products and services they have not yet discovered.

The End