# Centum Product Plan FY2 S2 V1 # Introduction Centum Catalog is a database where shoppers can find long deep articles about specific topics in the context of shopping, either when actively searching for specific information or when passively inviting fun information to learn about. # Customer Problem Also see “Appendix: Target Customer Profile” in this document. When Alex wants to relax, he spends hours on his phone learning about various information in this world, from historical events to ingredients. For this purpose, he also enjoys “online window shopping” during which he learns about brands and products. He hates TikTok. In fact, he even tried explicitly to form a habit of watching short videos on TikTok, but he was just not able to appreciate the shallow content format. Due to lack of a holistic solution, he jumps from one surface to another: Google Chrome discovery feed, Reddit forums, Wikipedia articles, YouTube videos, blog posts. This current solution is not ideal for him, because the experience is fragmented (e.g. getting kidnapped from Google Chrome to Reddit app), redundant (e.g. seeing same information from one webpage and then again on another webpage), and inadequate (e.g. lack of tools for collecting notes). # Product Solution Ideally, Alex will have access to a single holistic solution that can satisfy his desire for interesting information with the convenience of modern technology. Such solution will exhibit following attributes: 1. Comfortable: He will be able to read for hours without feeling disrupted. When he cannot read on screen, he will also be able to listen to the contents. 2. Fun: He will be able to dive into topics he finds fun at that moment (as opposed to getting limited to topics recommended by AI), and study the aspects of the topic that he finds interesting (e.g. “tea” section but not “fabric coloring” section on “saffron” topic). 3. Actionable: He will be able to take notes or even make purchases, in doing so completing a task. Until now, developing such solution has been practically impossible due to the prohibitively expensive cost of generating so many deep articles tailored to each person. Because the topics of interest for people are so diverse (consider the differences between “saffron” and “Apple Vision Pro”), it has been not feasible for a human to write expertly on all the topics or for a machine to scrape information into a standard template (”display resolution” attribute does not make any sense for the topic “saffron”). We apply Generative AI technology to overcome this challenge. For a tail distribution topic, for which we are not able to apply common sense understanding or standard template, Generative AI will study that topic, then write a report, at a cost that is so low that we can write that report in every context, for every consumer. Content Generation AI will also make it possible for us to deliver that information in multliple formats, such as audio, for the convenience of the consumers. Our product, Centum Catalog, is a collection of documents that dive deeply into each topic for each person. Consumers who enjoy deep contents can dive into topics, and check out other related topics, growing their understanding of the world each minute. Moreover, reading passively is now also a productive behavior, because AI companion helps the consumer collect information into their own wiki, which itself becomes a valuable source of information for others who follow them. Centum will the best place for deep contents. ### Daily Habit Alex opens phone, checks recommended articles on Google Chrome discovery feed, then sees news about a new product called Rabbit R1. After reading the entire article, Alex still feels like he wants to learn more about it. Alex opens Centum, searches for Rabbit R1, and there is a single page article describing interesting facts about the product. In the article, words are linked to other standalone documents. Alex reads that Rabbit R1 has a bright orange color, opens “orange” document, reads that Hermes is famous for orange colors, opens “Hermes” document, reads that Hermes collaborated with Apple on Apple Watch, opens “Apple Watch Hermes Edition” document, reads that Apple Watch was ranked 46 in Fortune’s top 100 greatest designs of modern times, opens that ranking, opens “Sony Walkman TPS-L2” which ranked 5th. He is satisfied for the day, then sleeps. Alex exits home to go jogging at the nearby park. He puts on a pair of earphones, opens Centum, and starts listening to audio of the reports. ### First Discovery (Product Distribution) Alex encounters 1P contents on 3P contents distribution channels such as 1) 1P text articles on 3P search engines, 2) 1P text articles on 3P social media, and 3) 1P audio files on 3P audio or video content channels. After reading a couple articles on Centum, Alex starts coming to Centum in search of better reports on topics he is interested in. ### Member Conversion Visitor becomes Inactive Member to enjoy features that are exclusive to Inactive Members. Same for conversion from Inactive Member to Active Member. ### Focus We will make Centum be the best place for deep contents. In order to focus sufficiently on our target, we will choose NOT to focus on 1) timely information (e.g. daily news), 2) visually engaging images and videos (as opposed to informative ones), and 3) objectively verifiable truth. # Enterprise ### Roadmap We will deliver benefits to stakeholders in the following order. **Centum Catalog:** We will first operate a 1P site to immediately deliver interesting contents and to demonstrate how to use our technology. Once we prove that consumers desire such product, we will develop then deliver the product to the consumers. We will monetize the product using membership fee and commercial shopping. **Centum Mall:** We will add a community feature on top of Centum Catalog. Centum Mall, like a Korean mall with many expressive select shops, will let shoppers express who they are with their curations and verified purchases. Shoppers will be able to connect with other shoppers and find better shopping information on our private community (similar to how Naver private web has better shopping contents than Google public web). We will also provide tools to help community members create engaging contents such as videos of shopping sessions or product reviews. **Centum Economy:** Finally, we will power an entire economy, with foundational services such as finance (points), distribution (delivery), and production (AI bespoke factory), and find business opportunities in areas like market data and private currency. ### Finance Cost of service will come from: AI model inferencing, web scraping, data storage, content distribution network, community content moderation, content marketing distribution. We may consider following business models for profit generation: membership (subscription) (e.g. premium features), sales commission (transaction fee) (e.g. affiliate marketing, dropship, directly charging sellers), ads, services for sellers, and data sales (e.g. contents). ### Defense **Risks** We will face risks that we noted in Enterprise Plan FY2 S1 V1. **Competitive Strategy** - Older Entrants To older entrants who have many existing customers and a low-cost-low-income ads-based business model, the cost of generating so much text (an entire wiki per person) will be prohibitively expensive. Also, older entrants review past data to make decisions on whether to invest in developing new products. This decision-making process makes it impossible for them to enter markets with too few suppliers, because they lack data definitionally. There is an over-supply in short-form videos market and there is an under-supply in the long-form texts market, and this is why latter is a great business opportunity for us and a difficult business opportunity for older entrants. - Newer Entrants We will be able to provide a greater satisfaction to our customers thanks to economies of scale, economies of scope, better customer understanding, better community, better contents, and better brand. # Action Plan We will iterate on this product plan in Centum Product Plan FY2 S3 V1, focusing on our differentiation from currently popular content apps. # Appendix: Pivot from Santana Centum is a modified product plan in response to what we learned from testing Santana (See [Product Plan FY2 S1 V2](../FY2%20S1%20(23%2011%2001-24%2001%2001)/Product%20Plan%20FY2%20S1%20V2%20e31ec4affdce42238f06e9f82c7b75e2.md)). Whereas Santana demonstrated that Generative AI can write great contents related to shopping, we discovered that 1) we must first compile a great catalog of true information so that it can be a solid foundation for our other services, and 2) we must not force consumers to have an interactive conversation (i.e. answer guiding questions) before providing them with a report since doing so introduces high friction to the product experience and damages the fun for them (even if interactive conversation on the report after the report generation may be fun). # Appendix: Centum Origin Story During my recent trip to Asia, I studied shopping habits of Korean people. (Clarification on why I studied Korean people and clarification on my respect for American people: There are also many activities that Koreans can learn from Americans, and Asians do keep learning a lot from Americans in activities that Asians consider Americans to be better at, but I consider shopping to be an activity that Koreans are better than Americans at.) In fact, I chose the code name Centum for this future product after visiting Busan “Centum City”: the world’s largest department store today. (Context regarding department stores and shopping malls: In theory, a mall and a department store are distinctly different. A department store is a retail business that sells products on behalf of brands to consumers. A mall is a real estate business that rents out store spaces to brands so that brands can sell directly to consumers. In practice in America, a mall and a department store are visibly different, because a mall feels like a diverse collection of distinctive stores whereas a department store feels like one coherent space with many products. In contrast, in practice in Korea, probably because Korean economy and retail industry developed more rapidly and therefore in a shorter time than in America, department stores are practically indistinguishable from malls. Even in a department store building, the store space for each brand is separated clearly from other brands, even if the stores are decorated very similarly since all the decorations are done by the single department store and not by each brand.) Select Shops stood out to me most prominently. Explaining Select Shops is a difficult task, not because they are so foreign, but because they are so similar to the kinds of stores that are very familiar to us. In order to help you gain an accurate understanding of what the Select Shops are, instead of latching onto similar concepts that are already familiar to you (that is, to prevent you from saying ”Oh, by Select Shops, you mean [a similar but different idea] which I am already familiar with?”), I will 1) provide a shallow description of Select Shops, 2) contrast Select Shops to similar stores especially in America, and 3) explain why Select Shops are interesting. Select Shops sell a variety of products, usually consumer discretionary products, that cater to a target demographic. For example, Junwon Select Shop might reminisce the early 2000s, and sell ripped jeans, cargo pants, baggy shirts, beanie hats, Heelys shoes, wired earphones, Dragon Ball toys, and snacks that were popular back then. A consumer who likes one of the products there are likely to find other products interesting. This makes it possible for the consumer to shop effectively. The consumer can get all the different items that the consumer may need to complete a ‘way to live’, without having to explore a wide space, and the consumer can also discover interesting products that has a high chance of appealing to them. Comparison to other kinds of shops: - Brand Stores like Nike sell products made by one company. Brands often also distribute products that they do not directly manufacture, but regardless, consumers trust brands for their production abilities. Select Shops do not make products that they sell. Shops merely “selected” the products to be there. Customers trust “taste” of Select Shops. - Category Killer Stores like a shoe store that sells shoes from many brands sell one product category to all kinds of people. For example, indoor running shoes, outdoor running shoes, basket ball shoes, so on. Category Killer Stores are great for customers who know they need to buy “shoes”. Select Shops sell all kinds of products to one kind of people. It will not be unusual to find shoes next to umbrellas next to home diffusers. Select Shops are great for customers who are interested in what they can buy to elevate the way they live. - Boutique Shops and Variety Stores are most difficult to compare against, not only because they are similar to Select Shops in appearance, but also because definitions for Boutique Shops and Variety Stores are loose and accommodate many different kinds of stores. Boutique Shops and Variety Stores are names we seldom if ever wondered about for the stores we are familiar with: these are local stores that carry a wide variety of products. For now, let us say Select Shops are one special kind of “Boutique Shops and Variety Stores”, then further add clarification by describing what are interesting about Select Shops. Before I move on to discuss interesting aspects of Select Shops, let me clarify that I am not praising Select Shops to be superior to other store formats in every context. In this document, I am just writing down what I encountered, what I observed, and how I started thinking about what eventually became Centum. In creating Centum, we will focus only on what is best for our customers, without limiting us to only select features that are exclusive to Select Shops. Select Shops are interesting, because they are loved by people visiting the shop as well as running the shop, and these people are different from the people who run the space: - Shoppers love Select Shops. As a shopper, I found it fun to visit these Select Shops, because it’s fun to be in a small space packed with a bunch of products that appeal to me, and I can also feel a connection with the selector who has a taste that is similar to mine. Some Select Shops, especially those with online presence, attach story to the products that they carry: not stories about the products, but stories about them and the products. A consumer will be rational to take out information about the recommender and assess products for what they are, but for some reason, because consumers are not rational creatures, these personal stories make the products more appealing to me. The selections are also impartial recommendations. It is true that the shop earns money when a consumer purchases this recommended bag at the shop, but the shop was also able to give that store space to a different bag. An Hermes store can only “recommend” Hermes bags using their store space, but a Select Shop can choose to recommend an Hermes bag or a different bag or a product that is not a bag at all. This makes it possible for me as a consumer to trust the Select Shops. - Running Select Shops is also a fun work. Filling a Select Shop with products that one selects is essentially identical to expressing self. In the past, retail businesses like department stores had to pay professional merchandisers significant salary for the job of selecting products and decorating stores, because the complexity of handling orders and products outweighed the fun of selecting and decorating, especially when selecting and decorating are done not to express self but on behalf of the employer. From these observations, I made the conclusion that there is a business opportunity in helping ordinary people shop from, and also run, Select Shops. Sometimes in history, we benefit by removing middlemen: the existence of a big Zara store (which manufactures and distributes their own products for cost efficiency) in a big mall (which brings down the cost of Zara operating a store in my city) makes it possible for me to buy cheap products. But there are also times when opening up an industry to many small producers can benefit not just the producers but also the consumers. I am reminded of YouTube. Before YouTube, few big studios made all the videos that consumers watched. Thanks to a combination of innovations that include YouTube as an infrastructure for video contents but also smartphones as video production tools, we now get to live in a world where our choices as consumers are not limited to few videos from few producers. While giving consumers many videos to watch and creators the opportunity to express and communicate very easily, YouTube also established a sustainable business providing infrastructure and space for the video contents. I think people will benefit from living in a world with many small stores which not only give them a way to shop better as consumers, but also a way to have fun while expressing who they are, as long as we can take on the cost of infrastructure and space such that they can have all the fun and none of the complexities. We also have necessary innovations in the world now. YouTube needed iPhones for content creation, and we have Generative AI. From all this, I saw an opportunity to create a sustainable business by using Generative AI to make a mall for Select Shops. +: Malls in Asia are also interesting in that they became places not just for shopping but also for city life. Similar to how cafes in Asia are places not just for drinking coffee, but also for relaxing alone or meeting others. Inside malls, there are businesses like playgrounds for kids, beauty spas for moms, and libraries with books and parks with plants for everyone. # Appendix: Ideation - Comparisons | | Offline | Online | | --- | --- | --- | | Sparse | Mall | Web | | Dense | Select Shop | Junwon | | Our New Solutions | Comparable Old Solutions | | --- | --- | | Junwon Catalog | Search Engine | | Junwon Mall | Social Media | - CUT Idea Spaces [Centum Product Plan FY2 S2 V1 CUT Idea Spaces](Centum%20Product%20Plan%20FY2%20S2%20V1/Centum%20Product%20Plan%20FY2%20S2%20V1%20CUT%20Idea%20Spaces%20c523da7029354367be996e62e9c0199a.csv) - CUT Feature: Other Languages Only support English for now. Test translation accuracy by translating to Korean from time to time. Later support Chinese, Japanese, Vietnamese, Spanish, French, Italian, German, Portuguese, Russian… # Appendix: Target Customer Profile Alex is a member of the intelligentsia class. After studying hard in school, then contributing skilled labor to the economy, Alex got the money to spend on elevating his life to one with beauty and comfort. He fulfilled his hunger and he now seeks to fulfill his taste. His purchasing power is not limited by money but by information. He is now spending a lot of time each day learning about what he can be buying. Other potentially defining attributes for our target customers are the following: Has college diploma, skilled labor profession, earns $100K+ per year, spends more on discretionary purchases (fashion, comfort, fitness, education, fun) than on necessary purchases (housing, food, transportation, medicine, utilities), lives in a big city like San Francisco, Seattle, New York, Los Angeles, Chicago. We are not interested in shopper who are “shallow” or “cheap”.